Remove ad_spdif from the normal codec list, and select it explicitly.
One goal was to decouple this from the normal codec selection, so
they're less entangled and the decoder selection code can be simplified
in the far future. This means spdif codec selection is now done
explicitly via select_spdif_codec(). We can also remove the weird
requirements on "dts" and "dts-hd" for the --audio-spdif option, and it
can just do the right thing.
Now both video and audio codecs consist of a single codec family each,
vd_lavc and ad_lavc.
Conceptually cleaner, although the API claims this is equivalent.
Originally, AVCodecContext fields were used, because not all supported
libavcodec/libavutil versions had the AVFrame fields.
This is not done for chroma_sample_location - it has no AVFrame field.
Helps with gif, probably does unwanted things with other formats.
This doesn't handle --end quite correctly, but this could be added
later.
Fixes#3924.
TrueHD is a fucked up audio codec with extremely small frame sizes. Some
of these frames start with full headers, which are usually marked as
keyframes, and from which decoding can be started (or at least that's
what you'd expect).
So for such tracks we should probably trust the keyframe flags. Doesn't
really improve seek behavior, though.
Some files have audio tracks with packets that do not have a keyframe
flag set at all. I don't think there's any audio codec which actually
needs keyframe flags, so always assume an audio packet is a keyframe
(which, in Matroska terminology, means it can start decoding from that
packet).
The file in question had these set:
| + Multiplexing application: Lavf57.56.100 at 313
| + Writing application: Lavf57.56.100 at 329
Garbage produced by garbage...
There are other such files produced by mkvmerge, though. It's not
perfectly sure whether these have been produced by FFmpeg as well
(mkvmerge often trusts the information in the source file, even if it's
wrong - so other samples could have been remuxed from FFmpeg).
Fixes#3920.
Don't force CLI usage. It can be imported, and generate_C_header() and
generate_C_definitions() can be called with a file argument instead of
writing to stdout always.
This reverts commit fae7307931.
Before the waf build system was used, we had a configure script written
in shell. To drop the build dependency on Python, someone rewrote the
Python scripts we had to Perl. Now the shell configure script is gone,
and it makes no sense to have a build dependency on both Perl and
Python.
This isn't just a straight revert. It adds the new Matroska EBML
elements to the old Python scripts, adjusts the waf build system, and of
course doesn't add anything back needed by the old build system.
It would be better if this used matroska.py/file2string.py directly by
importing them as modules, instead of calling them via "python". But for
now this is simpler.
this replaces the old fullscreen with the native
macOS fullscreen. additional the
--fs-black-out-screens was removed since the new
API doesn't support it in a way the old one did.
it can possibly be re-added if done manually.
Fixes#2857#3272#1352#2062#3864
The reST contents directive is added to mpv.rst.
In wscript_build.py, the --strip-elements-with-class=contents option is
needed for the rst2man call in order to prevent the TOC from appearing
in mpv.1.
When dropping a file on mpv, either on the window
or the App bundle icon, while holding the shift
key the dropped files will be appended to the
playlist.
Fixes#2166
Allow minimizing the borderless/fullscreen window by clicking on the
taskbar button or pressing Win+Down hotkey.
Also fixes#2229 and probably fixes#2451
Remove 'Quit mpv & remember playback position'
from the menu because it conflicts with the global
logout shortcut. add separator between 'Hide' and
'Quit' for consistency with other Apps.
also rename the 'Movie' menu to 'Video'. it's a
bit more generic.
Fixes#3865
According to MSDN, GetWindowLong and SetWindowLong have been
superseded by GetWindowLongPtr and SetWindowLongPtr.
It's a cosmetic code change in this case.
Was started by Uoti Urpala in commit 5f631d1c. Although it was made part
of demux_mkv.c, it's quite obvious that it's not based on any
pre-existing demux_mkv.c code (or ebml.c/.h for that matter). Anyone
else who has touched this code every since has already agreed to LGPL
relicensing.
Currently, if init_filter fails after lavf_ctx is allocated, uninit is called
which frees lavf_ctx, but doesn't clear the pointer in spdif_ctx. So, on the
next call of decode_packet, it thinks it is already initialized and uses it,
resulting in a crash on my system.
As documented in interface-changes.rst. This makes it much easier to
follow what the heck is going on.
Whether this is adequate for real-world use is unknown.
Enca is dead. libguess is relatively useless due to not having an
universal detection mode. On the other hand, libuchardet is actively
developed.
Manpages changes in the following commit.
This allows leaving autoload in auto-loaded scripts and to be used
in a special profile like "pseudo-gui" without being troublesome
to disable the behavior in profiles that get applied after
pseudo-gui.
Ex:
[someprofile]
script-opts=autoload-disabled=yes
Length property is deprecated and no longer works. This fixes
a bug when the total file duration wasn't visible if the
option to display milliseconds was activated.
Having empty space before the title in layout=*bar looks worse
than the floating buttons in layout=box.
Also disable both playlist buttons selectively according to the
current position.
Needs explicit logic. Fixes a pretty bad regression which prefers
vdpau-copy over native vaapi with direct rendering (with --hwdec=auto)
if libvdpau-va-gl1 is present. The reason is that vdpau-copy is above
vaapi, simply because all vdpau hwdecs are grouped and happened to be
listed before vaapi.
Although this is not that bad for copy-mode (unlike the case described
above), it's still a good idea to use our native vaapi code instead.
Apparently we don't always set the viewport to window dimensions
anymore, e.g. if nothing is actually rendered. This means the viewport
can contain old values.
The window screenshot code uses the viewport values to guess the default
framebuffer dimensions. With --force-window --idle --no-osc (which draws
nothing and issues a glClear() command only), taking a screenshot would
yield an image with the wrong size and possibly garbage in it. Fix this
by explicitly passing the currently known window dimensions. Abusing the
values stored in the viewport was questionable anyway.
We log a large number of formats, but we rarely log the result of the
probing. Change this.
The logic in try_format_exclusive() changes slightly, but should be
equivalent. EXIT_ON_ERROR() checks for FAILED(), which should be
exclusive to SUCCEEDED().
The way playback/loading is stopped on the demuxer layer makes it report
an error to the higher levels of the player. But if playback/loading was
explicitly aborted, printing such an error is confusing and misleading.
This was probably just an oversight anyway. Fix it by using the libmpv
API reported error field instead, which handles this better.
Remove more stuff that was needed only for legacy suboptions.
One user-visible change is that parent-options like --tv are now not
visible anymore. They lead to a special error message when used before,
but now they're simply not part of the option list anymore.
Fixes a segfault introduced in libwayland
e8ad23266f36521215dcd7cfcc524e0ef67d66dd, where a poison value has been
introduced to catch this kind of use-after-free bug.
Long planned. Leads to some sanity.
There still are some rather gross things. Especially g_groups is ugly,
and a hack that can hopefully be removed. (There is a plan for it, but
whether it's implemented depends on how much energy is left.)
Until now, this was only implemented for ao_alsa and AOs not using
push.c. ao_alsa.c relied on enabling funny underrun semantics for
avoiding resets on lower levels, while other AOs using push.c didn't do
anything.
Change this and at least make push.c copy silent data to the AO. This
still isn't perfect as keeping track of how much silence was played when
seems complex, so we don't do it. The consequence is that frame-stepping
will essentially randomize the A/V offset (it'll recover immediately
when unpausing, but still ugly). Also, in order to empty the currently
buffered audio on seeks etc., we still call ao_driver->reset and so on,
so the AO driver will still need to handle this specially.
The intent is to make behavior with ALSA less weird (for one we can
remove the code in ao_alsa.c that tries to trigger an initial
underflow). Also might help with #3754.
We want to avoid causing problems if libmpv is used in an application
that links cuda, or if the libav* libraries are linked with cuda,
as might happen if the scale_npp filter is used.
The test ended up failing if cuda.h wasn't present, even if cuda.h
isn't used during the actual build.
This test is attempting to establish if the ffmpeg being built
against has dynlink_cuda support. While it might theoretically be
possible to build against the older normally-linked-cuda version
of ffmpeg, it seems more trouble than it's worth.
For some reason, some types of accesses didn't warn, for example when
using mp.observe_property() in Lua. This was because the deprecation
handling code explicitly checks certain accesses. I'm not quite certain
why it was done this way. Just make it warn always.
This could be backported to the current release, if we cared.
This was a typo in the extensiuon spec and was probably always broken.
Could have led to broken builds when used with ancient ANGLE headers
(or possibly generic EGL headers).
The latest 375.xx nvidia drivers add support for P016 output
surfaces. In combination with an ffmpeg change to return those
surfaces, we can display them.
The bulk of the work is related to knowing which format you're
dealing with at the right time. Once you know, it's straight forward.
As threatened by the API changes document.
We can actually keep the deprecated --playlist-pos and --cache options,
since they are aliases and not used by the corresponding properties.
They are inconsistent, but do no harm. Keep them for now for the sake of
the command line user.
mpv_identify.sh partially stopped working, because it was never updated.
The shell magic can't deal with property names that contain "/", so we
can't replace "samplerate" with "audio-params/samplerate" - just remove
these properties. (How about you use ffprobe?)
As threatened by the API changes document.
This commit also removes or stubs equivalent calls in IPC and Lua
scripting.
The stubs are left to maintain ABI compatibility. The semantics of the
API functions have been close enough to doing nothing that this probably
won't even break existing API users. Probably.
At least with Nvidia drivers, some thread tries to access D3D11 objects
after ANGLE unloads d3d11.dll. Fix this by holding a reference to
d3d11.dll ourselves.
Might fix the crash in #3348. (I wish I knew why though.)
Since the recent release was named 0.22.0 instead of 0.21.1, bump all
mentions of 0.22.0 to 0.23.0. These were planned removals of deprecated
versions, which obviously didn't happen in 0.22.0.
Run "playlist-remove current" while the last playlist entry is being
played stopped playback. Fix this and return to the first entry instead.
Fixes#3808.
If video reaches EOF, subtitle timing will be switched to timing without
video frames. This means it calls osd_set_force_video_pts() and
overrides the PTS of whatever video frame is current (since the video
frame's PTS has nothing to do with the current playback position
anymore).
This was not reset when seeking back into video. Subtitles wouldn't show
up, or if there was a subtitle displayed, it would get stuck with it. In
particular, this could happen even if EOF was only temporary (such as
with --keep-open).
Fix this by clearing the override PTS whenever a video frame is shown.
Fixes#3770.
- win32-console-wrapper.c was inconsistently using the explicit Unicode
versions of some Windows API functions and structures.
- vo.c should use llabs for int64_t, since long is 32-bit on Windows.
- vo_direct3d.c had a potential use of an uninitialized variable if it
took the first goto error_exit.
This fixes the build in mingw-w64/Clang on MSYS2. It also disables the
use of gnu_printf in Clang, which was what was causing most of the
warnings. The Clang-compiled mpv binary appears to work, but there are
no guarantees yet, since until now mpv has only been tested with
mingw-w64/GCC on Windows.
Fixes#3800
This fixes waf setting the wrong LIBDIR for DEST_OS=win32 in
waflib/Tools/c_config.py:get_cc_version()
Any scripts assuming the implib and pkgconfig are in the wrong
place should be changed to move the .dll instead.
We're too lazy to maintain a full changelog, but there are still a bunch
of places which document specific changes. A user really should check
them on each update, even if there are transition periods (including
printing warnings etc.) for most changes. It's a good idea to give them
more exposure by adding them to the README.
The "default" entry (which is and always was mpv/mplayer's default) does
not have a description set in the ALSA API. (While "sysdefault"
strangely has.)
Instead of an empty description, this should show something nice, so
reuse the ao.c code for naming default devices (see previous commit).
It's still a bit ugly that audio-device-list will have a default entry
for "Autoselect device" and "Default (alsa)", but then again we probably
want to allow the user to force ALSA (i.e. prevent fallbacks to other
AOs) just because ALSA is so flaky and makes this a legitimate feature.
This will make it easier for AOs to add explicit default device entries.
(See next commit.)
Hopefully this change doesn't lead accidentally to bogus "Default"
entries to appear, but then it can only happen if the device ID is
empty, which would mean the underlying audio API returned bogus entries.
Use the device name as fallback. This is ugly, but still better than
skipping the description entirely. This can be an issue on ALSA, where
the API can return entries without proper description.
Deactivating this options makes it possible to
circumvent the default OS X behavior of using
points. Windows on HiDPI resolutions won't open
in double the size anymore and videos are display
in their native resolution when windowed.
Fixes#3716
av_reduce(&num, &den, 1, 14112000, 1000000) can return num=0, den=1.
This means a 1/14112000 timebase (as used by the mp3 demuxer) would
become invalid.
The intention of mp_get_codec_timebase() is to always return a valid
timebase. av_reduce() probably does the logically correct thing - so add
a fallback to the safe default timebase.
Also, increase the av_reduce() parameter to INT_MAX. Let's just pray
this doesn't cause any actual problems. libavformat does the same, but
might be in a different position due to using av_rescale() etc., while
we convert between fractional timestamps and floats.
This is a bit unintuitiv, but it appears hwdec backends have to unset
hwdec_priv manually in their uninit function. Normally with this idiom
you'd expect the common code to do this (and maybe even freeing the priv
struct). Since other hwdec backends do this quite consistently, just fix
vdpau for now.
Also add an assert to detect similar bugs sooner.
Fixes#3788.
The previous commit means subtitles were reinitialized on every seek
(even within a segment). This commit restores the old behavior.
To check whether the segment changed at all, we don't reset the current
start/end values. This assumes the decoder wrapper is always fed by a
stream which doesn't mix segment and non-segment packets, which is
currently always true.
This is needed to put the decoders into the correct state. In
particular, decoders will not initialize the current segment without
this flag. The intention of not setting the flag for seeks within the
segments were to avoid costly decoder reinits, but it seems this is
better handled explicitly in the decoder wrappers.
Implementation-wise, the values from the demuxer/codec header are merged
with the values from the decoder such that the former are used only
where the latter are unknown (0/auto).
Matroska actually has lots of colorimetry metadata that video tracks can
use, including mastering metadata (HDR signal peak) etc.
This commit adds the EBML definitions and parses the most basic fields.
Note that nothing uses these fields yet, this commit is just adding the
necessary parsing and infrastructure.
This happens when ALSA gives us more channels than we asked for, for
whatever reasons. It looks like this wasn't handled correctly. The mpv
and ALSA channel counts could mismatch, which would lead to UB.
I couldn't actually trigger this case, though. I'm fairly sure that
drivers or plugins exist that do it anyway. (Inofficial ALSA motto: if
it can be broken, then why not break it?)
If the input is already mono or stereo, or if channel map selection
results in mono or stereo, then disable further use of the champ ALSA
API (or rather, stop trusting its results). Then we behave like a simple
application that only wants to output mono or stereo.
See #3045 and #2905. I couldn't actually test these cases, but this
commit is supposed to fix them.
set_chmap() skipped _setting_ the ALSA chmap if chmap use was requested
to be disabled by setting dev_chmap.num=0 by the caller, but it still
queried the current ALSA channel map. We don't trust it that much, so
disable that as well.
But we still query and log it, because that could be helpful for
debugging. Otherwise we could skip the entire set_chmap() call in these
cases.
The intention was that if --blend-subtitles is enabled, the frame should
always be re-rendered instead of using e.g. a cached scaled frame. The
reason is that subtitles can change anyway, e.g. if you pause and change
subtitle size and such.
On the other hand, if the frame is marked as repeated, it should always
use the cached copy. Actually "simplify" this and drop the cache only if
playback is paused (which frame->still indicates indirectly).
Also see PR #3773.
unmap_current_image() is called after rendering. This essentially
invalidates the textures, so we can't assume that the image is still
present.
Also see PR #3773.
Enumerate all of the scaling-related options, even for the ``--cscale``
/ ``--tscale`` etc. variants. Unfortunately this breaks 80col quite
severely, but there's nothing I can do about it due to a bug in rst2man
preventing definition list labels from spanning multiple lines.
Also reorder some of the scaling-related options to be closer together
and in a more consistent order (for a top-to-bottom reading flow).
This allows us to define the tukey window (and other tapered windows).
Also add a missing option definition for `wblur` while we're at it, to
make testing out window-related stuff easier.
This deals with the estimation of buffered packets, which is used mostly
for display, but also things like pausing on low buffer levels.
If a stream is fully EOF (no more packets), we don't want to include it
in the total buffer amount. This also means we should make ds->eof less
flaky and more stable, so don't reset it in ds_get_packets() (this
function reset ds->eof just to retrigger a packet read attempt - we can
have this slightly simpler). This somewhat fixes buffering display when
e.g. issuing a refresh seek after re-enabling audio/video when playing
with subtitles only.
This makes no sense, as the flag is supposed to be used for vsync
purposes only (when literally outputting the screen again with no
changes at all), and redrawing is often used for OSD updates.
It's not that easy to decide whether a frame needs to be
reuploaded/rerendered. Using unique frame IDs for input makes it
slightly easier and more robust. This also removes the use of video PTS
in the interpolation path.
This should also avoid reuploading the video frame if it's just redrawn
in paused mode, or when using OSD/subtitles in cover art mode.
Coverart mode has the same issue as no-video mode, except that the video
chain is fully active. It shows only 1 frame at the start, which would
normally mean that only the subtitle at timestamp 0 is shown. Use the
no-video subtitle rendering mode in this case instead.
(This still doesn't handle subtitle display when playing cover-art
without audio, or playing a single image. This is because there's
nothing that will advance playback_pts.)
When pthread_cond_timedwait(), the condition we are checking for could
be true or false. This code assumed it was always false.
This should be an extremely obscure race condition, since it can happen
only if timeout and the condition changing sort of happen at the same
time, or the lock is held for a longer time (which it normally isn't).
But I could observe it a few times.
This code would just keep it busy while e.g. being paused. Even if it's
not paused, it couldn't help with anything since we obviously still lock
display to the externally updated PTS.
Even if a frame is dropped due to the libmpv API user not drawing a
frame, it should be set as current frame. This avoids dropping a frame
forever in certain circumstances such as cover art of the API user was
stuck at initialization or such.
This way people can still use the mouse to quickly check the elapsed time
without moving it all the way to the bottom while still having half the screen
to ignore mouse movement.
If a VO is created, but no video is playing (i.e. --force-window is
used), then until now no subtitles were shown. This is because VO
subtitle display normally depends on video frame timing. If there are no
video frames, there can be no subtitles.
Change this and add some code to handle this situation specifically. Set
a subtitle PTS manually and request VO redrawing manually, which gets
the subtitles rendered somehow.
This is kind of shaky. The subtitles are essentially sampled at
arbitrary times (such as when new audio data is decoded and pushed to
the AO, or on user interaction). To make a it slightly more consistent,
force a completely arbitrary minimum FPS of 10.
Other solutions (such as creating fake video) would be more intrusive or
would require VO-level API changes.
Fixes#3684.
* Fixes: when on the end of playlist only half of entries are displayed.
* Simplifies the logic of limited_list so it's easy to follow.
* limited_list's pos parameter is now 1 based which seem more natural.
* Few changes to comply with code style thorough the file.
* Small format change:
"Playlist: (%d/%d):" -> "Playlist [%d/%d]:"
"Chapters: (%d/%d):" -> "Chapters [%d/%d]:"
These can be used in input.conf for pretty formatting of lists as
with shift+clicking the OSC buttons.
Ex:
z script-message osc-playlist
Z script-message osc-chapterlist
x script-message osc-tracklist
Commit f72a900892 (and others) added support for ordered editions that
recursively refer to other ordered editions. However, this recursion
code incorrectly activated if the source files had ordered chapters
even if the main file only wanted to use them as raw video, resulting
in broken timeline info overall.
Ordered chapters can specify a ChapterSegmentEditionUID value if they
want to use a specific edition from a source file. Otherwise the
source is supposed to be used as a raw video file. The code checked
demuxer->matroska_data.num_ordered_chapters for an opened source file
to see whether it was using a recursive ordered edition, but demux_mkv
could enable a default ordered edition for the file using the normal
playback rules even if the main file had not specified any
ChapterSegmentEditionUID. Thus this incorrectly enabled recursion if a
source file had a default edition using ordered chapters. Check
demuxer->matroska_data.uid.edition instead, and ensure it's never set
if a file is opened without ChapterSegmentEditionUID.
Also fix what seems like a memory leak in demux_mkv.c.
Signed-off-by: wm4 <wm4@nowhere>
Just that actually changing it at runtime won't do anything.
This deals with a nasty initialization order issue with encoding.
Encoding is initialized after options have initialized, but before
--load-scripts is checked and executed. Encoding initialization accesses
FFmpeg API, thus it has to run after FFmpeg is initialized (which also
implies it's initialized after options/logging init). On the other hand,
it sets the encoding builtin profile, which possibly sets --load-scripts
to "no". That failed at this point because --load-scripts was marked as
fixed.
Just marking it as not fixed gets rid of the headache, even if it's not
perfectly orthodox.
--reset-on-next-file=all triggers m_config_backup_all_opts(), which
backups all options (even deprecated ones). Later, when the option
values are reset with m_config_restore_backups(), mp_on_set_option() is
called, which in turn calls mp_property_do_silent(), which in turn will
call mp_property_generic_option() and m_config_get_co(), which triggers
the deprecation message.
Unfortunately there's no good way to determine whether an option has
actually changed (there's no option value compare operation), so the
deprecated options have to be set no matter what. On the other hand, we
can't pass through additional flags through the property layer. So we
add a dumb global flag to silence the deprecation warnings in these
cases.
Fortunately m_config_get_co_raw() works to silence the warnings.
m_config_get_co() also resolves aliases (deprecated and non-deprecated),
but aliased options are handled differently by the option-property
bridge, so we don't need to do that here, so the only purpose of it is
to trigger a warning for deprecated (non-alias) options.
The tv-freq options and properties use different types, thus must be
treated as incompatible. Fixes an assertion with reset-on-next-file=all,
which tries to set the option.
Fixes#3708.
It turns out the glFlush() call really helps in some cases, though only
in audio timing mode (where we render, then wait for a while, then
display the frame). Add a --opengl-early-flush=auto mode, which does
exactly that.
It's unclear whether this is fine on OSX (strange things going on
there), but it should be.
See #3670.
Don't leave the buffering state while the demuxer is still marked as
having underflowed. It's unclear why this hasn't been done before - with
the logic being complicated as it is, maybe there was a reason for this.
This is actually still not very reliable, but should be better than what
was before: on stream switching decoders can read packets all while the
demuxer is executing a refresh seek, which creates the underrun
situation - but nothing really totally guarantees that the underrun
state remains stable when the demuxer is back at the current demuxer
position. Anyway, it's an improvement.
The rest of the touched condition is not changed, just moved around for
cosmetic reasons.
in very rare circumstances, doing a relative seek like +1s will end up
doing an absolute seek to 00:01. this guards against that possibility.
so far i've only ever seen this issue when using --ad=lavc:ac3_at and
doing several relative seeks in quick succession. this is likely either
a bug in the audiotoolbox decoder in ffmpeg, or simply due to inherent
latency in that hardware decoder which causes brief periods of time
where the current audio pts is unknown.
i've seen several mpegts samples where pts jumps backwards and repeats
itself. this usually happens on live tv streams from cable providers,
particularly when the stream switches from one advertisement to another.
when there is a huge delay between audio/video sync, it can take a
really long time to converge back. this speeds up the resync time by
increasing the max_change allowed per iteration.
Signed-off-by: wm4 <wm4@nowhere>
Thread-local storage in GCC is platform-specific, and some platforms that
are otherwise perfectly capable of running mpv may lack TLS support in GCC.
This change adds a test for GCC variant of TLS and relies on its result
instead of assumption.
Provided that LLVM's `__thread` support is similar to GCC, the test is
called "GCC/LLVM TLS".
Signed-off-by: wm4 <wm4@nowhere>
At this point, all other hwaccels provide -copy modes, and vdpau is the
exception with not having one. Although there is vf_vdpaurb, it's less
convenient in certain situations, and exposes some issues with the
filter chain code as well.
FFmpeg recently got "support" for mov edit lists. This is a terrible
hack that will fail completely at least with some decoders (in
particular wrappers for hardware decoding might be affected). As such it
makes no point to pretend they are supported, even if we assume that the
"intended" functionality works, that there are no implementation bugs
(good luck with all that messy code added to the already huge mov
demuxer), and that it covers enough of the mov edit list feature to be
of value.
So log an error if the FFmpeg code for mov edit lists appears to be
active - AV_PKT_FLAG_DISCARD is used only for "clipping" edit list
segments on non-key frame boundaries.
In the first place, FFmpeg committed this only because Google wanted it
in, and patch review did not even pick up obvious issues. (Just look how
there was no lavc version bump when AV_PKT_FLAG_DISCARD was added.)
We still pass the new packet flag to the decoders (av_common.c change),
which means we "support" FFmpeg's edit list code now. (Until it breaks
due to FFmpeg not caring about all the details.)
Had only tested with luajit which supports the \xHH syntax added
in Lua 5.2.
The arrow is troublesome to use since the ideal way to use it, as
the OSD code uses it, needs \alpha&H00<arrow>\r to work, which
does not in OSC's way of showing messages.
Both AVFrame.pts and AVFrame.pkt_pts have existed for a long time. Until
now, decoders always returned the pts via the pkt_pts field, while the
pts field was used for encoding and libavfilter only. Recently, pkt_pts
was deprecated, and pts was switched to always carry the pts.
This means we have to be careful not to accidentally use the wrong
field, depending on the libavcodec version. We have to explicitly check
the version numbers. Of course the version numbers are completely
idiotic, because idiotically the pkg-config and library names are the
same for FFmpeg and Libav, so we have to deal with this explicitly as
well.
Change a few other defaults accordingly:
- seekbarstyle=bar looks better with bottombar.
- Bigger scalewindowed and scalefullscreen make bottom/topbar more readable.
POSIX leaves poll() behavior on directories unspecified. While on
Linux, it seems to behave the same way as regular files (always
return immediately), this is not guaranteed. At least with OSX
10.12, it seems to wait, which essentially means that opening
directories will "hang".
Fixes#3530 and #3649.
Keyboard input in the console still isn't quite as flexible as it is in
the video window. Ctrl+<letter> and Ctrl+LEFT/RIGHT work, but
Ctrl+Alt+<letter> and Ctrl+<number> do not. Also, in the new Windows 10
console, a bunch of Ctrl keystrokes including Ctrl+UP/DOWN are handled
by the console window and not passed to the application.
Unlike in w32_common.c, we can't really translate keyboaard input
ourselves because the keyboard layout of the console window (in
conhost.exe) doesn't necessarily match the keyboard layout of mpv's
console input thread, however, using ToUnicode as a fallback when the
console doesn't return a unicode value could be a possible future
improvement.
Fixes#3625
The original version of this code in getch2-win.c fetched 128 console
events at once. This was probably to maximize the chance of getting a
key event if there were other events in the buffer, because it returned
the value of the first key event it found and ignored all others. Since
that code was written, it has been modified to receive console input in
an event-based way using an input thread, so it is probably not
necessary to fetch so many events at once any more. Also, I'm not sure
what it would have done if there were more than 128 events in the
console input buffer. It's possible that fetching multiple events at a
time also had performance advantages, but I can't find any other
programs that do this. Even libuv just fetches one console event at a
time.
Change read_input() to fetch only one event at a time and to consume all
available events before returning to WaitForMultipleObjects. Also remove
some outdated comments and pass the console handle through to the input
thread instead of calling GetStdHandle multiple times (I think this is
theoretically more correct because it is possible for the handles
returned by GetStdHandle to be changed by other threads.)
If the "default" device refuses to be opened as spdif device (i.e. it
errors due to the AES0 etc. parameters), we were falling back to the
iec958 device. This is needed on some systems for smooth operation with
PCM vs. spdif.
Now change it to try "hdmi" before "iec958", which supposedly helps in
other situations.
Better suggestions welcome. Apparently kodi does this too, although I
didn't check directly.
- Change connector selection to accept human readable names (such as
eDP-1, HDMI-A-2) rather than arbitrary numbers.
- Change GPU selection to accept GPU number rather than device paths.
- Merge connector and GPU selection into one --drm-connector.
- Add support for --drm-connector=help.
- Add support for --drm-* in EGL backend.
- Refactor KMS; reduce state sharing across drm_common.
The glFlush() call was made optional recently
since it's not needed in most cases. On OSX though
this is needed since we removed kCGLPFADoubleBuffer
from the context creation, so the glFlush() call
was added to the cocoa backend only.
The CGLFlushDrawable() call can be safely removed
since it only does something when a double
buffered context is used. Also fixes a small typo.
Fixes#3627.
In "dumb mode" (where most features are disabled and which only performs
some basic rendering) we explicitly copy a set of whitelisted options,
and leave all the other options at their default values. Add the new
--opengl-early-flush option to this whitelist. Also remove an option
field accidentally added in the commit adding --opengl-early-flush.
Don't require special property code for handling updates, and simply use
the UPDATE_AUDIO flag instead. Also make runtime changes to
--audio-client-name take effect.
Now a reload requested by an AO behaves in exactly the same way as
changing an AO-related options (like --audio-channels or
--audio-exclusive). This is good for testing and uniform behavior. (You
could go as far as saying it's a necessity, because the spotty and
obscure AO reload behavior is hard to reproduce and thus hard to test at
all.)
This affects changing audio configuration options. Explicitly flush and
uninitialize the audio output first before doing the rest. This should
ensure all state attached to the audio output is discarded and not used
during the reconfiguration.
Also add a comment to the reinit_audio_filters() call. It doesn't
necessarily restore the audio chain fully, but makes sure a seek is
issued if the amnount of buffered audio discarded was huge enough to
cause "problems".
It seems this can cause issues with certain platforms, so better to
disable it by default. The original reason for this isn't overly
justified, and display-sync mode should get rid of the need for it
anyway.
The new option is meant for testing, and will probably be removed if
nobody comes up and reports that enabling the option actually improves
anything.
The player tries to avoid splitting frames with spdif (sample alignment
stuff). This can in certain corner cases with certain drivers lead to
the situation that ao_get_space() returns a number higher than 0 and
lower than the audio frame size. The playloop will round this down to 0
bytes and do nothing, leading to a missed wakeup. This can lead to
underruns or playback completely getting stuck.
It can be reproduced by playing AC3 passthrough with no video and:
--ao=null --ao-null-buffer=0.256 --ao-null-outburst=6100
This commit attempts to fix it by allowing the playloop to write some
additional data (to get a complete frame), that will be buffered within
the AO ringbuffer even if the audio device doesn't want it.
If we really want client API users to use mpv_set_property() instead of
mpv_set_option(), then compatibility handling of deprecated options
should be included. Otherwise, there's the danger that client API users
either break too early (and without a warning), or mpv_set_option() will
continue to have a reason to exist.
Reduces code duplication between OpenGL backend and DRM VO.
(The control() for OpenGL backend isn't sufficiently similar to the
VO's control() to consider merging it as a whole - I extracted only the
FPS code.)
Rename the text subtitle options from --sub-text- to --sub-
and --ass- options to --sub-ass-.
The intention is to common sub options to prefixed --sub-
and special ASS option be seen as a special version of sub options.
The OSD options that work like the --sub- options are still named
--osd-.
Man page updated including a short note about renamed --sub-text-*
and --ass-* options to --sub-* and --sub-ass-*.
This is potentially incompatible if a program used negative timestamps
to deal with timestamp resets, which would potentially lead to mpv
producing and using negative timestamps.
"seek -10 absolute" will seek to 10 seconds before the end. This more or less
matches the --start option and negative seeks were otherwise useless (they just
clipped to 0).
Regression since commit bbcd0b6a. This code is just cursed, because it's
a fragile state machine with no proper tests, and which could be done in
a much simpler way. Without doubt this change will cause a regression in
some ridiculous corner case as well.
Fixes#3610 (the cause of it, not the behavior it resulted in).
When the vaapi decoder is used in copy mode, it creates a dummy
display to render to. In theory, this should support hardware
decoding on on a separate GPU that is not actually connected to
any output (like an iGPU which supports more formats than the
external GPU to which the monitor is connected).
However, before this change, only X11 displays were supported as
dummy displays. This caused some graphics drivers (namely
intel-driver) to core dump when they were not actually used as X11
module.
This change introduces support for drm libav displays, which
allows vaapi-copy to run on such cards which are not actually
rendering the X11 output.
Move the screensaver enable/disable determination to a central place,
and call it if the stop-screensaver property is changed.
Also, do not stop the screensaver when in idle mode (i.e. no file is
loaded).
Fixes#3615.
Other than being overly convoluted, this seems to make sense to me.
Except that to get the "rot" transform I have to set flip=true, which
makes no sense at all to me.
Combining rotation and cropping didn't work. It was just completely
broken.
I'm still not sure if this is correct. Chroma positioning seems to be
broken on rotation. There might also be a problem with non-mod-2 frame
sizes. Still, strictly an improvement for both rotated and non-rotated
rendering modes.
Also, this could probably be written in a more elegant way.
Commit aa1047a3 originally added this as:
+ // this is from the DarkPlaces engine, reduces to 3x3. Original code
+ // released under GPL2 or any later version.
According to Rudolf Polzer, the original author (a certain LH) was
actually asked whether it would be ok to put this code under LGPL, and
the author gave his agreement. This code is not from id Software either
(on which large parts of DarkPlaces is based on), which is the main
reason why DarkPlaces is under GPL.
So this note is just confusing, and always has been LGPL. Fix it.
We'd like to get log messages on the output as soon as possible in the
output. I also feel like using fflush() is nicer than using setvbuf().
Who knows how the latter behaves on win32.
The video code can deal fine with feeding software image formats to
hwdec interop drivers. In RPI's case, this is preferable for
performance, working around OpenGL bugs (see RPI firmware issue #666),
and because OpenGL rendering doesn't bring too many advantages due to
RPI supporting GLES 2.0 only.
Maybe a way to force the normal video path is needed later. But
currently, this can be tested by just not loading the hwdec interop
driver.
If you run command-line mpv and set --hwdec to something that does
not load the RPI interop layer, you'll even have to use --hwdec-preload
manually to get it enabled.
Was intended to put the GL layer above the standard console. (But
actually that was done already, and the oddness I'm seeing seems to
be an unrelated bug.)
The code for expanding the ~~ prefix used mp_find_config_file(), which
strictly looks for _existing_ files in any config path (e.g. not just
the user-local one, but also system-wide config). If no such file
exists, it simply returns NULL, which makes the code below just return
the literal, unexpanded path.
Change this so that it'll resolve the path to the user-local config
directory instead.
Requested in #3591.
This should make display-names usable on Windows. It returns a list of
GDI monitor names like "\\.\DISPLAY1". Since it may be useful to get the
monitor that Windows considers associated with the window (with
MonitorFromWindow,) this will always be returned as the first argument.
This monitor is the one used for display-fps and icc-profile-auto.
Seems like a valid use-case. Not sure if I like it calling back into the
config code. Care has to be taken for not letting the config path
resolving code dead-lock (which is why locking details in the msg.c code
are changed).
Fixes#3591.
I'm not sure if this option affects anything or if it's a placebo,
especially since the VO thread is now registered with MMCSS. Still, I
think --priority=high may have helped back when I used mplayer2 on a
netbook. It's also possible that encoding-mode users would want to set
--priority=idle.
Anyway, it was one of the last M_OPT_FIXED options, so fix that.
We always want to use __declspec(selectany) to declare GUIDs, but
manually including <initguid.h> in every file that used GUIDs was
error-prone. Since all <initguid.h> does is define INITGUID and include
<guiddef.h>, we can remove all references to <initguid.h> and just
compile with -DINITGUID to get the same effect.
Also, this partially reverts 622bcb0 by re-adding libuuid.a to the
build, since apparently some GUIDs (such as GUID_NULL) are not declared
in the source file, even when INITGUID is set.
AVIOContext.seekable is actually a bitfield. Currently, it has only
AVIO_SEEKABLE_NORMAL defined, but it might be extended with a hint for
non-byte seekability. Thus we should check it correctly.
This should still allow user-set default options to override built-in
pseudo-gui while respecting user-set pseudo-gui options.
Pros:
- user option in default profile overrides built-in pseudo-gui's options
Ex: screenshot-directory overrides built-in pseudo-gui's
- user can "fix" pseudo-gui if some option like "force-window=no" is set
in default by setting "force-window=yes" in [pseudo-gui]
- `mpv --profile=pseudo-gui` will work as before
Cons:
- --show-profile=pseudo-gui won't display the built-in's options
Original idea from wm4.
Documentation edits mostly by wm4.
Signed-off-by: wm4 <wm4@nowhere>
If the PTS goes backwards (whether it's a timestamp reset or some other
problem) would just use 0 as frame duration. (At least until the logic
for detecting divergence with the timestamps gets active.)
Trust the demuxer framerate in these cases instead, if it's available. I
think this improves behavior slightly with some broken files.
During init it will first call mp_load_builtin_scripts(), and then again
via mp_load_scripts().
This was harmless (a second attempt won't load it again if the first one
was successful), but it's unnecessary, and also looks confusing if the
scripts failed to load the first time.
demux_lavf.c forces seek to being determined as supported if
STREAM_CTRL_HAS_AVSEEK is returned as success. But it always succeeds
with current FFmpeg versions. (Seems like Libav commit cae448cf broke
this in early 2016.)
Now we can't determine via private API whether the underlying protocol
supports read_seek anymore. The affected protocols (mostly rtmp) also
set seekable=0, meaning they signal they're not seekable, even though
read_seek would work. (My guess is that this can't be fixed because even
though seekable is in theory a combination of elaborate flags [of which
only 1 is defined, AVIO_SEEKABLE_NORMAL], a seekable!=0 always means
it's byte-seekable in some way.)
So the FFmpeg API is being garbage _again_, and all what we can do is
determining this via protocol name and a whitelist.
Should fix the behavior reported in #1701.
Secondary subtitle streams (to be shown on the top of the screen along
main subtitle stream) were shown with normal alignment. This is because
we tell libass to override the alignment style (a relatively recent
change, see commit 2f1eb49e). This would behave differently with old
libass versions too.
To escape the mess, just set the alignment explicitly with an override
tag instead of modifying the style.
This should normally happen only if memory allocation for the state
happens, which should be extremely rare. But with Luajit on OSX, it can
happen if the magic compiler flags required by Luajit were not passed to
mpv compilation. Print an error to reduce confusion.
The intention of M_OPT_FIXED is to make options not runtime-changeable,
so trying to set them at runtime will always error. This is not wanted
for --profile and --include, for which there is no reason to block them
at runtime.
Fixes#3581.
When switching a subtitle track, the subtitle wasn't necessarily
updated, especially when playback was paused.
Some awfully subtle and complex interactions here.
First off (and not so subtle), the subtitle decoder will read packets
only on explicit update_subtitles() calls, which, if video is active, is
called only when a new video frame is shown. (A simply video frame
redraw doesn't trigger this.) So call it explicitly. But only if
playback is "initialized", i.e. not when it does initial track selection
and decoder init, during which no packets should be read.
The second issue is that the demuxer thread simply will not read new
packets just because a track was switched, especially if playback is
paused. That's fine, but if a refresh seek is to be done, it really
should do this. So if there's either 1. a refresh seek requested, or 2.
a refresh seek ongoing, then read more packets.
Note that it's entirely possible that we overflow the packet queue with
this in unpredicated weird corner cases, but the queue limit will still
be enforced, so this shouldn't make the situation worse.
This was dumb and could return something like "{name=123}" as an array.
Also, fix the error message if a key is not a string. lua_typename()
takes a type directly, not a stack item.
The last commit was fine - just making some enhancements.
Rename the function to parse_node_chapters(), since it really has not
much to do with Lua.
Don't use len<0 as check whether it makes sense to set chapters, and
instead check for mpctx->demuxer (that includes the possibility to set
chapters e.g. within a preload hook, but after chapters are initialized
from the demuxer).
Return M_PROPERTY_ERROR instead of M_PROPERTY_OK if the mpv_node has the
wrong type.
It's ok if a chapter has no title, so change the checks accordingly.
Remove a Yoda condition.
Notify that chapter metadata might have changed with mp_notify() (the
chapter list itself is already taken care by generic code).
Fix leaking the metadata allocations of the new chapter list.
Obviously, in the vast majority of cases, there's only one device
in the system, but doing this means we're more likely to get a
usable device in the multi-device case.
cuda would support decoding on one device and displaying on another
but the peer memory handling is not transparent and I have no way
to test it so I can't really write it.
The documentation around this stuff is poor, but I found an nvidia
sample that demonstrates how to use the interop API most efficiently.
(https://github.com/nvpro-samples/gl_cuda_interop_pingpong_st)
Key lessons are:
1) you can register the texture itself and have cuda write to it,
thereby skipping an additional copy through the PBO.
2) You don't have to be mapped when you do the copy - once you get a
mapped pointer, it remains valid. Magic!
This lets us throw out the PBOs as well as much of the explicit
alignment and stride handling.
CPU usage is slightly (~3%) lower for 4K content in one test case,
so it makes a detectable difference, and presumably saves memory.
Seems like this confused users quite often.
Instead of --profile=pseudo-gui, --player-operation-mode=pseudo-gui now
has to be used to invoke pseudo GUI mode. The old way still works, and
still behaves in the old way.
I would have been fine with this, but now I want to add another flag,
and the duplication would become more messy than having a strange
function for deduplication.
The property observation mechanism turns properties into integer IDs for
fast comparison. This means if two properties get the same ID, they will
receive the same notifications. Use this to make properties under
options/ receive notifications. The option-property bridge marks
top-level properties with the same name as the options.
This still might not work in cases the C code sets values on options
structs directly.
On x11, you can change the fullscreen via the window manager and without
mpv's involvement. In these cases, the internal fullscreen flag has to
be updated.
The hack used for this didn't really work properly. Change it
accordingly. The important thing is that the shadow copy of the option
is updated. This is still not really ideal.
Fixes#3570.
The documentation claims that --video-unscaled will still perform
anamorphic adjustments, and it rightfully should. The current reality is
that it does not, because the video-unscaled size was based on the wrong
set of variables. (encoded width/height instead of nominal display
width/height)
If one of the audio output format options is set, brually reinit the
audio chain.
Since the audio-channels property is still mapped to the deprecated
read-only property, "options/audio-channels" currently has to be used,
e.g. "cycle-values options/audio-channels 2 7.1".
The intention is to give libmpv users as much flexibility to load
scripts as using mpv from CLI, but without restricting libmpv users from
having to decide everything on creation time, or having to go through
hacks like recreating the libmpv context to update state.
Basically, make use of all the new code.
Unfortunately, the new code and it's "compromise semantics" with the old
behavior before the options/properties merge bites back: this
vid/aid/sid, deinterlace, and video-aspect properties have neutral
default settings (such as "auto"), but return the current value during
playback, so we would unnecessarily force this value when playback is
resumed. So leave those as they are.
Also, the colormatrix options were removed a long time ago.
Add this flag where needed. You shouldn't be able to set e.g. config-dir
in these situations.
Remove the mpctx->initialized check from the property/option bridge,
since it's in use strictly only after initialization. Likewise, the
apply-profile command doesn't need to check this.
When we rotate the inmage by 90° or 270°, chroma width and height need
to be swapped.
Fixes#3568.
But is the chroma sub location correct? Who the hell knows...
With the merging of options and properties, the mpv_set_option()
function is close to being useless, and mpv_set_property() can be used
for everything instead. There are certain conflicts remaining, which are
explained in depth in the docs. For now, none of this should affect
existing code using the client API.
Make mpv_set_property() redirect to mpv_set_option() before
initialization.
Remove some options marked as M_OPT_FIXED. The "pause" and "speed"
options cannot be written anymore without the playloop being notified by
it, so the M_OPT_FIXED does nothing. For "vo-mmcss-profile", the problem
was lack of synchronization, which has been added. I'm not sure what the
problem was with "frames" - I think it was only marked as M_OPT_FIXED
because changing it during playback will have no effect. Except for
pause/speed, these changes are needed to make them writable as
properties after mpv_initialize().
Also replace all remaining uses of CONF_GLOBAL with M_OPT_FIXED.
mp_new_client() blatantly accessed some mutex-protected state outside of
the mutex.
The destruction code is in theory OK, but with changes in the following
commits it'll be a bit hard to guarantee that it stays this way. Add a
simple flag that makes adding new clients impossible, so that having no
clients after shutdown_clients() remains guaranteed.
Instead of rechecking the timers every time after an event is read, do
it only once the event queue is empty. This is probably slightly more
efficient, and facilitates the next commit.
This is the actual decoder output, with no overrides applied. (Maybe
video-params shouldn't contain the overrides in the first place, but
damage done.)
This really shouldn't be in vd_lavc.c - move it to dec_video.c, where it
also applies aspect overrides. This makes all overrides in one place.
The previous commit contains some required changes for resetting the
image parameters change detection (i.e. it's not done only on video
aspect override changes).
Use the new mechanism, instead of wrapped properties. As usual, extend
the update handling to some options that were forgotten/neglected
before. Rename video_reset_aspect() to video_reset_params() to make it
more "general" (and we can amazingly include write access to
video-aspect as well in this).
A recent change merged the window-scaler option and property, but forgot
that the option is float for some reason, while the property uses
double. This led to undefined behavior. Fix it by changing the option
to double too.
Setting the osc or ytdl properties will now load/unload the associated
scripts. (For ytdl this does not mean the currently played URL will be
reloaded.)
Also add a changelog entry for this, which also covers the preceding
work for --terminal.
Remove wrapper properties for OSD and video position updates, use the
new mechanism for them. We can mark the options directly. Update
behavior will work for more options (since I've casually marked more
affected options than the old less direct mechanism covered).
My original idea was making mpv_initialize() a no-op, but it seems this
can't happen after all. The problem is especially with subtle
interactions in option parsing (basically all pre-parse options).
Instead, I might go into the opposite direction, and add a new API
function that takes over the role of mpv_create+mpv_initialize, and
which will take a list of options. This list will be for the purpose
of setting options that can be set only at initialization time (such
as config-dir). This would also make it more uniform with the command-
line player initialization.
Maybe.
In any case, for now revert parts of commit 453fea87 to remove the
initialization-related freedoms it added. Fortunately, this wasn't
released yet, so we remove it from the API as if it never happened.
(The rest of that commit is still fine, just not the additional
freedom.)
Move the MPV_LEAK_REPORT env query to mp_create(), where it will also be
used by the client API (it might be helpful, so why not). The same
applies to MPV_VERBOSE.
The prepare_playlist() call doesn't need to be in mp_initialize() and
can just be in mp_play_files() to reduce the size of mp_initialize().
Also, remove wakeup_playloop(), which is 100% redundant with
mp_wakeup_core_cb().
So client API users don't have to care about whether to set this before
or after mpv_initialize().
We still don't enable terminal at any point before mpv_initialize(),
because reasons.
This also subtly changes some behavior how terminal options are applied
while parsing. This essentially reverts the behavior as it was reported
in issue #2588. Originally, I was hoping to get rid of the pre-parse
option pass, but it seems this is absolutely not possible due to the way
config and command line parsing are entangled. Command line options take
priority over configfile options, so they have to be applied later - but
we also want to apply logging and terminal options as specified on the
command-line, but _before_ parsing the config files. It has to be this
way to see config file error messages on the terminal, or to hide them
if --no-terminal is used. libmpv considerations also factor into this.
Until now, the terminal thread always sent a quit command if the
terminal thread was torn down (whether it happened via terminal_uninit()
or a quit signal). This is not so good if we want to enable toggling
terminal use at runtime, since disabling the terminal would always make
the player quit. So we want terminal_uninit() not to send quit.
This can be easily fixed by using the "death byte" sent to the pipe used
for thread tear-down to indicate whether it was caused by a signal or
terminal_uninit().
Extend the flag-based notification mechanism that was used via
M_OPT_TERM. Make the vo_opengl update mechanism use this (which, btw.,
also fixes compilation with OpenGL renderers forcibly disabled).
While this adds a 3rd mechanism and just seems to further the chaos, I'd
rather have a very simple mechanism now, than actually furthering the
mess by mixing old and new update mechanisms. In particular, we'll be
able to remove quite some property implementations, and replace them
with much simpler update handling. The new update mechanism can also
more easily refactored once we have a final mechanism that handles
everything in an uniform way.
--quiet can be always set - the playloop checks it whenever rendering
the status line. Nothing special about it.
The ytdl- options are simply refetched by the Lua script every time a
stream is opened, so it makes sense to be able to change them at runtime
as well.
The VO options don't have a real reason to be marked with it anymore.
For audio files, this is identical to time-pos (except read-only).
For audio-video files, this returns the audio position. Unlike
time-pos, this is not quantized to a video frame.
For video-only files, this property is unavailable.
This was in the parser code all along. As far as I can tell, *cp was
intended. There is no need to check cp for NULL (nor does it make any
sense to do so every time around the loop) for AF_CONTROL_COMMAND.
However, s->matrixstr can be NULL, so checking for that separately is in
order.
For stereo and typical L/R-first channel arrangements, this avoids
undesirable phasing artifacts, especially obvious when speed is changed
and then reset. Without this, there is a very audible change in the
stereo field even when librubberband is no longer actually making any
speed changes.
Oops, this is kind of important, isn't it?
Stopped working for properties which don't implement
M_PROPERTY_GET_CONSTRICTED_TYPE directly, as do_action() goes to the
property directly, while m_property_do() does a fallback.
Some properties had a different type from their equivalent options (such
as mute, volume, deinterlace, edition). This wasn't really sane, as raw
option values should be always within their bounds. On the other hand,
these properties use a different type to reflect runtime limits (such as
range of available editions), or simply to improve the "UI" (you don't
want to cycle throuhg the completely useless "auto" value when cycling
the "mute" property).
Handle this by making them always return the option type, but also
allowing them to provide a "constricted" type, which is used for UI
purposes. All M_PROPERTY_GET_CONSTRICTED_TYPE changes are related to
this.
One consequence is that you can set the volume property to arbitrary
high values just like with the --volume option, but using the "add"
command it still restricts it to the --volume-max range.
Also deprecate --chapter, as it is grossly incompatible to the chapter
property. We pondered renaming it to --chapters, or introducing a more
powerful --range option, but concluded that --start --end is actually
enough.
These changes appear to take care of the last gross property/option
incompatibilities, although there might still be a few lurking.
For some odd reason, value ranges for the window-scale option and
property are different, and the property has a more narrow range. Change
it to the option range.
Also store the window-scale value into the option value when setting the
property, so it will be persistent if the window is closed and reopened.
Conflicts with the "playlist-pos" property. They're really a bit too
different, and since the --playlist-pos option is relatively new and
obscure, just rename it to get this out of the way.
Make the option type exactly the same as the underlying option's one. I
think this has no user-visible consequences, but makes more sense for
the option-property bridge.
These are not mapped as property, so the option-property bridge has to
skip them. Do this automatically if a property is not found. I know that
this affects --quiet and --really-quiet, but in theory there could be
more.
There was both user-agent and user_agent options, the former is deprecated in FFmpeg/FFmpeg@27714b462 master.
Libav uses both forms.
This avoids constant `[ffmpeg] http: the user-agent option is deprecated, please use user_agent option` warnings using ytdl_hook.
All option write accesses are now put through the property interface,
which means runtime option value verification and runtime updates are
applied. This is done even for command line arguments and config files.
This has many subtle and not-so-subtle consequences. The potential for
unintended and intended subtle or not-subtle behavior changes is very
large.
Architecturally, this is us literally jumping through hoops. It really
should work the other way around, with options being able to have
callbacks for value verification and applying runtime updates. But this
would require rewriting the entirety of command.c. This change is more
practical, and if anything will at least allow incremental changes.
Some options are too incompatible for this to work - these are excluded
with an explicit blacklist.
This change fixes many issues caused by the mismatch between properties
and options. For example, this fixes#3281.
There were multiple values under M_OPT_EXIT (M_OPT_EXIT-n for n>=0).
Somehow M_OPT_EXIT-n either meant error code n (with n==0 no error?), or
the number of option valus consumed (0 or 1). The latter is MPlayer
legacy, which left it to the option type parsers to determine whether an
option took a value or not. All of this was changed in mpv, by requiring
the user to use explicit syntax ("--opt=val" instead of "-opt val").
In any case, the n value wasn't even used (anymore), so rip this all
out. Now M_OPT_EXIT-1 doesn't mean anything, and could be used by a new
error code.
This makes m_config_set_option_raw() the function that is always called
on the lowest level (as leaf function for all other functions).
To do this, m_config_parse_option() has to do something special to deal
with "impure" options like --vf-add, which work on the previous option
value, instead of fully replacing it. m_config_set_option_raw() itself
always completely replaced the previous value.
This meant "cannot be used as per-file option" (wrt. playlist items).
Doesn't make too much sense anymore, especially given how obscure
per-file options are.
This _actually_ does what commit 8716c2e8 promised, and gives a slight
performance improvement for client API users which make a lot of
requests (like reading properties).
The main issue was that mp_dispatch_lock() (which client.c uses to get
exclusive access to the core) still called the wakeup callback, which
made mp_dispatch_queue_process() exit. So the playloop got executed
again, and since it does a lot of stuff, performance could be reduced.
If --blend-subtitles=yes is given, vo_opengl will call osd_draw()
multiple times, once for subtitles, and once for OSD. This meant that
the want_redraw flag was reset before the OSD was rendered, which in
turn meant that update_osd() was never called. It seems like removing
the per-OSD object want_redraw wasn't such a good idea. Fix it by
reintroducing such a flag for OSDTYPE_OSD only.
Also, the want_redraw flag is now unused, so kill it.
Another regression caused by commit 9c9cf125. Fixes#3535.
If we were waiting, and then exiting due to timeout, we still have to
recheck the condition protected by the condition variable/mutex in order
to get back to a consistent state. In this case, the queue was locked
with mp_dispatch_lock(), and mp_dispatch_queue_process() got to return
without waiting for unlock.
Also caused commit 8716c2e8. Probably an argument for replacing the
dispatch queue by a simple mutex.
Negative height is used to signal a flipped framebuffer. There's
absolutely no reason to pass this down to overlay_adjust(), and only
requires implementers to deal with an additional special-case.
Instead of using input_ctx for waiting, use the dispatch queue directly.
One big change is that the dispatch queue will just process commands
that come in (e.g. from client API) without returning. This should
reduce unnecessary playloop excutions (which is good since the playloop
got a bit fat from rechecking a lot of conditions every iteration).
Since this doesn't force a new playloop iteration on every access, this
has to be enforced manually in some cases.
Normal input (via terminal or VO window) still wakes up the playloop
every time, though that's not too important. It makes testing this
harder, though. If there are missing wakeup calls, it will be noticed
only when using the client API in some form.
At this point we could probably use a normal lock instead of the
dispatch queue stuff.
They're useless, and I have no idea what they're actually supposed to do
(wrt. pending input processing changes).
Also remove their implicit uses from the IPC handlers.
This does 3 kinds of changes:
- change sleeptime=x to mp_set_timeout()
- change sleeptime=0 to mp_wakeup_core() calls (to be more explicit)
- change commands etc. to call mp_wakeup_core() if they do changes that
require the playloop to be rerun
This is preparation for the following changes. The goal is to process
client API requests without having to rerun the playloop every time. As
of this commit, the changes should not change behavior. In particular,
the playloop is still implicitly woken up on every command.
Currently, calling mp_input_wakeup() will wake up the core thread (also
called the playloop). This seems odd, but currently the core indeed
calls mp_input_wait() when it has nothing more to do. It's done this way
because MPlayer used input_ctx as central "mainloop".
This is probably going to change. Remove direct calls to this function,
and replace it with mp_wakeup_core() calls. ao and vo are changed to use
opaque callbacks and not use input_ctx for this purpose. Other code
already uses opaque callbacks, or has legitimate reasons to use
input_ctx directly (such as sending actual user input).
This could in theory lead to missed updates if subtitles were switched
or external OSD overlays (via overlay-add) were updated. While the
change IDs of each of those were consistent, switching between two
separate OSD sources is not, and we have to explicitly trigger a change.
Regression since commit 9c9cf125. The new code is actually better,
because we do exactly what is needed, and don't just mess with the
update ID for libass-based OSD.
'cuda-gl' isn't right - you can turn this on without any GL and
get some non-zero benefit (with the cuda-copy hwaccel). So
'cuda-hwaccel' seems more consistent with everything else.
When playing audio-only, and changing the audio output device, playback
froze until the next time the playback core happened to wakeup (like
moving the mouse, or OSD redrawing). This is probably because of the
awful statemachine in fill_audio_out_buffers() - just make it recreate
the AO directly instead.
Remove the per-part force_redraw flags, and instead make the difference
between flagging dirty state and returning it to the player frontend
more explicit. The big issue is that 1. the OSD needs to know the dirty
state, and it should be cleared strictly when it is re-rendered
(force_redraw flag), and 2. the player core needs to be notified once,
and the notification must be reset (want_redraw flag).
The call in loadfile.c is replaced by making osd_set_sub() set the
change flag. Increasing the change flag on dirty state (the force_redraw
check in render_object()) should not be needed, because OSD part
renderers set it correctly (at least now).
Doing this just because someone pointed this out.
This also lets you just do "mpv --hwdec file.mkv", with the minor caveat
that the legacy syntax "--hwdec val" or "-hwdec val" (without "=") does
not work as expected anymore.
The previous commit merely copied the profile string to a file (plus
changing how RPI-specific defaults are initialized), now make some
changes on top of it. In particular, remove the --input-lirc option,
which was removed a long time ago, but forgotten from the libmpv
profile.
Move the embedded string with the builtin profiles to a separate
builtin.conf file. This makes it easier to read and edit, and you can
also check it for errors with --include=etc/builtin.conf. (Normally
errors are hidden intentionally, because there's no way to output error
messages this early, and because some options might not be present on
all platforms or with all configurations.)
Just wow. This function is implemented in ipc-win.c, and was surely be
meant to be called. But it wasn't called. This could in theory cause
crashes during exit if IPC clients were active.
Untested whether it really works.
This workaround prevented that libmpv users could accidentally crash
when the SIGPIPE signal was triggered by FFmpeg's OpenSSL/GnuTLS usage.
But it also modifies the global signal handler state, so remove it now
that this workaround is not required anymore.
This happened to break because the texture unit wasn't reset to 0, which
some code expects. The OSD code in particular set the OSD texture on the
wrong texture unit, with the result that OSD/OSC was not visible.
A minor cleanup that makes the code simpler, and guarantees that we
cleanup the GL state properly at any point.
We do this by reusing the uniform caching, and assigning each sampler
uniform its own texture unit by incrementing a counter. This has various
subtle consequences for the GL driver, which hopefully don't matter. For
example, it will bind fewer textures at a time, but also rebind them
more often.
For some reason we keep TEXUNIT_VIDEO_NUM, because it limits the number
of hook passes that can be bound at the same time.
OSD rendering is an exception: we do many passes with the same shader,
and rebinding the texture each pass. For now, this is handled in an
unclean way, and we make the shader cache reserve texture unit 0 for the
OSD texture. At a later point, we should allocate that one dynamically
too, and just pass the texture unit to the OSD rendering code. Right now
I feel like vo_rpi.c (may it rot in hell) is in the way.
The caller now has to call gl_sc_reset(), and _after_ rendering. This
way we can unset OpenGL state that was setup for rendering. This affects
the shader program, for example. The next commit uses this to
automatically manage texture units via the shader cache.
vo_rpi.c changes untested.
Stops Mesa from restricting us to OpenGL 3.0. It also tries to create
GLES 3 contexts for drivers which do not just return a higher context
when requesting GLES 2.
I don't know whether this code is a good or bad idea. A not-so-good
aspect is that we don't check for EGL 1.5 (or 1.4 extensions) for some
of the more advanced context attributes. But EGL implementations should
be able to tolerate it and return an error, and then we'd use the
fallback.
This used to be shared, but since vo_rpi is going to be removed,
untangle them. There was barely any actual code shared since the recent
changes anyway.
As a subtle change, we also stop opening libGLESv2.so explicitly in the
vo_opengl backend, and use RTLD_DEFAULT instead.
Minimal support just for testing.
Only the window surface creation (including size determination) is
really platform specific, so this could be some generic thing with
platform-specific support as some sort of sub-driver, but on the other
hand I don't see much of a need for such a thing.
While most of the fbdev usage is done by the EGL driver, using this
fbdev ioctl is apparently the only way to get the display resolution.
Add a function to egl_helpers.c for creating an EGL context and make
context_x11egl.c use it. This is meant to be generic, and should work
with other windowing APIs as well. The other EGL-using code in mpv can
be switched to it.
Edit the 0.21.0 section: remove the redundant vo_opengl items, move some
up. Move the additions (which are less important and which aren't
documented completely anyway) below the incompatible
changes/deprecations.
The wrong enum got copied here, so it was essentially using the transfer
characteristics as the primaries (instead of the primaries), which
accidentally worked fine most of the time (since the two usually
coincided), but broke on weird/mistagged files.
Fixes missing subtitle tracks if the first entry didn't have any.
Previously it just checked for the first entry in the playlist for
requested languages and if that entry happened to not have subtitles
they also wouldn't show up for the other entries.
It will skip languages if the first entry with subs has less or
different languages than the others.
Unrelated to http_dash_segments.
The consequence of this was that e.g. hardware decoding with VAAPI-EGL
could sometimes not work if the compiler didn't support C11. (Although I
found this one on RPI, which also uses this mechanism.)
If the shader fails to compile, and assertion could trigger in
gl_sc_gen_shader_and_reset() due to the code trying to recreate the
shader every time, and re-appending the uniforms every time. Just reset
the uniform array to fix this.
Some disturbed GL drivers might not return anything for glGetShaderiv()
if the GL state got "lost", so initialize variables just for additional
robustness.
Since vo_rpi is going to be deprecated, better port its features to the
vo_opengl backend.
The most tricky part is the fact that recreating dispmanx elements will
conflict with the GL context. Fortunately, RPI's EGL support is
reasonably compliant, and we can transplant the context to newly created
dispmanx elements, making this much easier. This means unlike vo_rpi,
the GL state will actually not be recreated.
This overlay support specifically skips the OpenGL rendering chain, and
uses GL rendering only for OSD/subtitles. This is for devices which
don't have performant GL support.
hwdec_rpi.c contains code ported from vo_rpi.c. vo_rpi.c is going to be
deprecated. I left in the code for uploading sw surfaces (as it might
be slightly more efficient for rendering sw decoded video), although
it's dead code for now.
When update_prop() successfully fetches a changed property value, it
sets prop->changed to true. mark_property_changed() only sets
prop->need_new_value if prop->changed is false, so this had the effect
of ignoring new property values until prop->changed was set back to
false in the next call to gen_property_change_event(). This meant that
when a property change event was generated for a property that was not
observed with MPV_FORMAT_NONE, it would contain the value associated
with the earliest property change, rather than the most recent, and the
property change event for the most recent change would never be
generated.
To fix this, mark_property_changed() should unconditionally set
prop->changed and prop->need_new_value, which will cause the property
value to be re-fetched and a property change event to be generated for
the most recent value.
The cuvid decoder already knows how to copy back to system memory
if NV12 frames are requested, and this will happen if the decoder
is used without the hwdec.
For convenience, let's add a wrapper hwdec so people don't have
to explicitly pick the cuvid decoder if they want this behaviour.
This will catch cases where mpv runs without error, but the
--list-options output isn't what we expect. Otherwise, we'll make a
broken completion file that will result in cryptic errors when pressing
tab, like:
_mpv:18: command not found: *:files:->mfiles
That's been the case for most of the zsh completion issues we've had
reported, that I can remember.
Also make uninitialized variable access fatal so that failures to parse
other options will also make the script die eventually, albeit with a
less nice message.
--list-protocol was printing a *:// entry, which looked strange at best.
The "*" protocol was used to always match everything, so stream_cb.c
could hook in custom protocols with a prefix chosen by the API user.
Change it instead so that an empty protocol list means "match all",
which also gets rid of the special-cased "*" entry.
This is an awful corner-case party, because we've started to allow the
user to send track-switching related commands before stream selection is
done in the loading stage. If mp_switch_track() is called before this
stream selection, it can behave inconsistently. So if we're in the
pre-loading phase, we must set the stream selection options to get
streams selected later, instead of calling this function. There seem to
be some annoying circumstances that exclude mp_switch_track() from
handling this logic too, so do it at all call-sites.
The nvidia examples use the old (as in CUDA 3.x) interop API which
is deprecated, and I think not even functional on recent versions
of CUDA for windows. As I was following the examples, I used this
old API.
So, let's update to the new API, and hopefully, it'll start working
on windows too.
This has all been made unnecessary recently. The change not to copy the
global option struct in particular can be made because now nothing
accesses the global options anymore in the demux and stream layers.
Some code that was accidentally added/changed in commit 5e30e7a0 is also
removed, because it was simply committed accidentally, and was never
used.
Mostly untested.
This is not compatible. It removes the URL fields for track range and
cdrom speed (what did this even do). The device is not not to be
prefixed with an additional "/" if it's put into the URL. I can't be
bothered to keep these things compatible, just rip your damn CDs
instead.
Instead of deselecting the video stream plainly, use the slightly more
robust error_on_track() function. Also give it an error code (although
I'm not sure if this one is confusing, it's better than the one before).
Instead, parse manually. This is to get rid of the option API usages,
which seem unnecessary and shoehorned. (Just look at the URL pseudo
parsing and the dumb url_options map. They were pretty much artifacts
from refactoring old mplayer code.)
It was used to determine whether the VO supports VOCTRL_SET_PANSCAN.
With all those changes to property semantics this became unnecessary,
and its only use was dropped at some point.
Just another corner-caseish potential issue. Unlike unreffing the image
manually, unref_current_image() also takes care of properly unmapping
hwdec frames. (The corner-case part of this is that it's probably never
mapped at this point, but it's apparently not entirely guaranteed.)
The " || vimg->mpi" part virtually never seems to trigger, but on the
other hand could possibly create unintended corner cases (for example by
trying to upload a NULL image, which would then be marked as an error
and render a blue screen).
I guess it's a leftover from over times, where a NULL image meant
"redraw the current frame". This is now handled by actually passing
along the current frame.
Nvidia's "NvDecode" API (up until recently called "cuvid" is a cross
platform, but nvidia proprietary API that exposes their hardware
video decoding capabilities. It is analogous to their DXVA or VDPAU
support on Windows or Linux but without using platform specific API
calls.
As a rule, you'd rather use DXVA or VDPAU as these are more mature
and well supported APIs, but on Linux, VDPAU is falling behind the
hardware capabilities, and there's no sign that nvidia are making
the investments to update it.
Most concretely, this means that there is no VP8/9 or HEVC Main10
support in VDPAU. On the other hand, NvDecode does export vp8/9 and
partial support for HEVC Main10 (more on that below).
ffmpeg already has support in the form of the "cuvid" family of
decoders. Due to the design of the API, it is best exposed as a full
decoder rather than an hwaccel. As such, there are decoders like
h264_cuvid, hevc_cuvid, etc.
These decoders support two output paths today - in both cases, NV12
frames are returned, either in CUDA device memory or regular system
memory.
In the case of the system memory path, the decoders can be used
as-is in mpv today with a command line like:
mpv --vd=lavc:h264_cuvid foobar.mp4
Doing this will take advantage of hardware decoding, but the cost
of the memcpy to system memory adds up, especially for high
resolution video (4K etc).
To avoid that, we need an hwdec that takes advantage of CUDA's
OpenGL interop to copy from device memory into OpenGL textures.
That is what this change implements.
The process is relatively simple as only basic device context
aquisition needs to be done by us - the CUDA buffer pool is managed
by the decoder - thankfully.
The hwdec looks a bit like the vdpau interop one - the hwdec
maintains a single set of plane textures and each output frame
is repeatedly mapped into these textures to pass on.
The frames are always in NV12 format, at least until 10bit output
supports emerges.
The only slightly interesting part of the copying process is that
CUDA works by associating PBOs, so we need to define these for
each of the textures.
TODO Items:
* I need to add a download_image function for screenshots. This
would do the same copy to system memory that the decoder's
system memory output does.
* There are items to investigate on the ffmpeg side. There appears
to be a problem with timestamps for some content.
Final note: I mentioned HEVC Main10. While there is no 10bit output
support, NvDecode can return dithered 8bit NV12 so you can take
advantage of the hardware acceleration.
This particular mode requires compiling ffmpeg with a modified
header (or possibly the CUDA 8 RC) and is not upstream in ffmpeg
yet.
Usage:
You will need to specify vo=opengl and hwdec=cuda.
Note that hwdec=auto will probably not work as it will try to use
vdpau first.
mpv --hwdec=cuda --vo=opengl foobar.mp4
If you want to use filters that require frames in system memory,
just use the decoder directly without the hwdec, as documented
above.
This is a bug fix, and the text alignment functionality probably got
lost sometime along the way.
For ASS subtitles, this could have unintended consequences, so it's hard
to get right - thus it's not applied to ASS subtitles.
For other text subtitles, this should be fine, though. It still works on
ASS subtitles as promised by the manpage if --no-sub-ass is used.
The first one is printed even if the user disabled video (or there's no
video), so just remove it. The second one uses deprecated sub-option
syntax, so remove that as well.
This time it's emulation that's supposed to work (not just dummied out).
Unlike the previous emulation, no mpv code has to be disabled, and
everything should work (albeit possibly a bit slowly). On the other
hand, it's not possible to implement this kind of emulation without
compiler support. We use GNU statement expressions and __typeof__ in
this case.
This code is inactive if stdatomic.h is available.
Don't access MPOpts directly, and always use the new m_config.h
functions for accessing them in a thread-safe way.
The goal is eventually removing the mpv_global.opts field, and the
demuxer/stream-layer specific hack that copies MPOpts to deal with
thread-safety issues.
This moves around a lot of options. For one, we often change the
physical storage location of options to make them more localized,
but these changes are not user-visible (or should not be). For
shared options on the other hand it's better to do messy direct
access, which is worrying as in that somehow renaming an option
or changing its type would break code reading them manually,
without causing a compilation error.
Instead of copying the options around... just don't. video.c now has
full control over when options are updated. (It still gets notified from
outside, but it decides when the updated options are copied: when
m_config_cache_update() is called.) So there's no need for tricky
stuff, and it can be simplified a bit.
Also change lcms.c. We could do it like video.c, and get the options
from the global config store. But it seems simpler to just provide a
pointer to an option struct, which is arbitrarily mutated from the
outside (from the perspective of lcms.c).
Setting --icc-profile had no effect, until a vo_opengl option was
changed at runtime. We must initialize the renderer for the initial
option state too.
For some reason, the ICC profile gets loaded twice. The next commit
happens to fix this.
It has to copy each option, whether it's deprecated or not. This would
print a warning on every deprecated sub-option, even if it's not used.
Yep, this is very stupid.
At least m_config_get_co() gets actually slightly cleaner, because it
separates the search and the deprecation handling.
And introduce a global option which does this. Or more precisely, this
deprecates the global wasapi and coreaudio options, and adds a new one
that merges their functionality. (Due to the way the sub-option
deprecation mechanism works, this is simpler.)
I decided that it's too much work to convert all the VO/AOs to the new
option system manually at once. So here's a shitty hack instead, which
achieves almost the same thing. (The only user-visible difference is
that e.g. --vo=name:help will list the sub-options normally, instead of
showing them as deprecation placeholders. Also, the sub-option parser
will verify each option normally, instead of deferring to the global
option parser.)
Another advantage is that once we drop the deprecated options,
converting the remaining things will be easier, because we obviously
don't need to add the compatibility hacks.
Using this mechanism is separate in the next commit to keep the diff
noise down.
Instead of requiring each VO or AO to manually add members to MPOpts and
the global option table, make it possible to register them automatically
via vo_driver/ao_driver.global_opts members. This avoids modifying
options.c/options.h every time, including having to duplicate the exact
ifdeffery used to enable a driver.
Instead of adding a lock_frame to the list when mp_dispatch_lock() is
called, just set a simple flag. This uses the fact that the lock is not
recursive, and can happen once per mp_dispatch_queue_process(). It
avoids the dynamic allocation, and makes error checking slightly
stricter.
Again, this is actually redundant and exists only for error-checking.
It'd actually need only a counter, because the actual locking is done by
"parking" the target thread in mp_dispatch_queue_process() and then
setting queue->idling=false. Only when mp_dispatch_unlock() sets it to
true again other work can proceed again. Document this too.
A deadlock bug was reported with the following test program:
mpv_handle *mpv = mpv_create();
mpv_set_option_string(mpv, "ytdl", "yes");
mpv_initialize(mpv);
mpv_terminate_destroy(mpv);
The cause of this is loading the ytdl.lua script, which triggers a
certain code path that calls mp_dispatch_queue_process() recursively. It
does so to wait until the script is loaded, and we want to keep that.
Reentrancy was not supported by mp_dispatch, which leads to the
deadlock. Rewrite the locking so that it does. We mainly get rid of the
"exclusive_lock" mutex. Instead we use the existing lock/condition
variable to wait until we can grab a logical lock.
Note that the lock_frame business can be replaced with a simple counter.
Instead of checking the lock_frame address, it'd simply increment and
store the counter when entering mp_dispatch_queue_process(), and then
compare the counter to decide whether or not to wait. But I think the
additional error checking done by the lock_frame list is valuable.
Fixes#3489.
Why do these API calls even still exist? I don't know, and maybe they
don't make any sense anymore. But whether they should be removed or not
is not a decision I want to make now. I want to get rid of
mp_dispatch_suspend/resume(), though. So implement the client APIs
slightly differently.
Whitelisting supported codecs is (probably) still better than just
allowing everything, given the weird FFmpeg API. I'm also assuming
Libav doesn't even have the codec ID, but I didn't check.
Also add a --teletext-page option, since otherwise it decodes every
teletext page and shows them in succession.
And yes, we can't use av_opt_set_int() - instead we have to set it as
string. Because FFmpeg's option system is terrible.
This affects the "cycle" command. If we switched to the next track, and
it failed to initialize, we just deselected everything.
Change it so that if initialization fails early (typically decoder
selection), we try to continue with the track after that. (Even if
nothing can be selected, the loop will terminate when trying to select
nothing.
Fixes#3446.
This was mistakenly added. It was removed from the vo_opengl_hq defaults
at an earlier time, but the documentation was not updated, which is why
it made it back into the profile.
Fixes#3485.
These never made any sense. They checked the --vo/--ao option, and
applied the profile corresponding to the first entry. So the only way to
get any use of those was to use the --ao or --vo option explicitly. You
can get the same functionality by making a manual profile, making these
force the ao/vo, and then using --profile on command line instead of
--vo/--ao.
This works by first parsing a config file into the default profile, and
applying it once parsing the whole file is finished.
This won't work across config files (not even if you include other
config files via "include=file.conf").
Apparently this was supposed to be handled - but badly at best. Make
unset values always have the value "" instead of NULL to avoid
special-cases.
In particular, this fixes passing NULL to a %s format specifier to
printf in show_profile(). Glibc prints this as "(null)", but it's
undefined behavior, and other libcs can crash.
On a VOCTRL_UPDATE_PLAYBACK_STATE store the state, and use it to
initialize the next time the task list becomes available.
This actually fixes#3482. Revert commit f2e25e9e because it's not
needed anymore.
With the recent vo_opengl changes it doesn't do anything anymore.
I don't think a deprecation period is necessary, because the command
was always marked as experimental.
With the conversion from sub-options to global options, this becomes
useless. This change also comes slightly too soon, because not all VOs
have been changed yet.
vo_opengl sub-option were always rather annoying to handle. It seems
better to make them global options instead. This is simpler and easier
to use. The only disadvantage we are aware of is that it's not clear
that many/all of these new global options work with vo_opengl only.
--vo=opengl-hq is also deprecated.
There is extensive compatibility with the old behavior. One exception is
that --vo-defaults will not apply to opengl-hq (though with opengl it
still works). vo-cmdline is also dysfunctional and will be removed in a
following commit.
These changes also affect opengl-cb.
The update mechanism is still rather inefficient: it requires syncing
with the VO after each option change, rather than batching updates.
There's also no granularity (video.c just updates "everything", and if
auto-ICC profiles are enabled, vo_opengl.c will fetch them on each
update).
Most of the manpage changes were done by Niklas Haas <git@haasn.xyz>.
This is still rather basic.
run_reconfig() and run_control() update the options because it's needed
for panscan (and other video scaling options), and fullscreen, border,
ontop updates. In the old model, these options could be accessed only
while both playback thread and VO threads were locked (i.e. during
synchronous calls like vo_control()), so this should be sufficient in
order not to miss any updates. In the future, a more fine-grained update
mechanism could be added to handle these updates "exactly".
x11_common.c contains an evil hack, as I see no reasonable way to handle
this properly. The VO thread can't "lock" the main thread, so this is
not simple.
The way option runtime changes are handled is pretty bad in the current
codebase. There's a big option struct (MPOpts), which contains almost
everything, and for which no synchronization mechanism exists. This was
handled by either making some options read-only after initialization,
duplicating the option struct, using sub-options (in the VO), and so on.
Introduce a mechanism that creates a copy of the global options (or
parts of it), and provides a well-defined way to update them in a
thread-safe way.
Most code can remain the same, just that all the component glue code has
to explicitly make use of it first.
There is still lots of room for improvement. For example, the update
mechanism could be better.
If the win32 taskbar progress update is sent before the VO window is
created, then w32_common.c will ignore it because the actual taskbar
object was not created yet. (At least this is what I suspect happens.
The window is already created at this point, but not mapped.)
Hopefully fix this is fixed by creating until after the window is
created, i.e. the VO has been configured at least once.
Untested (who wants to boot into Windows just to wait until it has
applied all of its stupid updates).
Also not explicit is whether update_vo_playback_state() will actually be
called soon enough in all cases. It probably is.
Probably fixes#3482.
Reduce accesses to the renderer opts in vo_opengl.c, and instead add
accessors for them to video.c.
I suppose gamma and maybe icc-auto could be moved to vo_opengl.c
options. Also, the output colorspace could probably be adjusted to what
is really used, not just the options (although it's possible that this
commit changes this, due to video.c mutating its own copy of the options
according to actual renderer capapbilities).
But don't deal with this now.
Normally I'd prefer a bunch of smaller functions with fewer parameters
over a single function with a lot of parameters. But future changes will
require messing with the parameters in a slightly more complex way, so a
combined function will be needed anyway. The now-unused "global"
parameter is required for later as well.
iconv distinguishes between euc-kr and cp949, while libguess
and libuchardet doesn't (only returns euc-kr). EILSEQ occurs
when the input encoding of iconv is set to euc-kr and if the subs
contain letters not included in euc-kr. Since cp949 is a extension
of euc-kr, choose cp949 instead.
Signed-off-by: wm4 <wm4@nowhere>
And replace the sort-of duplicated explanations.
(It's a bit funny to use weblinks to the generated web version of itself
instead of proper RST links, but I think I don't care.)
The --cache option and cache property conflict, so one of them has to be
renamed. The option is probably used frequently, so initiate
deprecation/rename of the property.
Deprecated in favor of user-shaders, which are functionally equivalent
but superior. (Except in the case of scaler-shader, which has no direct
replacement, but it turned out to be a very unpopular feature either way
- most custom scalers don't fit into the mpv kernel infrastructure and
are therefore implemented as user shaders either way)
Signed-off-by: wm4 <wm4@nowhere>
Create the core thread right in mpv_create(), and reduce what
mpv_initialize() does further. This is simpler, and allows the API user
to do more before calling mpv_initialize(). The latter is not the real
goal, rather we'd like mpv_intialize() reduced to do almost nothing. It
still does a lot, but nothing truly special anymore that is absolutely
required for basic mpv workings.
One thing we want the user to be able to do is changing properties
before mpv_initialize() to reduce the special status of
mpv_set_option().
Make some existing properties behave more like options. This mostly
means they don't deny access if the associated component is not active,
but redirects to the option.
One kind of fishy change is that we apply --brightness etc. only if
they're not set to the default value. This won't necessarily work with
--vo=xv, but affects only cases where 1. the Xv adapter has been changed
to non-defaults, and 2. the user tries to reset them with mpv by passing
e.g. --brightness=0. We don't care about Xv, and the noted use-case is
dumb, so this change is acceptable.
These conflict with options of the same name, and prevent a "full"
unification. Not addressed is the "cache" property, and possibly a few
properties that behave differently from their equivalent options.
Now options are accessible through the property list as well, which
unifies them to a degree.
Not all options support runtime changes (meaning affected components
need to be restarted for the options to take effects). Remove from the
manpage those properties which are cleanly mapped to options anyway.
From the user-perspective they're just options available through the
property interface.
E.g. --vf=scale=no-arnd didn't work, because it didn't recognize no-arnd
as flag option.
The top-level command line parser is not affected by this, because it
uses the result of m_config_option_requires_param() differently and
assumes unknown parameters do not necessarily require a parameter. (The
suboption parser can't do this.)
Positional parameters cause problems because they can be ambiguous with
flag options. If a flag option is removed or turned into a non-flag
option, it'll usually be interpreted as value for the first sub-option
(as positional parameter), resulting in very confusing error messages.
This changes it into a simple "option not found" error.
I don't expect that anyone really used positional parameters with --vo
or --ao. Although the docs for --ao=pulse seem to encourage positional
parameters for the host/sink options, which means it could possibly
annoy some PulseAudio users.
--vf and --af are still mostly used with positional parameters, so this
must be a configurable option in the option parser.
seek_fn is supposed to return the new file offset, or a negative error
code. Our code doesn't use the offset, and only wants to know if any
errors happened. The int cast is completely broken and might treat a
successful seek as failed depending on whether the sign bit will be set.
Instead, add a hacky OPT_ASPECT option type, which only exists to accept
a "no" parameter, which in combination with the "--no-..." handling code
makes --no-video-aspect work again.
We can also remove the code in m_config.c, which only existed to make
"--no-aspect" (a deprecated alias) to work.
This is a really old weird MPlayer feature. While the MPlayer requires
you to use the sub-option syntax in these cases, mpv "flattens" them to
normal options. The still-supported alternate sub-option syntax remains
a weird artifact that hopefully nobody uses.
For example you can do "-sub-text font=Foo:color=0.5" instead of using
"--sub-text-font=Foo --sub-text-color=0.5". For --sub-text this is an
accidental feature, but it used to be documented for
--demuxer-rawaudio and some others.
This should just be removed, but for now only print a warning to preempt
complaints from weird users wanting this feature back.
The client API can do this (and there are apparently some libmpv using
projects which rely on this). But it's just unnecessary bloat as it
requires a separate code path from the option parser. It would be better
to remove this code. Formally deprecate it, including API bump and
warning in the API changes file to make it really clear.
OPT_ALIAS redirects the options at a higher level, instead of
introducing "duplicate" options with different name but same backing
storage. This uses the OPT_REPLACED mechanisms - only the deprecation
warning had to be made conditional. Note that e.g. --no-video still
works, because the "--no-..." redirection and OPT_ALIAS are orthogonal.
The deprecated --sub -> --sub-file alias had to be dropped, because it
essentially conflicts with --no-sub. If anyone complains, this could
probably still be undone by letting m_config_find_negation_opt do a
special mapping for --no-sub. (Which would be dumb, but simple and
effective.)
Instead of adding "no-"-prefixed aliases to the internal option list,
which will act like normal options, do it in the parsing stage. This
turns out to be simpler (and cheaper), and avoids adding aliased
options.
This is basically dead code, and even the commit that added this code 4
years ago said that this should be for debugging only.
(Though it is possible that the clamp callback was used for something
else, and then unused again. Also, some of the clamping code remains and
is used for internal checking, e.g. clamp_double().)
Before this commit, all VOs had to toggle the option flag themselves,
now command.c does it.
I can't really comprehend why it required every VO to do this manually.
Maybe it was for rejecting the property/option change if the VO didn't
support a specific capability. But then it could have checked the VOCTRL
result. In any case, I don't care, and successfully changing the
property without doing anything (With some VOs) is fine too. Many things
work this way now, and it's simpler overall.
This change will be useful for cleaning up VO option handling.
Just a minor refactor along the planned option change. This commit will
make it easier to update (i.e. copy) the VO options without copying
_all_ options. For now, behavior should be equivalent, though.
(The VO options were put into a separate struct quite early - when all
global variables were removed from the source code. It wasn't clear
whether the separate struct would have any actual purpose, but it seems
it will now. Awesome, huh.)
More appropriate. Originally it really was for automatically added
options, but now it even needed some stupid comments to indicate that
it was used for simply hiding options.
It's actually redundant with whether m_option_type.free is set. Some
option types were flagged inconsistently. Its only use was for running
an additional sanity check without any real functionality.
Some files not only use rounded timestamps, but they also do it
incorrectly. They may jitter between up to 4 specific frame durations.
In this case, I found a file that mostly used 41ms and 42ms, but also
had 40ms and 43ms outliers (often but not always following each other).
This breaks the assumption of the framerate estimation code that the
frame duration can deviate up to 1ms. If it jitters around 4 possible
frame durations, the maximum deviation is 3ms. Increase it accordingly.
The change might make playback of "true VFR" video via display-sync mode
worse, but it's not like it was particularly good in the first place.
Also, the check whether to usen the container FPS needs to be stricter.
In the worst case, num_dur is 1, which doesn't really indicate any
evidence that the framerate is correct. Only if there are "enough"
frames the deviation check will become meaningful. 16 is an arbitrary
value that has been designated "enough" by myself.
Also otuput the frame duration values for --dump-stats.
It seems like many GL implementations (including Mesa) choke on this,
while others are fine. We still think that this use of the GL API is
allowed by the standard (at least in the Mesa case), so to reduce
confusion, explicitly check the "controversial" calls, and use an
appropriate error message.
Like it's done for audio and video. Just to be uniform.
I'm sorry for deleting the anti-ffmpeg vitriol. It's still all true, but
since we decided to always set the timebase, the crappiness is isolated
to FFmpeg internals.
These are different AVCodecContext fields. pkt_timebase is the correct
one for identifying the unit of packet/frame timestamps when decoding,
while time_base is for encoding. Some decoders also overwrite the
time_base field with some unrelated codec metadata.
pkt_timebase does not exist in Libav, so an #if is required.
Normally, OSD can be disabled with --osd-level=0. But this also disables
terminal OSD, and some users want _only_ the terminal OSD. Add
--video-osd=no, which essentially disables the video OSD.
Ideally, it should probably be possible to control terminal and video
OSD levels independently, but that would require separate OSD timers
(and other state) for both components, so don't do it. But because the
current situation isn't too ideal, add a threat to the manpage that
might be changed in the future.
Fixes#3387.
This is for text subtitles. libavformat currently always reads text
subtitles completely on init. This means the underlying stream is
useless and will consume resources for various reasons (network
connection, file handles, cache memory).
Take care of this by closing the underlying stream if we think the
demuxer has read everything. Since libavformat doesn't export whether it
did (or whether it may access the stream again in the future), we rely
on a whitelist. Also, instead of setting the stream to NULL or so, set
it to an empty dummy stream. This way we don't have to litter the code
with NULL checks.
demux_lavf.c needs extra changes, because it tries to do clever things
for the sake of subtitle charset conversion.
The main reason we keep the demuxer etc. open is because we fell for
libavformat being so generic, and we tried to remove corresponding
special-cases in the higher-level player code. Some of this is forced
due to ass/srt mkv/mp4 demuxing being very similar to external text
files. In the future it might be better to do this in a more
straight-forward way, such as reading text subtitles into libass and
then discarding the demuxer entirely, but for aforementioned reasons
this could be more of a mess than the solution introduced by this
commit.
Probably fixes#3456.
Cleaner and makes it easier to change the underlying stream.
mp_property_stream_capture() still directly accesses it directly via
demux_run_on_thread(). This is evil, but still somewhat sane and is not
getting into the way here.
Not sure if I got all field accesses.
It's just wasted memory.
One corner case is when a file grows during playback, but this is rare
and usually happens on-disk only. The cache size was generally limited
before this change already, so no reason to care.
As an unrelated change, move the cache size info to the resize_cache()
function. There's really no reason not to do this, and it's slightly
more informative if the user changes the cache size at runtime.
Because VOCTRL_CHECK_EVENTS is processed asynchronously (as of 088a007,)
the GUI thread no longer gets regular wakeups, so the old check that
made sure the video window matched the parent window's size in --wid
embedding mode did not run very often. This made --wid embedding not
very usable.
Instead of polling for window size changes, use Windows hooks to react
to them when they happen. When the parent window is owned by the same
process as the video window, use a WH_CALLWNDPROC hook. When the parent
window is not owned by the same process, WinEvents must be used, which
are not as smooth, but still work for this purpose.
Since neither SetWindowsHookEx nor SetWinEventHook take a context
parameter to send data to the hook function, the hook functions must
find the child window by its class instead, so there are a few changes
to ensure this is fast and the class is unique.
This also fixes up the logic to handle window destruction. When a parent
window is destroyed, its children are also destroyed, so this gives us a
way to react to parent window destruction without polling.
If the video has the same size as the screen, starting with --fs and
then leaving fullscreen doesn't actually leave fullscreen.
The reason is that mpv tries to restore the previous window size if
necessary (otherwise, you'd end up with a Window of nearly the same size
as the screen with some WMs). It will typically restore with the
rectangle set exactly to the screen if no other position or size is
forced. This triggers pre-EWMH fullscreen mode, which WMs detect using
various heuristics.
Apparently we triggered this with mutter (but strangely no other WMs).
It's possible that pre-EWMH fullscreen mode actually requires removing
decorations, and mutter either ignores this. But this is speculation and
I haven't checked.
Work this around by reducing the requested size by 1 pixel if it
happens.
This was observed with mutter 3.18.2.
Fixes#2072.
If spdif is enabled, the channel layout has no meaning other than
setting the number of channels. The number of channels must be fixed to
achieve the exact bitrate required.
Fixes#3445.
It doesn't necessarily have to mean anything bad.
We're still too lazy to provide any more detailed information (e.g.
whether this happened to likely bad interleaving, excessive amount of
packets like with some ASS subs, or that the readahead user option is
limitted by the packet queue size).
This should actually be rather safe - we already check whether the
estimated value jitters less than the (possibly untrustworthy) nominal
one. Remove a "safety" check that disabled this code for small
deviations, and make it trigger sooner into playback. Also lower the log
level of messages about using the estimated display FPS down to verbose.
Normally there's another mechanism for smoothing out minor estimation
differences, but that is not good enough here.
This possibly improves behavior as reported in #3433, which can be
reproduced with --vo=null:fps=48.426 --display-fps=48 (though it doesn't
consider the jitter introduced by a real VO).
Doing this required synchronizing with the VO thread, which could lead
to audio dropouts if the VO was frozen (which can happen in practice if
e.g. an opengl_cb user is not doing what the API demands).
Add a way to send asynchronous VOCTRLs, and use that for the playback
state. In theory, it would be better to make this status update a
several function and to "merge" several queued update, but that would be
slightly more effort/code, and the update is so infrequent that the
merging would never happen anyway.
The change to vo_destroy() is to make sure all queued asynchronous
reuqests are finished before making the vo_thread exit.
Even though it's only used on MS Windows, it's run on any platform with
any VO, which makes this worse.
run_control() dereferences an uint32_t as int. Whether this is allowed
depends on what uint32_t is typedefed to (dereferencing an unsigned int
as int should be fine). Fix it by always using int. The uint32_t type
never really made sense.
There are situations where the resampler is destroyed and recreated
during playback. If recreating the resampler unexpectedly fails, the
filter function is supposed to return an error. This wasn't done
correctly, because get_out_samples() accessed the resampler before the
check. Move the check up to fix this.
When fetching the playlist property, playlist_entry_from_index would be
called for each playlist entry, which traversed a linked list to get the
entry corresponding to the specified index. This was very slow for large
playlists. Since get_playlist_entry is called for each index in order,
it can avoid a full traversal of the linked list by using the next
pointer on the previously requested entry.
Instead of passing through double float timestamps opaquely, pass real
timestamps. Do so by always setting a valid timebase on the
AVCodecContext for audio and video decoding.
Specifically try not to round timestamps to a too coarse timebase, which
could round off small adjustments to timestamps (such as for start time
rebasing or demux_timeline). If the timebase is considered too coarse,
make it finer.
This gets rid of the need to do this specifically for some hardware
decoding wrapper. The old method of passing through double timestamps
was also a bit questionable. While libavcodec is not supposed to
interpret timestamps at all if no timebase is provided, it was
needlessly tricky. Also, it actually does compare them with
AV_NOPTS_VALUE. This change will probably also reduce confusion in the
future.
This affects A-B loops and --loop-file, and audio. Instead of dropping
audio by resetting the AO, try to make it seamless by not sending data
after the loop point, and after the seek send new data without a reset.
The code actually kept going out of EOF mode into resync mode back into
EOF mode when the playloop had to wait after an audio EOF caused by the
endpts. This would break seamless looping (as added by the next commit).
Apply endpts earlier, to ensure the filter_audio() function always
returns AD_EOF in this case.
The idiotic ao_buffer makes this an amazing pain in the ass.
Instead of letting it keep decoding by trying to find a new frame,
"plug" the frame queue by not removing it. (Or actually, by putting
it back instead of discarding it.)
Matters for seamless looping (following commits), and possibly some
other corner cases.
The added function vf_unread_output_frame() is a bit of a sin, but still
reasonable, since its implementation is trivial.
The --image-display-duration option controls how long an image is
displayed. It's also possible to display the image forever (until manual
user interaction stops playback).
With this, the core drops the old method to "drain" video (i.e. waiting
for the last frame duration on end of playback). Instead, we reuse
MPContext.time_frame. The old mechanism was disabled for non-images
anyway.
Fixes#3425.
When an ogg track upodates metadata, we have to perform a complicated
runtime update due to the demux.c architecture. A detail was broken and
an array was allocated with the previous number of streams, which
usually led to invalid memory write accesses at least on the first
update.
See github commit comment on commit b9ba9a89.
The touched code is for seek resets and such - we simply want to reset
the entire resample state. But I noticed after a seek a tiny bit of
audio is missing (mpv's audio sync code inserted silence to compensate).
It turns out swr_drop_output() either does not reset some internal state
as we expect, or it's designed to drop not only buffered samples, but
also future samples.
On the other hand, libavresample's avresample_read(), does not have this
problem. (It is also pretty explicit in what it does - return/skip
buffered data, nothing else.)
Is the libswresample behavior a bug? Or a feature? Does nobody even
know? Who cares - use the hammer to unfuck the situation. Destroy and
deallocate the libswresample context and recreate it. On every seek.
Change the last parameter from a bool to an int, which is supposed to
take bit-flags. The at this point only flag is MPSEEK_FLAG_DELAY, which
replaces the previous bool parameter. The old false parameter becomes 0,
the old true parameter becomes MPSEEK_FLAG_DELAY.
Since the old "immediate" parameter is now essentially inverted, two
coalesced immediate and delayed seeks end up as delayed instead of
immediate. This change doesn't matter, since there are no relative
immediate seeks anyway.
The accepts_packet packet callback is supposed to deal with subtitle
decoders which have only a small queue of current subtitle events (i.e.
sd_lavc.c), in case feeding it too many packets would discard events
that are still needed.
Normally, the number of subtitles that need to be preserved is estimated
by the rendering pts (get_bitmaps() argument). Rendering lags behind
decoding, so normally the rendering pts is smaller than the next video
frame pts, and we simply discard all subtitle events until the rendering
pts.
This breaks down in some annoying corner cases. One of them is seeking
backwards: the VO will still try to render the old PTS during seeks,
which passes a high PTS to the subtitle renderer, which in turn would
discard more subtitles than it should. There is a similar issue with
forward seeks. Add hacks to deal with those issues.
There should be a better way to deal with the essentially unknown
"rendering position", which is made worse by screenshots or rendering
with vf_sub. At the very least, we could handle seeks better, and e.g.
either force the VO not to re-render subs after seeks (ugly), or
introduce seek sequence numbers to distinguish attempts to render
earlier subtitles when a seek is done.
If the PEAK tag is invalid, return an error.
Make the error signalling conventions more uniform by strictly returning
a negative value on error, and treating >=0 as success.
The demuxer layer usually doesn't log per-stream information, and even
the replaygain information was logged only if it came from tags.
So log it in af_volume instead.
...and ignore it. The main purpose is for retrieving per-track
replaygain tags. Other than that per-track tags are not used or accessed
by the playback core yet.
The demuxer infrastructure is still not really good with that whole
synchronization thing (at least in part due to being inherited from
mplayer's single-threaded architecture). A convoluted mechanism is
needed to transport the tags from demuxer thread to user thread. Two
factors contribute to the complexity: tags can change during playback,
and tracks (i.e. struct sh_stream) are not duplicated per thread.
In particular, we update the way replaygain tags are retrieved. We first
try to use per-track tags (common in Matroska) and global tags
(effectively formats like mp3). This part fixes#3405.
With the previous commit, ao_alsa.c now has 3 possible ways to pause
playback. Actually all 3 of them need get_delay() to fake its return
value, so don't duplicate that code.
Also much of the code looks a bit questionable when considering
inconsistent pause/resume calls from outside, so ignore redundant calls.
push.c does not handle this automatically, and AOs using push.c have to
handle it themselves. Also, ALSA is low-level enough that it needs
explicit support in user code. At least I haven't found any option that
does this.
We still can get away relatively cheaply by abusing underflow-handling
for this. ao_alsa.c already configures ALSA to handle underflows by
playing silence. So we purposely induce an underflow when opening the
device, as well as when pausing or resetting the device.
This introduces minor misbehavior: it doesn't account for the additional
delay the initial silence adds, unless the device has fully played the
fragment of silence when the player starts sending data to it. But
nobody cares.
The current stdatomic check verifies the availability of the function by
calling atomic_load(). It also uses this test to check if linking
against libatomic is needed or not.
Unfortunately, on specific architectures (namely SPARC), using
atomic_load() does *not* require linking against libatomic, while other
atomic operations do. Due to this, mpv's wscript concludes that
stdatomic is available, and that linking against libatomic is not
needed, causing the following link failure:
[190/190] Linking build/mpv
audio/out/ao.c.13.o: In function `ao_query_and_reset_events':
/home/peko/autobuild/instance-0/output/build/mpv-0.18.1/build/../audio/out/ao.c:399: undefined reference to `__atomic_fetch_and_4'
In order to fix this, the stdatomic check is adjusted to call
atomic_fetch_add() instead, which does require libatomic. Thanks to
this, the wscript realizes that linking against libatomic is needed, and
the build works fine.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
If the normal stream cache init fails, and a file cache was initialized
before, we free the file cache as well. But since the file cache is
chained to the real stream, the real stream will also be freed. This has
to be prevented by clearing the pointer to the original stream in the
uncached_stream field.
This could in particular be triggered by using --cache-initial=1000 and
aborting playback during loading. (Without that option, stream cache
init failure is far less likely.)
Forgotten in previous commit.
Also minor semi-related change: remove the extra "," from the
mpv_sub_api enum, which I accidentally added in the previous commit.
(C99 is fine with trailing ",", C89 strictly speaking not. So do
this for maximum compatibility.)
Relative seeks backwards with external audio tracks does not always work
well: it tends to happen that video seek back further than audio, so
audio will remain silent until the audio's after-seek position is
reached. This happens because we strictly seek both video and audio
demuxer to the approximate desirted target PTS, and then start decoding
from that.
Commit 81358380 removes an older method that was supposed to deal with
this. It was sort of bad, because it could lead to the playback core
freezing by waiting on network.
Ideally, the demuxer layer would probably somehow deal with such seeks,
and do them in a way the audio is seeked after video. Currently this is
infeasible, because the demuxer layer assumes a single demuxer, and
external tracks simply use separate demuxer layers. (MPlayer actually
had a pseudo-demuxer that joined external tracks into a single demuxer,
but this is not flexible enough - and also, the demuxer layer as it
currently exists can't deal with dynamically removing external tracks
either. Maybe some time in the future.)
Instead, add a gross hack, that essentially reseeks the audio if it
detects that it's too far off. The result is actually not too bad,
because we can reuse the mechanism that is used for instant track
switching. This way we can make sure of the right position, without
having to care about certain other issues.
It should be noted that if the audio demuxer is used for other tracks
too, and the demuxer does not support refresh seeking, audio will
probably be off by even a higher amount. But this should be rare.
This code is for resyncing audio-only streams (e.g. switching between
audio tracks if no video track is active). This must not be run if the
video PTS just isn't known yet. (Although the case in which this changes
anything is probably very obscure, if it can even happen. Still, it's a
bit more correct.)
This is a correction to commit 91a3bda6.
In display-sync mode, the very first video frame is idiotically fully
timed, even though audio has not been synced yet at this point, and the
video frame is more like a "preview" frame. But since it's fully timed,
an underflow is detected if audio takes longer than the display time of
the frame (we send the second frame only after audio is done).
The timing code will try to compensate for the determined desync, but it
really shouldn't. So explicitly discard the timing info in this specific
case. On the other hand, if the first frame still hasn't finished
display, we can pretend everything is ok.
This is a hack - ideally, we either would send a frame without timing
info (and then send it again or so when playback starts properly), or we
would add real pause support to the VO, and pause it during syncing.
Play a trick to make the packet pos field monotonically increasing over
segment boundaries if the source demuxers return monotonically
increasing pos values. This allows the demuxer to uniquely identify
packets with the pos field, and can do refresh seeks using that.
Normally, the packet pos field is used as a fallback for determining the
playback position if the demuxer returns no proper duration. But
demux_timeline.c always will, and the packet pos fields usually make no
sense in relation to the returned file size anyway if the timeline
source demuxers originate from separate streams.
Remove the explicit whitelisting of formats for refresh seeks. Instead,
check whether the packet position is somewhat reliable during demuxing.
If there are packets without position, or the packet position is not
monotonically increasing, then do not use them for refresh seeks.
This does not make sure of some requirements, such as deterministic
seeks. If that happens, mpv will mess up a bit on stream switching.
Also, add another method that uses DTS to identify packets, and prefer
it to the packet position method. Even if there's a demuxer which
randomizes packet positions, it hardly can do that with DTS. The DTS
method is not always available either, though. Some formats do not have
a DTS, and others are not always strictly monotonic (possibly due to
libavformat codec parsing and timestamp determination issues).
If an audio track is enabled during playback, then make it resume at the
exact "current position", instead of playing audio before that position.
This was already done for video.
If the packet read function returns, and EOF was detected, and a seek
was issued in the meantime, then don't use the EOF result. The seek will
be processed later, and reset the EOF state anyway.
The main effect is probably that we don't return EOF to the decoders
(which the playback core resets before issuing the seek), and that we
won't log an EOF message.
Not important, but slightly more correct.
When switching tracks, we normally have the problem that data gets lost
due to readahead buffering. (Which in turn is because we're stubborn and
instruct the demuxers to discard data on unselected streams.) The
demuxer layer has a hack that re-reads discarded buffered data if a
stream is enabled mid-stream, so track switching will seem instant.
A somewhat similar problem is when all tracks of an external files were
disabled - when enabling the first track, we have to seek to the target
position.
Handle these with the same mechanism. Pass the "current time" to the
demuxer's stream switch function, and let the demuxer figure out what to
do. The demuxer will issue a refresh seek (if possible) to update the
new stream, or will issue a "normal" seek if there was no active stream
yet.
One case that changes is when a video/audio stream is enabled on an
external file with only a subtitle stream active, and the demuxer does
not support rrefresh seeks. This is a fuzzy case, because subtitles are
sparse, and the demuxer might have skipped large amounts of data. We
used to seek (and send the subtitle decoder some subtitle packets
twice). This case is sort of obscure and insane, and the fix would be
questionable, so we simply don't care.
Should mostly fix#3392.
Always require them, instead of just for some components which have hard
requirements on correct atomic semantics. They should be widely
available, and are supported by all recent gcc and clang compiler
versions. We even have the fallbacks builtins, which should keep this
working on very old gcc releases.
In particular, w32_common.c recently added a hard requirement on
atomics, but checking this properly in the build system would have been
messy. This commit makes sure it always works.
The fallback where weak atomic semantics are always fine is in theory
rather questionable as well.
Exactly the same situation as with ao_alsa in commit 0b144eac (except
that we can detect the situation better under wasapi).
Essentially, wasapi will allow us to output any sample format, and not
just the one configured by the user in the audio system settings.
VOCTRL_CHECK_EVENTS is called on every frame. This is by design, and is
supposed to check the event queue of the windowing API.
With the decoupled GUI thread in w32_common.c this doesn't make too much
sense, and the purpose of VOCTRL_CHECK_EVENTS is really reduced to
checking event flags. Even worse, waiting on the GUI thread can
interfere with playback, since win32 sometimes blocks the event loop
(e.g. clicking the window title bar).
Change the code such that we really only query the event flags. Use
atomics to avoid having to add a new mutex. (We assume we always have
real atomics available. The build system doesn't check this properly,
and it could fall back to dummy atomics, which are not atomic.)
Should help with #3393. Doesn't help if the core happens to send a
synchronous request, most commonly via VOCTRL_SET_CURSOR_VISIBILITY or
VOCTRL_UPDATE_PLAYBACK_STATE.
This commit adds an --audio-channel=auto-safe mode, and makes it the
default. This mode behaves like "auto" with most AOs, except with
ao_alsa. The intention is to allow multichannel output by default on
sane APIs. ALSA is not sane as in it's so low level that it will e.g.
configure any layout over HDMI, even if the connected A/V receiver does
not support it. The HDMI fuckup is of course not ALSA's fault, but other
audio APIs normally isolate applications from dealing with this and
require the user to globally configure the correct output layout.
This will help with other AOs too. ao_lavc (encoding) is changed to the
new semantics as well, because it used to force stereo (perhaps because
encoding mode is supposed to produce safe files for crap devices?).
Exclusive mode output on Windows might need to be adjusted accordingly,
as it grants the same kind of low level access as ALSA (requires more
research).
In addition to the things mentioned above, the --audio-channels option
is extended to accept a set of channel layouts. This is supposed to be
the correct way to configure mpv ALSA multichannel output. You need to
put a list of channel layouts that your A/V receiver supports.
Pointless anyway. With superficial checking I couldn't find any decoder
which actually outputs this, and AO chmap negotiation would properly
ignore them anyway in most cases.
Assume you use a large value like --audio-delay=20. Then until now the
player would just have seeked normally to a "too late" position, and
played silence for about 20 seconds until audio in the correct time
range is coming again.
Change this by offsetting seeks by the right amount. This works for both
external and muxed files. If a seek isn't precise, then it works only
for external files.
This might cause issues with very large delay options. Hr-seek skipping
could take a lot of time (especially because it affects video too), the
demuxer queue could overflow, and other weird corner cases could appear.
But we just try this on best-effort basis, and if the user uses extreme
values we don't guarantee good behavior.
Prevents segfaults when a fullscreen switch is issued before fully
initializing the VO.
Doesn't change anything since the schedule_resize is only there to
resize in case the image size switches, which happens long after init.
We send a refcounted frame to the encoder, but then disrespect
refcounting rules and write to the frame data without making sure the
buffer is really writeable.
In theory this can lead to reallocation on every frame is the encoder
really keeps a reference. If we really cared, we could fix this by
providing a buffer pool. But then again, we don't care.
This is logical: the function makes sense only in situations where you
are going to write to the audio data. To make it worse,
av_buffer_realloc() also handles this situation, but only if the buffer
size changes (simply because it can't realloc memory in use), so we have
to explicitly force reallocation by unreffing the buffers first.
The problem was that when in fullscreen, switching between images did
not issue a resize event, causing none of the images to be rendered
correctly.
This fixes the problem by issuing a resize event with the screen width
and height.
This commit also moves the zeroing of the events field to when it gets
retrieved by mpv rather than randomly after a resize in the vo/backend
code.
ssurface_handle_configure()'s width and height are just hints given by
the compositor, the application's free to not respect those strictly and
to compensate for e.g. aspect ratio.
This prevents crazy scenarios in which pictures with portrait aspect
ratios have a huge black area to make them 16:9 or whatever the
compositor feels like.
With X11 it was usually left up to the window manager to prevent huge
windows from being out of range, but no Wayland compositor will do
this right now.
Hugely improves usability when using mpv as an image viewer.
Missed during the recent changes.
Also simplify error checking code and check for POLLNVAL
as well (the display fd was never actually checked to be valid).
This requires changing the pixel upload alignment because the odd sizes
might not be aligned to multiples of 4.
Anyway, the restriction has no real benefit and the sizes in between 32
and 64 might be worth using, so just drop it.
Following testing after ebe798a, this is a more than sufficient size to
cover our use case.
The old default was a drop of about 58 dB PSNR using the old code, and
this new default is about 65 dB PSNR, so it's actually an improvement
despite resulting in a smaller size.
There was no outlier whatsoever when comparing sizes around the 64
neighbourhood (with every step corresponding to a PSNR drop of about
0.07 dB), so I picked this since it's a power of two and requires no
change to the current 3dlut-size parsing logic.
I also tested smaller sizes such as 32x32x32 which performed almost as
well on colorful samples, but this results in noticeable black boost in
the dark regions, which is pretty undesirable. Therefore, we should
avoid going much further below 64x64x64.
Either way, this new size is so fast to compute that the 3dlut cache is
almost useless on my end. In fact, it might even be slower to load the
profile from the cache than to recompute it from scratch. (For caches on
a disk. For cache on a tmpfs, it makes no difference)
It seems vo_x11_check_events() was supposed to return the currently
flagged events and reset them. But there are many places where
vo_x11_check_events() is called without checking its return value. This
could lead to forgotten events.
Change the code such that they can't get lost.
This code had the exact same texture indexing bug that the original
scaler code had before the introduction of the LUT_POS macro to fix it.
We can re-use this same macro here, and the performance drop is
virtually entirely negligible. The benefit is greatly improved LUT
accuracy as the 3DLUT size decreases - in particular, the old LUT
started introducing more and more black crush the lower your LUT size is
(because the error was essentially an over-contrast bias, with a
magnitude linearly related to the lut size).
The new code improves black stability as the LUT size decreases, and
only at very low values (16 and below) do black levels start noticeably
getting affected (due to crude linearization of the nonlinear response
curve).
The default value of 3dlut-size is definitely generous enough for this
to make no difference out of the box, but it also causes no performance
drop at all on my machine so I see no harm in improving the logic.
Furthermore, this means we could easily decrease the default 3dlut size
in a future commit, perhaps even down to 64x64x64 as a default. (But
more testing is warranted here)
The FFmpeg API is incredibly weird and inconsistent about this. This is
also a FFmpeg-only issue and nothing like this is in Libav - which
doesn't really show FFmpeg in a very positive light.
(To make it even worse: this is a full-blown Libav API incompatibility,
even though this crap was added for Libav ABI-compatibility. It's
absurd.)
Quoting the FFmpeg header for the AVFrame.channels field:
/**
* number of audio channels, only used for audio.
* Code outside libavutil should access this field using:
* av_frame_get_channels(frame)
* - encoding: unused
* - decoding: Read by user.
*/
int channels;
It says "should" not must, and it doesn't even mention
av_frame_set_channels(). It's also in the section for public fields (not
below a marker that indicates private fields in a public struct, like
it's done e.g. in AVCodecContext).
But not using the accessor will cause silent failures on ABI changes.
The failure that happened due to this code didn't even make it apparent
what was wrong. So just use the idiotic accessor.
Also harmonize the FFmpeg-cursing in the code. (It's fully justified.)
Fixes#3295.
Note that mpv will still check the exact library version numbers, and
reject mismatches - to protect itself from such issues in the future.
Otherwise it behaves dumb. (Although you could argue it shouldn't try to
guess whether speed changes work, but instead simply disable DS if they
don't work.)
It used not to work - but now it apparently does. Not sure when that got
fixed in FFmpeg, but there's no longer a reason to keep this hack.
This also gets rid of the check for the read_seek2 field, which is not
part of the public API.
Both backends have code to close each FD of their wakeup_pipe array.
This array is default-initialized with 0, which means if the backends
exit before the wakeup pipe is created (e.g. when probing), they would
close FD 0.
Initialize the FDs with -1. Then we call close(-1) in these situations,
which is perfectly allowed and has no bad consequences.
This fits natively into the vo/backend and allows to simplify the
polling code.
One new change is the fact that surface_handle_enter flags VO_EVENT_WIN_STATE
and VO_EVENT_RESIZE instead of only VO_EVENT_WIN_STATE. Before this, the code
hackily relied on the timeout and the loop in the wait_frame function to track
and set the scaling factor. Instead, this triggers mpv to run a schedule_resize
and adjust the new VO output dimensions immediately. This is also more accurate
since surface_handle_enter() gets called when a surface is created, moved and
resized, which is exactly what the rest of the player might be interested in.
This uses GLSL mix() instead of going through an indirect texture
access. Easy to implement and might require less resources on some
devices, since the oversample code was already essentially just a
special case of this.
Could be made the new default (as per issue #2685), but that should be
done in a separate commit.
Until now, this has been either handled over vo.event_fd (which should
go away), or by putting event handling on a separate thread. The
backends which do the latter do it for a reason and won't need this, but
X11 and Wayland will, in order to get rid of event_fd.
There's no need to call wl_display_flush() since all the client-side
buffered data has already been flushed prior to polling the fd.
Instead only check for POLLIN and the usual ERR+HUP.
Don't just cause vo_opengl to update the ICC profile every time the
window is moved. Instead, explicitly check if the screen was changed.
Mostly untested.
Some client API users simply don't like such filenames. For their sake,
don't return them, but return a dummy filename instead. (Returning a
latin1-ized version would work too, but is slightly more work.)
Also remove the "\n" from the replacement dummy filename. This was
accidental.
mixer.c didn't really deserve to be separate anymore, as half of its
contents were unnecessary glue code after recent changes. It also
created a weird split between audio.c and af.c due to the fact that
mixer.c could insert audio filters. With the code being in audio.c
directly, together with other code that unserts filters during runtime,
it will be possible to cleanup this code a bit and make it work like the
video filter code.
As part of this change, make the balance code work like the volume code,
and add an option to back the current balance value. Also, since the
balance semantics are unexpected for most users (panning between the
audio channels, instead of just changing the relative volume), and there
are some other volumes, formally deprecate both the old property and the
new option.
Since mixer->ao is always NULL now (it was really just forgotten to be
removed), the uninit call never actually cleared the af field, leaving
a dangling pointer that could be accessed by volume control.
This makes the difference between passing VA_FRAME_PICTURE or
VA_BOTTOM_FIELD for progressive frames (that should be force-
deinterlaced) to VAProcPipelineParameterBuffer.flags. VA-VPP doesn't
really seem to care, and we can get rid of mp_refqueue_is_interlaced()
entirely. It could be argued it's better to pass field flags instead of
the progressive flag.
"Real" frame flag vs. what we pretend it to be. It always used the real
flag, and thus never deinterlaced unflagged frames, even if the
suboption was set to "no".
The hw_subfmt field roughly corresponds to the field
AVHWFramesContext.sw_format in ffmpeg. The ffmpeg one is of the type
AVPixelFormat (instead of the underlying hardware format), so it's a
good idea to switch to this too for preparation.
Now the hw_subfmt field is an mp_imgfmt instead of an opaque/API-
specific number. VDPAU and Direct3D11 already used mp_imgfmt, but
Videotoolbox and VAAPI had to be switched.
One somewhat user-visible change is that the verbose log will now always
show the hw_subfmt as image format, instead of as nonsensical number.
(In the end it would be good if we could switch to AVHWFramesContext
completely, but the upstream API is incomplete and doesn't cover
Direct3D11 and Videotoolbox.)
Old-style commands using _ as separator (e.g. show_progress) were still
used in some places, including documentation and configuration files.
This commit updates all such instances to the new style (show-progress)
so that commands are easier to find in the manual.
Since it turns out that knowing what exactly a file was tagged with can
be useful for debugging purposes, expose this as a property so I can
check it more easily.
This is mostly useful for sig-peak (since nom-peak is currently entirely
calculated by us), but I added both for consistency.
--deinterlace=auto is the default, and has the obscure semantics that
deinterlacing is disabled, unless the user has manually inserted a
deinterlacing filter.
While in software decoding this doesn't matter, and we will happily
insert 2 yadif filters (if the user has already added one), or not
remove the yadif filter (if deinterlacing is disabled, but the user has
added the filter manually), this is different with hardware deinterlacer
filters. These support VFCTRL_SET_DEINTERLACE for toggling deinterlacing
filtering at runtime. It exists mainly for legacy reasons, and possibly
because it makes switching deinterlacing modes more efficient. It might
also gives us an entry-point for VO deinterlacing, maybe. For whatever
reasons this mechanism exists, we still support and use it.
This commit fixes that video.c always used VFCTRL_SET_DEINTERLACE to
disable deinterlacing, even if --deinterlace=auto was set. Fix this by
checking the value of the option directly.
For some reason, the lack of version info was preventing mpv from
appearing in the Default Programs dialog. Re-add it, but don't set the
string version numbers from version.h, because that's what was causing
trouble when the version info was removed. Like the binary version
numbers, these are now hardcoded to 2.0.0.0, which probably doesn't
matter.
The new version info block is also slightly different to the old one. It
fills out all the binary VERSIONINFO fields and makes better use of
macros. It also removes the \000 line terminators from the string
version info, since as far as I can tell, this was just cargo-culting
for an old broken version of the Microsoft resource compiler, and
binutils' windres terminates the strings properly without them.
This should get mpv working on Windows 7 machines without hardware
accelerated graphics adapters. It already worked on Windows 8 and up
because those systems would silently fall back to WARP if there was no
graphics hardware installed.
The normal MPGL_CAP_SW flag is not set, so unlike other opengl backends,
this will choose a software adapter even if opengl:sw is not specified.
The reason for this is, unlike on Linux, where vo_xv and vo_x11 can be
used, mpv on Windows does not have any VO to fall back on when hardware
acceleration isn't available, so if software adapters are rejected, the
user won't see any video output when using the default settings. WARP
seems to perform quite well, so it should be used in this case.
The cache reader thread actually unlocks the mutex protecting the
underlying stream while reading from it. That's why other code goes out
of its way to run certain stream operations on the cache thread. Do the
same.
We could have this simpler by creating a mechanism that would "park" the
cache thread and make it wait for the lock (while we have it) in order
to gain exclusive access. This could be done in the future.
Eagerly execute seeks to the underlying stream in the cache seek
entrypoint itself. While asynchronous execution is a goal of the cache,
it doesn't matter too much for seeks. They always were executed within
the lock, so the reader was blocked anyway. It's not necessary to ensure
async. execution here either, because seeks are relatively rare, and the
demuxer can just stay blocked for a while.
Fixes: mpv http://samples.mplayerhq.hu/V-codecs/DIV5/ayaneshk-test.avi
This code was supposed to adjust existing conversion filters (to make
them output a different format). But the code was just broken,
apparently a refactoring accident. It accessed af instead of af->prev.
The bug tended to add new conversion filters, even if an existing one
could have been used. (Can be tested by inserting a dummy lavrresample
filter followed by a format filter which forces conversion.)
In addition, it's probably better to return the actual error code if
reinitializing the filter fails. It would then respect an AF_FALSE
return value, which means format negotiation failed, instead of a
generic error.
af_volume has a volumedb sub-option, which allows the user to set an
explicit volume. Until recently, the player read back this value and
used it as initial softvol volume. But now it just overwrites it.
Instead of overwriting it, multiply the different gain values. Above
all, this will do the right thing if only softvol is used, or if the
user only adds the af_volume filter manually.
Normally, you want downmixing to happen first thing in the filter chain.
This is reflected in codec downmixing, which feeds the filter chain
downmixed audio in the first place. Doing this has the advantage of
needing less data to process. But the main motivation is that if there
is a drc filter in the chain, you want to process it the downmixed
audio.
Add an idiotic heuristic to achieve this. It tries to detect whether the
audio was indeed automatically downmixed (or upmixed). To detect what
the output format is going to be, it builds the filter chain normally,
and then retries with the heuristic applied (and for extra paranoia,
retries without the heuristic again if it fails to successfully rebuild
the filter chain for unknown reasons). This is simple and will work in
almost all cases.
Doing it in a more complete way is rather hard, because filters are so
generic. For example, we know absolutely nothing about the behavior of
af_lavfi, which creates an opaque filter graph with libavfilter. We
don't know why a filter would e.g. change the channel layout on its
output. (Our heuristic bails out in this case.) We're also slave to the
lowest common denominator of how our format negotiation works, and how
libavfilter's works.
In theory, we could make this mechanism explicit by introducing a
special dummy filter. The filter chain would then try to convert between
input and output formats at the dummy filter, which would give the user
more control over how downmix happens. On the other hand, the user could
just insert explicit conversion filters instead, so this would probably
have questionable value.
The algorithm and functionality is the same, but the code becomes much
simpler and easier to follow.
The assumption that there is only 1 conversion filter (lavrresample)
helps with the simplification, but the main change is to use the same
code for format/channels/rate. Get rid of the different AF_CONTROL_SET_*
controls, and change the af->data parameters directly. (af->data is
badly named, but essentially is a placeholder for the output format.)
Also, instead of trying to use the af_reinit() loop to init inserted
conversion filters or filters with changed output formats, do it inline,
and move the common code to a filter_reinit() function. This gets rid of
the awful retry variable.
In general, this should not change any runtime behavior.
This happens to be better for the af_volume filter (for softvol), and
saves some code too. It's "better" because you want to affect the
final filtered audio, such as after a manually inserted drc filter.
Drop the code for switching the volume options and properties between
af_volume and AO volume controls. interface-changes.rst mentions the
changes in detail.
Do this because this was exceedingly complex and had other problems as
well. It was also very hard to test. It's just not worth the trouble.
Some leftovers like AOCONTROL_HAS_PER_APP_VOLUME will be removed at a
later point.
Fixes#3322.
I made this call up because I sort of thought this makes senssssse. I'm
now convinced that it does not, and even is actively harmful. I'm still
quite in the dark how the DirectD 11 video API is supposed to work with
synchronization, but at least for normal graphics this call would not
make much sense.
When selecting a device that simply doesn't exist with --audio-device,
AudioUnit will still initialize and start playback without complaining.
But it will never call the audio render callback, which leads to audio
playback simply not progressing.
I couldn't find a way to get AudioUnit to report an error at all, so
here's a crappy hack that takes care of this in most cases. We assume
that all devices have a kAudioDevicePropertyDeviceIsAlive property.
Invalid devices will error when querying the property (with 'obj!' as
status code).
This is not the correct fix, because we try to double-guess AudioUnit's
behavior by accessing a lower label API. Suggestions welcome.
Although it appears to be accepted by the function, MSGL_STATUS messages
are never passed to the client API. Consequently "status" has the same
meaning as "v" and is useless.
Commit 771a8bf5 added code to avoid unnecessary vf_reconfig() calls for
unrelated reasons, but forget to consider that it has to be called at
least once if the input format changes. As a consequence it got "stuck"
due to not being able to decode more frames.
vo_frame can have more than 1 frame - the extra frames are future
references, which are sometimes useful for filtering (vo_opengl
interpolation). There's no harm in reducing the number of frames sent to
the VO requested amount of future frames, so do that.
Doesn't actually reduce the number of concurrently in use frames in
practice.
If the status line is wider than the reported terminal size, then cut it
off instead of causing the terminal to scroll down for the next line.
This is done in the most primitive way possible, assuming ASCII.
This was actually done in the past as far as I'm aware; do it again.
(Probably differently.)
Some filters support VFCTRL_SET_DEINTERLACE. This affects most hardware
deinterlace filters. They can be inserted by the user manually, or auto-
inserted by vf.c itself as conversion filter (vf_d3d11vpp). In these
cases, we shouldn't insert or remove filters outselves, and instead
VFCTRL_SET_DEINTERLACE should be invoked to switch the mode.
This wasn't done correctly in the recently refactored code and could
have broken with --deinterlace. (The refactor only considered switching
via property in this case.) Fix it by making it a proper part of the
filter_reconfig() function, and making set_deinterlacing() (which is
called by the property handler) merely call filter_reconfig() in all
cases to do the real work.
We can even avoid rebuilding the filter chain - though only if no other
auto-filters are inserted. It probably also provides a slightly cleaner
way to implement functionality in the VO while still inserting video
filter fallbacks correctly if required.
The test scenario at hand was hardware decoding a file with d3d11 and
with deinterlacing enabled. The file switches to a non-hardware
dedocdeable format mid-stream. This failed, because it tried to call
vf_reconfig() with the old filters inserted, with was fatal due to
vf_d3d11vpp accepting only hardware input formats.
Fix this by always strictly removing all auto-inserted filters
(including the deinterlacing one), and reconfiguring only after that.
Note that this change is good for other situations too, because we
generally don't want to use a hardware deinterlacer for software
decoding by default. They're not necessarily optimal, and VAAPI VPP even
has incomprehensible deinterlacer bugs specifically with software frames
not coming from a hardware decoder.
Instead of using the "vf" command code (which changes filters at runtime
on user input), use the general filter-insertion code. The latter was
added later, and is more suitable for automatically inserted filters.
The old code failed in particular when using watch-later saving, which
stored the filter list in the resume config file. If a user changed the
hardware decoding mode via command line, the stored filter chain was out
of date and could cause failure due to not working with hardware or
software decoding mode. Storing the deinterlace filter in the filter
list was unavoidable, because it was part of the user state. (The new
code only edits the actually instantiated filters.)
These mostly happen in situations where the correct behavior is
relatively new and not found in the wild (therefore not worth
implementing) and/or extremely complicated (and thus not worth worrying
about the potential edge cases and UI changes).
Still, it's best to document these where they happen to guide the poor
souls maintaining these files in the future.
This uses eglPostSubBufferNV to trigger ANGLE to check the window size
and update the size of the swapchain to match, which is recommended
here: https://groups.google.com/d/msg/angleproject/RvyVkjRCQGU/gfKfT64IAgAJ
With the D3D11 backend, using eglPostSubBufferNV with a 0-sized update
region will even skip the Present() call, meaning it won't block for a
vsync period. Hopefully ANGLE will have a less hacky way of doing this
in future. See the relevant ANGLE issue: http://anglebug.com/1438Fixes#3301
This can for example happen with vo_opengl_cb, if it is used with a GL
implementation that does not supports FBOs. (mpv itself should never
attempt to use FBOs if they're not available.)
Without this check it would trigger an assert() in our dummy
glBindFramebuffer wrapper.
Suspected cause of #3308, although it's still unlikely.
For example it should be set to IMGFMT_D3D11NV12 if it isn't already.
Otherwise, an assertion in vf.c could trigger.
This probably couldn't be provoked yet.
This moves some of the bulky user-shader specific logic into the file
dedicated to it. Rather than expose video.c state, variable lookup is
now done via a simulated closure.
This greatly improves the result when decoding typical (ST.2084) HDR
content, since the job of tone mapping gets significantly easier when
you're only mapping from 1000 to 250, rather than 10000 to 250.
The difference is so drastic that we can now even reasonably use
`hdr-tone-mapping=linear` and get a very perceptually uniform result
that is only slightly darker than normal. (To compensate for the extra
dynamic range)
Due to weird implementation details, this only seems to be present on
keyframes (or something like that), so we have to cache the last seen
value for the frames in between.
Also, in some files the metadata is just completely broken /
nonsensical, so I decided to apply a simple heuristic to detect
completely broken metadata.
This involves multiple changes:
1. Brightness metadata is split into nominal peak and signal peak.
For a quick and dirty explanation: nominal peak is the brightest value
that your color space can represent (i.e. the brightness of an encoded
1.0), and signal peak is the brightest value that actually occurs in
the video (i.e. the brightest thing that's displayed).
2. vo_opengl uses a new decision logic to figure out the right nom_peak
and sig_peak for all situations. It also does a better job of picking
the right target gamut/colorspace to use for the OSD. (Which still is
and still should be treated as sRGB). This change in logic also
fixes#3293 en passant.
3. Since it was growing rapidly, the logic for auto-guessing / inferring
the right colorimetry configuration (in pass_colormanage) was split from
the logic for actually performing the adaptation (now pass_color_map).
Right now, the new logic doesn't do a whole lot since HDR metadata is
still ignored (but not for long).
This has two reasons:
1. I tend to add new fields to this metadata, and every time I've done
so I've consistently forgotten to update all of the dozens of places in
which this colorimetry metadata might end up getting used. While most
usages don't really care about most of the metadata, sometimes the
intend was simply to “copy” the colorimetry metadata from one struct to
another. With this being inside a substruct, those lines of code can now
simply read a.color = b.color without having to care about added or
removed fields.
2. It makes the type definitions nicer for upcoming refactors.
In going through all of the usages, I also expanded a few where I felt
that omitting the “young” fields was a bug.
Commit 883d3114 seems to have (accidentally?) dropped the FBOTEX_FUZZY
from the output_fbo resize, which means that current master will keep
resizing and resizing the FBO as you change the window size, introducing
severe memory leaking after a while. (Not sure why that would cause
memory leaks, but I blame nvidia)
Either way, it's bad for performance too, so it's worth fixing.
vo_vaapi is the only thing which can't scale RGBA on the GPU. (Other
cases of RGBA scaling are handled in draw_bmp.c for some reason.)
Move this code and get rid of the osd_conv_cache thing.
Functionally, nothing changes.
This affects VOs (or other code which render OSD) which does not support
the LIBASS format, but only RGBA. Instead of having a converter stage in
osd.c, make mp_ass_packer_pack() output directly in RGBA.
In general, this is work towards refcounted subtitle images.
Although we could keep the "converter" design, doing it this way seems
simpler, at least considering the current situation with only 2 OSD
formats. It also prevents copying & packing the data twice, which will
lead to better performance. (Although I guess this case is not important
at all.)
It also fixes --force-rgba-osd-rendering when used with vo_opengl,
vo_vdpau, and vo_direct3d.
The intention is to let mp_ass_packer_pack() produce different output
for the RGBA and LIBASS formats. VOs (or whatever generates the OSD)
currently do not signal a preferred format, and this mechanism just
exists to switch between RGBA and LIBASS formats correctly, preferring
LIBASS if the VO supports it.
This is how PBOs are normally supposed to be used.
Unfortunately I can't see an any absolute improvement on nVidia binary
drivers and playing 4K material. Compared to the "old" PBO path with 1
buffer, the measured GL time decreases significantly, though.
GL generally does not support flipping the image on upload, meaning
negative strides are not supported. vo_opengl handles this by flipping
rendering if the stride is inverted, and gl_pbo_upload() "ignores"
negative strides by uploading without flipping the image.
If individual planes had strides with different signs, this broke. The
flipping affected the entire image, and only the sign of the first plane
was respected.
This is just a crazy corner case that will never happen, but it turns
out this is quite simple to support, and actually improves the code
somewhat.
This introduces a gl_pbo_upload_tex() function, which works almost like
our gl_upload_tex() glTexSubImage2D() wrapper, except it takes a struct
which caches the PBO handles. It also takes the full texture size (to
make allocating an ideal buffer size easier), and a parameter to disable
PBOs (so that the caller doesn't have to duplicate the gl_upload_tex()
call if PBOs are disabled or unavailable).
This also removes warnings and fallbacks on PBO failure. We just
silently try using PBOs on every frame, and if that fails at some point,
revert to normal texture uploads. Probably doesn't matter.
This makes the geometry of the sizing borders more like the ones in
Windows 10. It also fixes an off-by-one error that made the right and
bottom borders thinner than the left and top borders, which made it
difficult to resize the window when using the Windows 7 classic theme
(because it has pretty thin sizing borders to begin with.)
We don't support this anymore.
This tries to exit in a controlled way after command line options are
applied in order to honor logging options and, in case of libmpv, not to
kill the host. Not sure if it would be better to just vomit text to
stderr and call abort().
Working towards refcounted sub images, and also for removing bitmap
packers from VOs.
I'm not sure why we even have this overlay-add command. It was sort of
"needed" before opengl-cb was introduced, and before Lua scripts could
put ASS drawings on OSD without conflicting with the OSC. But now trying
to use it doesn't make too much sense anymore.
Still keep it because we're trying to be nice, but throw performance out
of the window. Now image data is copied 2 more times before displaying
it. This also makes using the command a bit simpler.
The libavcodec wmapro decoder will skip some bytes at the start of the
first packet and return each time. It will not return any audio data in
this state.
Our own code as well as libavcodec's new API handling
(avcodec_send_packet() etc.) discard the PTS on the first return, which
means the PTS is never known for the first packet. This results in a
"Failed audio resync." message.
Fixy it by remember the PTS in next_pts. This field is used only if the
decoder outputs no PTS, and is updated after each frame - and thus
should be safe to set.
(Possibly this should be fixed in libavcodec new API handling by not
setting the PTS to NOPTS as long as no real data has been output. It
could even interpolate the PTS if the timebase is known.)
Fixes the failure message seen in #3297.
Change all producer of libass images to packing the bitmaps into a
single larger bitmap directly when they're output. This is supposed to
help working towards refcounted sub bitmaps.
This will reduce performance for VOs like vo_xv, but not for vo_opengl.
vo_opengl simply will pick up the pre-packed sub bitmaps, and skip
packing them again. vo_xv will copy and pack the sub bitmaps
unnecessarily - but if we want sub bitmap refcounting, they'd have to be
copied anyway.
The packing code cannot be removed yet from vo_opengl, because there are
certain corner cases that still produce unpackad other sub bitmaps.
Actual refcounting will also require more work.
We have a warning mechanism for removed and for replaced options, but
none yet for options which have been simply deprecated.
For the following commit.
(Fun fact: just adding the m_option field increases binary size by
14KB.)
No method of taking a screenshot was implemented at all. vo_opengl
lacked window screenshotting, because ANGLE doesn't allow reading the
frontbuffer. There was no way to read back from a D3D11 texture either.
Implement reading image data from D3D11 textures. This is a low-quality
effort to get basic screenshots done. Eventually there will be a better
implementation: once we use AVHWFramesContext natively, the readback
implementation will be in libavcodec, and will be able to cache the
staging texture correctly. Hopefully. (For now it doesn't even have a
AVHWFramesContext for D3D11 yet. But the abstraction is more appropriate
for this purpose.)
OK, this was dumb. The file didn't have much to do with ANGLE, and the
functionality can simply be moved to d3d.c. That file contains helpers
for decoding, but can always be present (on Windows) since it doesn't
access any D3D specific libavcodec APIs. Thus it doesn't need to be
conditionally built like the actual hwaccel wrappers.
Instead of hard-coding a big list, move some of the functionality
to csputils. Affects both the auto-guess blacklist and the peak
estimation.
Also update the comments.
Too many "exceptions" these days, it's easier to just hard-code a
whitelist instead of a blacklist. And besides, it only really makes
sense to avoid adaptation for BT.601 specifically, since that's the one
we auto-guess based on the resolution.
I'm not even sure why we ever consulted *_src to begin with, since that
just describes the current image format - and not the original metadata.
(And in fact, we specifically had logic to work around the impliciations
this had on linear scaling)
image_params is *the* authoritative source on the intended (i.e.
reference) image metadata, whereas *_src may be changed by previous
passes already. So only consult image_params for picking auto-generated
values.
Also, add some more missing "wide gamut" and "non-gamma" curves to the
autoconfig blacklist. (Maybe it would make sense to move this list to
csputils in the future? Or perhaps even auto-detect it based on the
associated primaries)
User request and not that hard. Closes#3157.
Note that FFmpeg doesn't support this and there's no signalling in HEVC
etc., so the only way users can access it is by using vf_format
manually.
Mind: This encoding uses full range values, not TV range.
This is actually not entirely trivial since it involves negative Yxy
coordinates, so the CMM has to be capable of full floating point
operation. Fortunately, LittleCMS is, so we can just blindly implement
it.
Most devices seems to require special signalling (e.g. via HDMI
metadata) to actually decode HDR signals and treat them as such, so it's
probably worth warning the potential user about the fact that mpv pretty
definitely does *not* set any of this metadata signalling.
This HDR function is unique in that it's still display-referred, it just
allows for values above the reference peak (super-highlights). The
official standard doesn't actually document this very well, but the
nominal peak turns out to be exactly 12.0 - so we normalize to this
value internally in mpv. (This lets us preserve the property that the
textures are encoded in the range [0,1], preventing clipping and making
the best use of an integer texture's range)
This was grouped together with SMPTE ST2084 when checking libavutil
compatibility since they were added in the same release window, in a
similar timeframe.
Some bugs in this code are exposed by e.g. playing lossless audio files
with --ad-lavc-threads=16. (libavcodec doesn't really support threaded
audio decoding, except for lossless files.) In these cases, a major
amount of audio can be buffered, which makes incorrect handling of this
buffering obvious.
For one, draining the decoder can take a while, so if there's a new
segment, we shouldn't read audio.
The segment end check was completely wrong, and used the start value.
The main framebuffer is not the default framebuffer for the dxinterop
backend. Bind the main framebuffer and use the appropriate attachment
when reading the window content.
Fix#3284
The size check introduced in commit d941a57b did not consider that Xv
can round up the image size to the next chroma boundary. Doing that
makes sense, so it can't certainly be considered server misbehavior.
Do 2 things against this: allow if the server returns a larger image (we
just crop it then), and also allocate a properly aligned image in the
first place.
Instead of hardcoding what the libavcodec ac3 encoder expects, configure
it based on the AVCodec fields.
Unfortunately, it doesn't export the list of sample rates, so that is
done manually. This commit actually fixes the rate always to 48Khz. I
don't even know whether the other rates worked. (Possibly did, but
they'd still change the spdif parameters, and would work differently
from ad_spdif.c.)
Workaround for an awful corner-case. The new decode API "locks" the
decoder into the EOF state once a drain packet has been sent. The
problem starts with a file containing a 0-sized packet, which is
interpreted as drain packet.
This should probably be changed in libavcodec (not treating 0-sized
packets as drain packets with the new API) or in libavformat (discard
0-sized packets as invalid), but efforts to do so have been fruitless.
Note that vd_lavc.c already does something similar, but originally for
other reasons.
Fixes#3106.
See previous commit. (The mixing case was never supported, so this has
equivalent functionality.)
This also implicitly fixes a bug: the old code allocated image data for
the cropped surface size only, which means the
video_surface_get_bits_y_cb_cr call could write beyond the allocated
image memory.
For now, this affects taking screenshots only.
If the video mixer is actually needed, we want to go through the video
mixer for screenshots too - which is why this function simply always
used the video mixer, and then copied the surface to RAM as RGBA.
Add reading the surface as nv12 if possible. If the format is correct,
and no special video processing (like deinterlacing) is used, then it's
read directly without going through the video mixer.
There's no particular reason for doing this, other than avoiding the
video mixer (like vo_opengl can do it now). Also, the next commit will
make use of this in vf_vdpaurb.c.
The GL_ARB_timer_query extension and thus the GL_TIME_ELAPSED constant
don't exist for GLES.
For ES the EXT_disjoint_timer_query is used so take the constant from
that else provide the constant manually.
See pr #3216 which introduced this error.
Until now, we've always converted vdpau video surfaces to RGB, and then
mapped the resulting RGB texture. Change this so that the surface is
mapped as NV12 plane textures.
The reason this wasn't done until now is because vdpau surfaces are
mapped in an "interlaced" way as separate fields, even for progressive
video. This requires messy reinterleraving. It turns out that even
though it's an extra processing step, the result can be faster than
going through the video mixer for RGB conversion.
Other than some potential speed-gain, doing this has multiple other
advantages. We can apply our own color conversion, which is important in
more complex cases. We can correctly apply debanding and potentially
other processing that requires chroma-specific or in-YUV handling.
If deinterlacing is enabled, this switches back to the old RGB
conversion method. Until we have at least a primitive deinterlacer in
vo_opengl, this will stay this way. The d3d11 and vaapi code paths are
similar. (Of course these don't require any crazy field reinterleaving.)
Otherwise stale references will survive forever. Could leak hardware
video surfaces.
In particular, the mpv vdpau code crashed with an assertion when exiting
after toggling deinterlacing, because not all references were released.
1. this basically reverts commit de4c74e5a4.
even with CVDisplayLinkCreateWithActiveCGDisplays and
CVDisplayLinkSetCurrentCGDisplayFromOpenGLContext we still have to
explicitly set the current display ID, otherwise it will just always
choose the display with the lowest refresh rate. another weird thing is,
we still have to set the display ID another time with
CVDisplayLinkSetCurrentCGDisplay after the link was started. otherwise
the display period is 0 and the fallback will be used.
if we ever use the callback method for something useful it's probably
better to use CVDisplayLinkCreateWithActiveCGDisplays since we will need
to keep the display link around instead of releasing it at the end.
in that case we have to call CVDisplayLinkSetCurrentCGDisplay two times,
once before and once after LinkStart.
2. add windowDidChangeScreen delegate to update the display refresh rate
when mpv is moved to a different screen.
We have two problems here.
1. CVDisplayLinkGetActualOutputVideoRefreshPeriod, like the name suggests,
returns a frame period and not a refresh rate. using this as screen_fps
just leads to a slideshow. why didn't this break video playback on OS X
completely? the answer to this leads us to the second problem.
2. it seems that CVDisplayLinkGetActualOutputVideoRefreshPeriod always
returns 0 if used without CVDisplayLinkSetOutputCallback and hence always
fell back to CVDisplayLinkGetNominalOutputVideoRefreshPeriod. adding a
callback to CVDisplayLink solves this problem. the callback function at
this moment doesn't do anything but could possibly used in the future.
This can also remove all the stuff for lazily attaching the texture. It
doesn't matter if the dxinterop backend changes the bound framebuffer
during a VOCTRL, since the renderer does not rely on the GL state being
preserved.
Since there are not many sub-rectangles, this doesn't cost too much. On
the other hand, it avoids frequent warnings with vo_xv.
Also, the second copy in mp_blur_rgba_sub_bitmap() can be dropped.
The previous few commits changed sd_lavc.c's output to packed RGB sub-
images. In particular, this means all sub-bitmaps are part of a larger,
single bitmap. Change the vo_opengl OSD code such that it can make use
of this, and upload the pre-packed image, instead of packing and copying
them again.
This complicates the upload code a bit (4 code paths due to messy PBO
handling). The plan is to make sub-bitmaps always packed, but some more
work is required to reach this point. The plan is to pack libass images
as well. Since this implies a copy, this will make it easy to refcount
the result.
(This is all targeted towards vo_opengl. Other VOs, vo_xv, vo_x11, and
vo_wayland in particular, will become less efficient. Although at least
vo_vdpau and vo_direct3d could be switched to the new method as well.)
The sub-bitmaps get extended by --sub-gauss, so we have to compute the
bounding box on the original subs. Not sure if this is really
eqauivalent to what the code did before, and I don't have the sample
anymore. (But this approach sure is a _shitty_ hack.)
Implement it directly in sd_lavc.c as well. Blurring requires extending
the size of the sub-images by the blur radius. Since we now want
sub_bitmaps to be packed into a single image, and we don't want to
repack for blurring, we add some extra padding to each sub-bitmap in the
initial packing, and then extend their size later. This relies on the
previous bitmap_packer commit, which always adds the padding in all
cases.
Since blurring is now done on parts of a large bitmap, the data pointers
can become unaligned, depending on their position. To avoid shitty
libswscale printing a dumb warning, allocate an extra image, so that the
blurring pass is done on two newly allocated images. (I don't find this
feature important enough to waste more time on it.)
The previous refactor accidentally broke this feature due to a logic bug
in osd.c. It didn't matter before it happened to break, and doesn't
matter now since the code paths are different.
Until now, subtitle renderers could export SUBBITMAP_INDEXED, which is a
8 bit per pixel with palette format. sd_lavc.c was the only renderer
doing this, and the result was converted to RGBA in every use-case
(except maybe when the subtitles were hidden.)
Change it so that sd_lavc.c converts to RGBA on its own. This simplifies
everything a bit, and the palette handling can be removed from the
common code.
This is also preparation for making subtitle images refcounted. The
"caching" in img_convert.c is a PITA in this respect, and needs to be
redone. So getting rid of some img_convert.c code is a positive side-
effect. Also related to refcounted subtitles is packing them into a
single mp_image. Fewer objects to refcount is easier, and for the libass
format the same will be done. The plan is to remove manual packing from
the VOs which need single images entirely.
Apply the padding internally to each input bitmap, instead of requiring
this for the semi-public API.
Right now, everything still uses packer_pack_from_subbitmaps() to fill
the input bitmap sizes, but that's going to change with the following
commit. Since bitmap_packer.in is mutated during packing anyway, it's
more convenient to add the padding automatically.
Also, guarantee that every sub-bitmap has a padding border around it.
Don't let the padding overlap. Add padding even on the containing
borders.
This is simpler, doesn't cost much in memory usage, and is convenient
for one of the following commits.
I do not understand why github requires adding this crap to source code
repositories themselves, instead of making them part of the repository
configuration. Remove CONTRIBUTING.md to compensate for github crap
accumulating.
This is the ES equivalent to ARB_timer_query. It enables the performance
timers on ANGLE. All the added functions should be identical in
semantics to their desktop GL equivalents.
The OpenGL 3.0+ and ES specs are quite clear on what values are
accepted for the attachment object name parameter. And there's no
overlap for the default framebuffer. Sigh.
Probably fixes Mesa raising an error in this case and might fix#3251.
Regression by the previous vo_opengl change.
Until now, we've used system-specific API (GLX, EGL, etc.) to retrieve
the depth of the default framebuffer. (We equal this to display depth
and use the determined depth for dithering.)
We can actually retrieve this value through standard GL API, and it
works everywhere (except GLES 2 of course). This simplifies everything a
great deal.
egl_helpers.c is empty now. But I expect that some EGL boilerplate will
be moved to it, so don't remove it yet.
This was somehow done under the assumption that ANGLE would somehow
always use RG in ES2 mode. But there's no basis for this. Even if ANGLE
supports NV12 textures with drivers that do not allow for texture_rg,
this cas eis too obscure to worry about. So do the robust and correct
thing instead, and disable this code if texture_rg is not available.
Commit 74e3d11 resulted in the background overlay not getting destroyed
when mpv quits. Add back a piece of code that was removed in that commit
to restore correct functionality.
Fixes issue #3100
This helps with shitty APIs and even shittier drivers (I'm looking at
you, ALSA). Sometimes they won't send proper wakeups. This can be fine
during playback, when for example playing video, because mpv still will
wakeup the AO outside of its own wakeup mechanisms when sending new data
to it. But when draining, it entirely relies on the driver's wakeup
mechanism. So when the driver wakeup mechanism didn't work, it could
hard freeze while waiting for the audio thread to play the rest of the
data.
Avoid this by waiting for an upper bound. We set this upper bound at the
total mpv audio buffer size plus 1 second. We don't use the get_delay
value, because the audio API could return crap for it, and we're being
paranoid here. I couldn't confirm whether this works correctly, because
my driver issue fixed itself.
(In the case that happened to me, the driver somehow stopped getting
interrupts. aplay froze instead of playing audio, and playing audio-only
files resulted in a chop party. Video worked, for reasons mentioned
above, but drainign froze hard. The driver problem was solved when
closing all audio output streams in the system. Might have been a dmix
related problem too.)
When we're draining, don't wakeup the core on every buffer fill, since
unlike during normal playback, we won't actually get more data. The
wakeup here conceptually works like wakeups with condition variables, so
redundant wakeups do not hurt, so this is just a minor change and
nothing of consequence.
(Final EOF also requires waking up the core, but there is separate code
to send this notification.)
Also dump the p->still_playing field in trace logging.
Of course we can't just skip updating the OSD if the playloop was woken
up for the purpose of removing OSD after an OSD timer expired.
Fixes e.g. OSD bars sometimes sticking along when seeking while paused.
Normally, OSD is updated every time the playloop is run. This has to be
done, because the OSD may implicitly reference various properties,
without knowing whether they really need to be updated or not. (There's
a property update mechanism, but it's mostly unavailable, because OSD is
special-cased and can not use the client API mechanism properly.)
Normally, these updates are no problem, because the OSD is only actually
printed when the OSD text actually changes.
But commit d23ffd24 added a rate-limiting mechanism, which tries to
limit OSD updates at most every 50ms (or the next video frame). Since it
can't know in advance whether the OSD is going to change or not, this
simply waked up the player every 50ms.
Change this so that the player is updated only as part of general
updates determined through mp_notify(). (This function also notifies the
client API of changed properties.) The desired result is that the player
will not wake up at all in normal idle mode, but still update properties
that can change when paused, such as the cache.
This is mostly a cosmetic change (in the sense of making runtime
behavior just slightly better). It has the slightly more negative
consequence that properties which update implicitly (such as "clock")
will not update periodically anymore.
This is a common idiom used in MSDN docs and Raymond Chen's example
programs to get a HINSTANCE for the current module, regardless of
whether it's an .exe or a .dll. Using GetModuleHandle(NULL) for this is
technically incorrect, since it always gets a handle to the .exe, even
when the executing code (in libmpv) is running in a .dll. In this case,
using the wrong HINSTANCE could cause namespace issues with window
classes, since CreateWindowEx uses the HINSTANCE to search for the
matching window class name.
See:
https://blogs.msdn.microsoft.com/oldnewthing/20050418-59/?p=35873https://blogs.msdn.microsoft.com/oldnewthing/20041025-00/?p=37483
There were two mistakes:
- SDL's RGB565 is always equivalent to ffmpeg's RGB565 (both are packed 16bit
native-endian integers in RGB=565 form) - this was wrongly reversed on
big endian platforms.
- SDL's RGB888 doesn't actually mean RGB24, but XRGB8888 (i.e. 32bit
packed integer, top 8 bits unused).
- Use RGB0 not RGBA when there is no alpha.
This is mainly for the nnedi3 user shader. With all whose NN weights
hardcoded into the shader source code, the shader file could be as
large as 300 kB.
This reverts commit b51957fab5.
Breaks big time. It appears to ignore explicitly configured paths within
the libav* .pc files, which for example breaks mpv-build.
Although D3D11 video decoding is unuspported on Windows 7, the
associated APIs almost work. Where they fail is texture creation, where
we try to create D3D11_BIND_DECODER surfaces. So specifically try to
detect this situation.
One issue is that once the hwdec interop is created, the damage is done,
and it can't use another backend (because currently only 1 hwdec backend
is supported). So that's where we prevent attempts to use it.
It still can fail when trying to use d3d11va-copy (since that doesn't
require an interop backend), but at that point we don't care anymore -
dxva2(-copy) is tried before that anyway.
Distros and users alike should be made aware of the fact that old
FFmpeg versions are bad. When users come to us with FFmpeg-related
trouble, the answer is “update FFmpeg” more often than not
(and no further support will be provided until they have done so),
so instead we just nag them about it here.
Instead of having 9 different properties, requiring 18 different
VOCTRLs to read them all, they are now exposed as a single property.
This is not only cleaner (since they're all together) but also allows
querying all 9 of them with only a single VOCTRL (by using
mp.get_property_native).
(The extra factor of 2 was due to an extra query being needed to get the
type, which is now also unnecessary)
This makes it much easier to access performance metrics from within a
lua script, and also makes it easier to just show a readable, formatted
version via show-text.
User hooks can now use an extra WHEN expression to specify when the
shader should be run. For example, this can be used to only run a chroma
scaling shader `WHEN CHROMA.w LUMA.w <`.
There's a slight semantics change to user shaders: When trying to bind a
texture that does not exist, a shader will now be silently skipped
(similar to when the condition is false) instead of generating an error.
This allows shader stages to depend on an optional earlier stage without
having to copy/paste the same condition everywhere.
(In other words: there's an implicit condition on all of the bound
textures existing)
When using --hwdec=auto, systems that don't provide
D3D11_CREATE_DEVICE_VIDEO_SUPPORT, which probably includes all Windows
Vista and 7 systems, will print an error message. Reduce the log level
to verbose when probing and skip the error message entirely if d3d11.dll
is not present.
This commit is in a similar spirit to 991af7d.
This comes up often, see e.g. #3220. The issue is that if the stream
input is not seekable, the demuxer is marked as not seekable. But if the
stream cache is enabled, the file still _might_ be seekable to a degree.
We recently disabled seeking in this mode because it can cause very
weird issues, mostly because if stream-layer seeking fails, the demuxers
will arbitrarily misbehave. On the other hand, it can work if the seek
is within the cached range, which is why the user can still enable it
with --force-seeking. There is a weird trade-off between allowing this
and not crapping up too easily, so just informing the user about the
possibility seems best.
Since the libavformat API is crap, we have to apply tons of heuristics
to check whether seeking will work. (No, checking it at seek time isn't
going to work either, because if a seek fails, the demuxer will be in an
undefined state. Because the libavformat API is crap.)
For clang, it's enough to just put (void) around usages we are
intentionally ignoring the result of.
Since GCC does not seem to want to respect this decision, we are forced
to disable the warning globally.
The default behavior of vo_opengl has pretty much always been 'show the
source colors as-is, without caring to adapt it to the target device'.
This decision is mostly based on the fact that if we do anything else,
lots of people will complain.
With the rise of content like BT.2020, however, it turns out more people
complain about this content being very desaturated than people complain
about this content not matching VLC - so let's just map ultra-wide gamut
content back down to standard gamut by default.
Instead of measuring the actual upload time, this instead measures the
time needed to render + map the texture via vdpau. These numbers are
still useful, since they're part of the critical path.
This is plumbed through a new VOCTRL, VOCTRL_PERFORMANCE_DATA, and
exposed as properties render-time-last, render-time-avg etc.
All of these numbers are in microseconds, which gives a good precision
range when just outputting them via show-text. (Lua scripts can
obviously still do their own formatting etc.)
Signed-off-by: wm4 <wm4@nowhere>
To avoid blocking the CPU, we use 8 time objects and rotate through
them, only blocking until the last possible moment (before we need
access to them on the next iteration through the ring buffer). I tested
it out on my machine and 4 query objects were enough to guarantee
block-free querying, but the extra margin shouldn't hurt.
Frame render times are just output at the end of each frame, via MP_DBG.
This might be improved in the future. (In particular, I want to expose
these numbers as properties so that users get some more visible feedback
about render times)
Currently, we measure pass_render_frame and pass_draw_to_screen
separately because the former might be called multiple times due to
interpolation. Doing it this way gives more faithful numbers. Same goes
for frame upload times.
When ANGLE is using D3D11 and not running in DirectComposition mode,
DXGI will hook the video window's message loop and override Alt+Enter to
trigger a transition to exclusive fullscreen mode (which doesn't even
work with mpv's renderer for some reason.) This behaviour can be
disabled by getting a pointer to the IDXGIFactory associated with the
D3D11 device and calling MakeWindowAssociation with the appropriate
flags.
Since the main thread is shared by other things in the player, using STA (single
threaded aparement) may have caused problems. Instead initialize in MTA
(multithreaded apartment).
Instead of implicitly resetting the options to defaults and then
applying the options, they're always applied on top of the current
options (in the same way adding new options to the CLI command line
will).
This does not apply to vo_opengl_cb, because that has an even worse mess
which I refuse to deal with.
Enable m_sub_options_copy() to copy nested sub-options, and also enable
it to create an option struct from defaults. We can get rid of most of
the crap in assign_options() now.
Calling handle_scaler_opt() to get a static allocation for scaler name
is still needed. It's moved to reinit_scaler(), which seems to be a
better place for it. Without it, dangling pointers could be created when
options are changed. (And in fact, this fixes possible dangling pointers
for window.name.) In theory we could create a dynamic copy, but that
seemed even more messy.
Chance of regressions.
Commit 026b75e7 actually enabled changing icc options at runtime (via
vo_cmdline), but it didn't quite work. In particular, changing the icc-
profile option just kept the old profile, because it was cached
accordingly.
As part of this, change gl_lcms.opts from a struct to a pointer to a
struct. We properly copy it, instead of allowing possibly dangling
strings, like it was done in a working but unclean way before.
Also, reinit the whole rendering chain when the auto icc profile
changes, just like it's done when icc options are changed.
Passing the bstr thing as pointer makes no sense. Everywhere else bstr
structs are passed by value because they're so small. Only when it's
supposed to receive a return value they're not.
Originally, video.c did not access any CMS things (other than lut3d
being set on it), but this has changed. In practice, almost all accesses
to it have moved to video.c. vo_opengl only created it, and set the auto
icc profile path.
Complete the move.
Some things wrt. option handling are a bit fishy. (But when is this not
the case.)
icc-profile-auto was not tested, but the distributed human CI will take
care of it.
It gets printed on every alt+tab or desktop switch under mutter and
weston, and offers no useful information since it's handled by
destroying the previous entry.
Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
This commit will cause the wayland backend and vo to correctly report
the display frame rate. This didn't work as VOCTRL_GET_DISPLAY_FPS was
received way too early, before the window was created (and thus
current_output set).
The VO will now signal VO_EVENT_WIN_STATE after window initialization
and upon a resize.
Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
This algorithm works really well. Setting it is a much better
"out-of-the-box" experience than just clipping, which will always look
ugly.
In other words, with this default, users of mpv will just be able to
play HDR content without even realizing it's HDR (pretty much).
Instead of doing HDR tone mapping on an ad-hoc basis inside
pass_colormanage, the reference peak of an image is now part of the
image params (alongside colorspace, gamma, etc.) and tone mapping is
done whenever peak_src != peak_dst.
To get sensible behavior when mixing HDR and SDR content and displays,
target-brightness is a generic filler for "the assumed brightness of SDR
content".
This gets rid of the weird display_scaled hack, sets the framework
for multiple HDR functions with difference reference peaks, and allows
us to (in a future commit) autodetect the right source peak from
the HDR metadata.
(Apart from metadata, the source peak can also be controlled via
vf_format. For HDR content this adjusts the overall image brightness,
for SDR content it's like simulating a different exposure)
The wayland protocol exposes scaling done by the compositor to
compensate for small window sizes on small high DPI displays. If the
program ignores the scaling done, what'll happen is the compositor is
going to ask the program to be scaled down by N times the window size and
then it'll upscale the program's surface by N times. The scaling
algorithm seems to be bilinear so the scaling is quite obvious.
This commit sets up callbacks to listen for the scaling factor of each
output and, on rescale events, notifies the compositor that the
surface's scale is what the compositor asked for and changes the
player's surface to the appropriate size, causing no scaling to be done
by the compositor.
Compositors not supporting this interface will ignore the callbacks and do
nothing, keeping program behaviour the same. For compositors supporting
and using this interface (mutter), this will fix the rendering to be pixel
precise as it should be.
Both the opengl wayland backend and the wayland vo have been fixed to support
this. Verified to not break either on weston and mutter.
Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
Commit 0348cd08 was too naive/simple, and always inserted the d3d11vpp
filter if any d3d11 output image formats were supported, even if it
makes no sense. For example --vf=format=rgb8 already breaks it.
It needs to take the set of supported input formats into account. the
weird format negotiation makes this hard. As a simple and cheap
solution, make some assumptions about the supported formats of a filter.
I hope to simplify this one day by using another format negotiation
algorithm, but this can probably wait.
No reason to do so. See also commit 240ba92b.
Since now many mp_images will never have a pixel aspect ratio set,
redefine a 0/0 aspect ratio to "undefined" instead invalid. This also
brings it more in line with how decoder vs. container aspect ratios are
handled.
Most callers seem to be fine with the new behavior.
mp_image_params_valid() in particular has to be adjusted, or some things
stop working due to mp_images not becoming valid after setting size and
format.
Position the window around the original window center on video size change
(when switching to the next file with different resolution, for example)
instead of keeping the position of its top-left corner fixed.
This is quite unexpected. It's caused by mp_image_set_size(), which is
used to update certain fields which can be format-dependent, but which
is actually also supposed to reset the pixel aspect ratio.
We now have a video filter that uses the d3d11 video processor, so it
makes no sense to have one in the VO interop code. The VO uses it for
formats not directly supported by ANGLE (so the video data is converted
to a RGB texture, which ANGLE can take in).
Change this so that the video filter is automatically inserted if
needed. Move the code that maps RGB surfaces to its own inteorp backend.
Add a bunch of new image formats, which are used to enforce the new
constraints, and to automatically insert the filter only when needed.
The added vf mechanism to auto-insert the d3d11vpp filter is very dumb
and primitive, and will work only for this specific purpose. The format
negotiation mechanism in the filter chain is generally not very pretty,
and mostly broken as well. (libavfilter has a different mechanism, and
these mechanisms don't match well, so vf_lavfi uses some sort of hack.
It only works because hwaccel and non-hwaccel formats are strictly
separated.)
The RGB interop is now only used with older ANGLE versions. The only
reason I'm keeping it is because it's relatively isolated (uses only
existing mechanisms and adds no new concepts), and because I want to be
able to compare the behavior of the old code with the new one for
testing. It will be removed eventually.
If ANGLE has NV12 interop, P010 is now handled by converting to NV12
with the video processor, instead of converting it to RGB and using the
old mechanism to import that as a texture.
Main use: deinterlacing.
I'm not sure how to select the deinterlacing mode at all. You can
enumate the available video processors, but at least on Intel, all of
them either signal support for all deinterlacers, or none (the latter is
apparently used for IVTC). I haven't found anything that actually tells
the processor _which_ algorithm to use.
Another strange detail is how to select top/bottom fields and field
dominance. At least I'm getting quite similar results to vavpp on Linux,
so I'm content with it for now.
Future plans include removing the D3D11 video processor use from the
ANGLE interop code.
This avoids a copy of the video image and lowers vsync jitter. Since
there are now two options to add to the window_attribs list, it has been
made dynamic.
This makes vf_vdpaupp use the deinterlacer helper code already used by
vf_vavpp. I nice side-effect is that this also removes some traces of
code originating from vo_vdpau.c, so we can switch it to LGPL.
Extend the refqueue helper with a deint setting. If not set,
mp_refqueue_should_deint() always returns false, which slightly
simplifies vf_vdpaupp. It's of no consequence to vf_vavpp (other than it
has to set it to get expected behavior).
A lot of real-world shaders start off with comments explaining the usage
or license, generating lots of "empty" passes. This simply change allows
us to skip them, which silences the warning spam and prevents us from
having to store and copy around these empty passes.
It also adds a more useful failure check: Attempting to use a user
shader that doesn't define any passes at all.
This requires the GL_EXT_texture_norm16 extension and works in ANGLE.
A default precision had to be set for sampler3Ds, otherwise the shaders
would fail to compile.
This has often been requested for use on OSD. I don't really like having
such "special" properties, but whatever. Hopefully this will be the only
case.
Untested because I'm too damn lazy.
Fixes#2828.
Remove the opengl-hq option default that caused it not to autoselect
ANGLE (unlike --vo=opengl). Details see commit d5df90a2.
Back then the intention was to use ANGLE by default, since it integrates
much nicer with the Windows compositor (instead of native OpenGL, which
tends to cause crazy glitches). On the other hand, many opengl-hq
capabilities are not available with older ANGLE builds, so it didn't
make any sense to autoselect ANGLE for it.
With the GL_EXT_texture_norm16 extension recently added to ANGLE, it has
essentially reached feature parity to desktop GL for the subset we are
using. (Even the integer texture hack for high bit depth input could be
dropped now.)
It (probably) still does not support nnedi3, due to the weird way the NN
coefficients are imported. Also, it uses half-floats instead of 16 bit
fixed-point textures for technical reasons, which implies about 5 bits
of precision loss. If anyone actually manages to distinguish the two
dithering texture formats in a double-blind test, I will fix it.
Abstracts the annoying framerate-doubling behavior.
Same deal as with refqueue introduction: the code size blows up, but at
least it can be reused for other filters.
Calling this right at start of filter_ext() also fixes a small
regression from previous commit. The change in reference surfaces due to
the first update_pipeline() with deinterlacing enabled changed behavior
of mp_refqueue_next() and mp_refqueue_has_output(). Since
update_pipeline() was called between those, the frame output logic got
inconsistent, and the first deinterlaced frame was duplicated from the
previous non-deinterlaced frame.
Also reset the number of ref-frames when switching back to non-deint
mode.
If the deinterlacer separates fields, the framerate must be doubled.
Since we have no stable and reliably framerate anywhere, we've been
calculating it by taking the time halfway to the next frame.
vf_vavpp actually used the past frame to calculate the frame duration,
which is sort of ok, but will skip the 2nd field in a stream (since the
first frame has no past PTS). This is annoying for testing, so use the
future frame PTS instead, which means the last field of the stream will
be dropped instead.
Move the handling of the future/past frames and the associated dataflow
rules to a separate source file.
While this on its own seems rather questionable and just inflates the
code, I intend to reuse it for other filters. The logic is annoying
enough that it shouldn't be duplicated a bunch of times.
(I considered other ways of sharing this logic, such as an uber-
deinterlace filter, which would access the hardware deinterlacer via a
different API. Although that sounds like kind of the right approach,
this would have other problems, so let's not, at least for now.)
SRW locks are available since Windows Vista. They work essentially like
Linux futexes. In particular, they can be statically initialized, and do
not require deinitialization. This makes them ideal for implementing
PTHREAD_MUTEX_INITIALIZER.
We still need CRITICAL_SECTION for recursive mutexes.
This must be called if a texture shared between D3D devices is updated.
Often enough, the shared devices will be the same device, but ANGLE
forces using shared surfaces. I suppose there is no guarantee the driver
will do the expected thing. Internally, the driver could for example not
insert the required barriers before the shared texture is used.
Fixes#320 (which is closed as 'not our problem' but eh)
Relevant xorg bug: https://bugs.freedesktop.org/show_bug.cgi?id=70931
For me this happened when (accidentally) trying to play a 8460x2812 jpg
file with mpv. Like the referenced bug, xvinfo reports "maximum XvImage
size: 8192 x 8192". So the returned XvImage is 8192x2812 and memory
corruption happens.
Only after handling this BadShmSeg X11 errors are shown.
Rename it to get out of OpenGL's namespace. The gl_ prefix is used by
other mpv functions, but no OpenGL ones.
The "slice" parameter was never actually used, and all callers passed 0
for it.
The main change is actually that e first copy to a "staging" memory
frame, and then upload this at once. The old non-PBO code called
glTexsubImage2D for each OSD sub-bitmap.
The new non-PBO code path is a bit faster now if there are many small
sub-bitmaps (on Linux/nVidia). It's also a bit simpler, so this is a
win.
(Although I don't particularly appreciate the mixed normal/PBO texture
code.)
Some of these checks became pointless after dropping ES 2.0 support for
extended filtering.
GL_EXT_texture_rg is part of core in ES 3.0, and we already check for
this version, so testing for the extension is redundant.
GL_OES_texture_half_float_linear is also always available, at least as
far as our needs go.
The functionality we need from GL_EXT_color_buffer_half_float is always
available in ES 3.2, and we explicitly check for ES 3.2, so reject this
extension if the ES version is new enough.
For some reason, GLES has no glMapBuffer, only glMapBufferRange.
GLES 2 has no buffer mapping at all, and GL 2.1 does not always have
glMapBufferRange. On those PBOs remain unsupported (there's no reason to
care about GL 2.1 without the extension).
This doesn't actually work on ANGLE, and I have no idea why. (There are
artifacts on OSD, as if parts of the OSD data weren't copied.) It works
on desktop OpenGL and at least 1 other ES 3 implementation. Don't enable
it on ANGLE, I guess.
Not sure how much can be gained with this, as we can't use it properly
yet. For now, this is used only before rendering, which probably does
overwhelmingly nothing.
In the future, this should be used after temporary passes, which could
possibly reduce memory usage and even memory bandwidth usage, depending
on the drivers.
Center window position after applying W and H parameters of the --geometry
option. Passing valid X and Y values will still override the position.
Fixes#2397.
Before that position of the window top-left corner was remaining the same
when the window was scaled.
Right now VOCTRL_SET_UNFS_WINDOW_SIZE is called only by window-scale. This
change will not affect resizes made by the user (dragging window edge).
Fixes#3164.
Center the window on the original window center instead of the screen center
when the window has been resized due to requested window size exceeding the
size of the screen.
If user moved the window, he probably did it for the reason and he probably
don't want it to get back to the center of the screen when he is resizing it
(with window-scale for example).
Properly update stored client area size when the window is resized in
reinit_window_state due to window size exceeding the size of the screen.
This was causing wrong behavior with window-scale - when window size was
becoming too big the window was resized but the video was not.
I've got a broken webm that fails to seek correctly with "--start=0".
The problem is that every index entry points to 1 byte before cluster
start (!!!). demux_mkv tries to resync to the next cluster, but since it
already has read 2 bytes with ebml_read_id(), it doesn't get the first
cluster, but the following one. Actually, it can be any amount of bytes
from 1-4, whatever happens to look valid at this essentially random byte
position.
Improve this by resyncing from the original position, instead of the one
after the EBML element ID has been attempted to be read.
The file shows the following headers:
| + Muxing application: google at 177
| + Writing application: google at 186
Indeed, the file was downloaded with youtube-dl. I can only guess that
Google got it completely wrong.
Following commit 84ccebd9, the internal helpers don't allow GL_RGB and
GL_RGBA as internal formats for FBO attachments anymore.
While OpenGL itself is perfectly fine with it, I don't see much of a
reason to bother, and mixing sized and unsized internal formats is
confusing anyway.
Just remove these formats.
This code evolved into an ifdef mess as support for cancellation on
Windows was added. Make the Windows-specific code completely separate.
It looks cleaner, and it also means that some of the posix code is not
uselessly enabled on Windows. The latter made msvcrt.dll output warnings
because it does not like -1 passed as FD to read/write. (The same would
be harmless on POSIX.)
Cropping usually happens by adjusting the plane start pointers and the
image size. The former is obviously not possible for opaque hwaccel
formats, but the latter must work.
Since the code already takes care of aligning the top/left crop origin
to chroma alignment, simply set the crop origin to 0/0 in the hwaccel
case. Also add a message if such an adjustment happens.
Supporting this isn't worth much; the main usefulness is with debugging.
ES 2.0 has this weird rule that not the internalformat parameter
determines the internal format, but the combination of all texture
parameters. GL_OES_texture_half_float thus does not specify e.g. a
GL_RGBA16F format, but requires passing GL_RGBA as format and
GL_HALF_FLOAT_OES as type. We won't bother with this, since ES 2.0 is a
lost cause anyway.
This also removes the OpenGL error when the code is trying to create a
f16 FBO for testing whether FBOs work.
gl_video_upload_image() can fail in the hardware decoding case. In this
case rendering continued "normally", which meant that pass_get_img_tex()
would kill the process with an assertion failure.
Fix this by allowing gl_video_upload_image() to fail, and exit rendering
early enough to skip code which requires an image to be present. (Maybe
this is still a bit too subtle, but better than before.)
Set an error flag, and render the blue screen we introduced for shader
errors. (For this purpose also move the rendering of it to final output,
to ensure it's visible at all.) The error flag is temporary, because the
associated failure might also be temporary, unlike shader compilation
errors.
ANGLE doesn't handle this very strictly. But if they change this in the
future, it shouldn't brick us.
Not quite happy with this glsl_extensions fields, but it is quite
unintrusive after all.
No reason not to, and makes the following commit slightly simpler.
In fact, this makes the shaders more correct too. Normally, "#extension"
must come before any normal shader text, including the "precision"
directive. Not sure why this worked before. (Probably didn't.)
ANGLE was missing texture() overloads in the shader compiler for
GL_TEXTURE_EXTERNAL_OES textures. Support has been added upstream,
so we can use it now.
With the new hooks mechanism, user shaders and such are actually loaded
before rendering starts, instead of being loaded during rendering. This
is used to cache them (instead of e.g. reparsing them every frame).
The cached state wasn't cleared correctly in some situations. Namely,
resizing didn't correctly enable/disable prescale hooks.
Reorganize how these reinitializations are handled. Get rid of
reinit_rendering(), whose meaning was pretty unclear. Call the required
functions to reset or recreate state directly wherever they are needed.
For some reason, the d3d9/dxva2/d3d11 DLLs are still optional. But we
don't need to try so hard to keep exact references. In fact, there's no
reason to unload them at all.
So load them once in a central place. For simplicity, the d3d9/d3d11
backends both load all DLLs. (They will error out only if the required
DLLs could not be loaded.)
In theory, we could just call LoadLibrary multiple times (without
calling FreeLibrary), but I'm slightly worried that this could be
detected as a "bug", or that the reference count could even have a low
static limit that could be hit soon.
wscript builds hwdec_dxva2gldx.c if gl-dxinterop is enabled, while
video/dxva2.c depends on d3d-hwaccel. If d3d-hwaccel is disabled, then
hwdec_dxva2gldx.c will fail to link, because it uses
d3d9_surface_in_mp_image(), defined in dxva2.c.
Fix this by removing the use of this function. It has barely any value
at this point anyway. Just use the libavcodec documented way to get the
surface directly.
Fixes#3150.
Use dynamic memory allocation, as the static allocation is starting to
get annoying.
Currently, SC_MAX_ENTRIES is essentially still a static upper limit on
the number of shaders. But in future we could try a more clever cache
replacement strategy, which does not keep stale entries forever if the
maximum happens not to be reached.
The new uniforms introduced by 362015c have exceeded the uniform limit
when using high-radius tscale. In addition, the SC limit of 32 entries
might be pushing it with user shaders.
Just make these value a bigger to delay the onset of this same failure
mode. Maybe in the future it should be reworked to grow dynamically?
Either way, we *can* always predict a static upper bound on the number
of uniforms and shader cache entries, it's just that we forgot to do so.
Fixes#3151
This makes it so that users with actual HDR displays can just set their
config to target-trc=st2084 and get native HDR output. This will look a
bit silly for SDR content (everything will be really bright), but for
lack of a better tone mapping situation (including reverse tone mapping)
this is the easiest thing to do for now.
Ideally the brightness metadata should be part of the colorspace struct
or something (with mpv always adapting where necessary), but it depends
on the TRC and not the primaries so it's a bit more complicated than
that.
Since dumb mode is affected by tone mapping (which I'll call a feature,
not a bug), we need to copy over the configuration - in particular, the
defaults. (To prevent a render failure)
Since HDR content is now auto-detected as such, we should probably do
something smarter in the "no configuration" case, such as outputting
gamma 2.2 instead.
This decision will affect the majority of users of stock configurations
who just play back appropriately tagged HDR files, so having a good
default behavior is important. "Output the HDR content as-is" is
definitely not likely to give the user a good result.
This now lets us auto-detect appropriately tagged HDR content using
FFmpeg's new TRC entries (when available).
Hidden behind an #if because Libav stable doesn't have it yet.
Make it dynamic and never remove entries from it.
For now, this is better than possibly creating dangling pointers all
over the place in the gl_user_shader struct.
Untested.
This is now a configurable option, with tunable parameters.
I got inspiration for these algorithms off wikipedia. "simple" seems to
work pretty well, but not well enough to make it a reasonable default.
Some other notable candidates:
- Local functions (e.g. based on local contrast or gradient)
- Clamp with soft knee (linear up to a point)
- Mapping in CIE L*Ch. Map L smoothly, clamp C and h.
- Color appearance models
These will have to be implemented some other time.
Note that the parameter "peak_src" to pass_tone_map should, in
principle, be auto-detected from the SEI information of the source file
where available. This will also have to be implemented in a later
commit.
Due to the way color management in mpv worked historically, the subtitle
blending function was written to preserve the linearity of the input.
(In the past, the 3DLUT function required linear inputs)
Since the 3DLUT was refactored to accept the video color directly, the
re-linearization after blending is now virtually always redundant.
(Notably, it's also redundant when CMS is turned off, so this way of
writing the code stopped making sense a long time ago. It is a remnant
from before the pass_colormanage function was as flexible as it is now)
Currently, this relies on the user manually entering their display
brightness (since we have no way to detect this at runtime or from ICC
metadata). The default value of 250 was picked by looking at ~10 reviews
on tftcentral.co.uk and realizing they all come with around 250 cd/m^2
out of the box. (In addition, ITU-R Rec. BT.2022 supports this)
Since there is no metadata in FFmpeg to indicate usage of this TRC, the
only way to actually play HDR content currently is to set
``--vf=format=gamma=st2084``. (It could be guessed based on SEI, but
this is not implemented yet)
Incidentally, since SEI is ignored, it's currently assumed that all
content is scaled to 10,000 cd/m^2 (and hard-clipped where out of
range). I don't see this assumption changing much, though.
As an unfortunate consequence of the fact that we don't know the display
brightness, mixed with the fact that LittleCMS' parametric tone curves
are not flexible enough to support PQ, we have to build the 3DLUT
against gamma 2.2 if it's used. This might be a good thing, though,
consdering the PQ source space is probably not fantastic for
interpolation either way.
Partially addresses #2572.
This is much more readable than hard-coding magic IDs all over the file,
and removes the need for all the explanatory comments that were a direct
result of this.
This macro takes care of rotation, swizzling, integer conversion and
normalization automatically. I found the performance impact to be
nonexistant for superxbr and debanding, although rotation *did* have an
impact due to the extra matrix multiplication. (So it gets skipped where
possible)
All of the internal hooks have been rewritten to use this new mechanism,
and the prescaler hooks have finally been separated from each other.
This also means the prescale FBO kludge is no longer required.
This fixes image corruption for image formats like 0bgr, and also fixes
prescaling under rotation. (As well as other user hooks that have
orientation-dependent access)
The "raw" attributes (tex, tex_pos, pixel_size) are still un-rotated, in
case something needs them, but ideally the hooks should be rewritten to
use the new API as much as possible. The hooked texture has been renamed
from just NAME to NAME_raw to make script authors notice the change (and
also deemphasize direct texture access).
This is also a step towards getting rid of the use_integer pass.
This replaces the previous TRANSFORM by WIDTH, HEIGHT and OFFSET where
WIDTH and HEIGHT are RPN expressions. This allows for more fine-grained
control over the output size, and also makes sure that overwriting
existing textures works more cleanly.
(Also add some more useful bstr functions)
This allows users to add their own near-arbitrary hooks to the vo_opengl
processing pipeline, greatly enhancing the flexibility of user shaders.
This enables, among other things, user shaders such as CrossBilateral,
SuperRes, LumaSharpen and many more.
To make parsing the user shaders easier, shaders are now loaded as
bstrs, and the hooks are set up during video reconfig instead of on
every single frame.
These are "sequence points" where the image could be rendered out to an
FBO, hooked, and re-loaded if any such hook exists. This is perfect for
things like the current user shaders system, as well as optional effects
like unsharp masking.
Note that since we have to pick *some* FBO to store the optionally
hooked texture, we just store it in an array indexed by an increasing
counter. Since we only ever store as many as MAX_TEXTURE_HOOKS + all
internal hook points entries, this is guaranteed to be enough space.
This commit also removes some of the now unused FBOs.
The hook mechanism allows arbitrary processing stages to get dispatched
whenever certain named textures have been "finalized" by the code.
This is mostly meant to serve as a change that opens up the internal
processing in pass_read_video to user scripts, but as a side benefit all
of the code dealing with offsets and plane alignment and other such
confusing things has been rewritten.
This hook mechanism is powerful enough to cover the needs of both
debanding and prescaling (and more), so as a result they can be removed
from pass_read_video entirely and implemented through hooks.
Some avenues for optimization:
- The prescale hook is currently somewhat distributed code-wise. It might be
cleaner to split it into superxbr and NNEDI3 hooks which can each be
self-contained.
- It might be possible to move a large part of the hook code out to an
external file (including the hook definitions for debanding and
prescaling), which would be very much desired.
- Currently, some stages (chroma merging, integer conversion) will
*always* run even if unnecessary. I'm planning another series of
refactors (deferred img_tex) to allow dropping unnecessary shader
stages like these, but that's probably some ways away. In the meantime
it would be doable to re-add some of the logic to skip these stages if
we know we don't need them.
- More hook locations could be added (?)
Instead of rounding down, we round to the nearest float. This reduces
the maximum possible error introduced by this rounding operation. Also
clarify the comment.
Remove non-texture_rg compatibility from LUT sampling. OpenGL without
texture_rg support will always trigger dumb-mode, and dumb-mode does not
use LUTs. It used not to, and that was when this made sense.
Fixes broken colors with --vf=format=0bgr (but only if deband is
disabled).
0bgr means the first byte is padding, while the following three bytes
are bgr. From the vo_opengl perspective, it has 4 physical components
with 3 logical components. copy_img_tex() simply copied 3 components
from the physical representation, which means the last component (r) was
sliced off.
Fix this by not using p->color_swizzle for packed formats, and instead
let packed formats set the per-plane swizzle in texplane.swizzle. The
latter applies the swizzle as part of operation in copy_img_tex(), which
essentially moves physical to logical representations.
Unfortunately, debanding (and thus with opengl-hq defaults) is still
broken.
Make the find_plane_format function take a bit count.
This also makes the function's comment true for the first time the
function and its comment exist. (It was commented as taking bits, but
always took bytes.)
Only a few very low bit depth internal formats can be rendered to in
pure ES2 (GL_RGB565 is the "best" one).
Seems like the only potentially reasonable renderable formats in ES2
could be provided via GL_OES_rgb8_rgba8, or half-floats, so don't
bother with this at all.
Remove the attempted cleverness; keep it dumb.
This strictly calculates the average speed over an at least 1 second
window (longer if I/O blocks it).
Since this doesn't reset the speed anymore when reading stops by going
idle, the results might actually be more accurate now.
Now that we know in advance whether an implementation should support a
specific format, we have more flexibility when determining which format
to use.
In particular, we can drop the roundabout ES logic.
I'm not sure if actually trying to create the FBO for probing still has
any value. But it might, so leave it for now.
Even if everything else is available, the need for first class arrays
breaks it. In theory we could fix this since we don't strictly need
them, but I guess it's not worth bothering.
Also give the misnamed have_mix variable a slightly better name.
This merges all knowledge about texture format into a central table.
Most of the work done here is actually identifying which formats exactly
are supported by OpenGL(ES) under which circumstances, and keeping this
information in the format table in a somewhat declarative way. (Although
only to the extend needed by mpv.) In particular, ES and float formats
are a horrible mess.
Again this is a big refactor that might cause regression on "obscure"
configurations.
This uses the normal autoprobing rules like "auto", but rejects anything
that isn't flagged as copying data back to system memory.
The chunk in command.c was dead code, so remove it instead of updating
it.
MSDN documents this as "Introduced in Windows 8.1.". I assume on Windows
7 this field will simply be ignored. Too bad for Windows 7 users.
Also, I'm not using D3D11_VIDEO_PROCESSOR_NOMINAL_RANGE_16_235 and
D3D11_VIDEO_PROCESSOR_NOMINAL_RANGE_0_255, because these are apparently
completely missing from the MinGW headers. (Such a damn pain.)
We don't have any reason to disable either. Both are loaded dynamically
at runtime anyway. There is also no reason why dxva2 would disappear
from libavcodec any time soon.
ANGLE is _really_ annoying to build. (Requires special toolchain and a
recent MSVC version.) This results in various issues with people
having trouble to build mpv against ANGLE (apparently linking it
against a prebuilt binary doesn't count, or using binaries from
potentially untrusted sources is not wanted).
Dynamically loading ANGLE is going to be a huge convenience. This commit
implements this, with special focus on keeping it source compatible to
a normal build with ANGLE linked at build-time.
In theory this was needed for the previous commit (but wasn't in
practice, since for hwdec the LUMINANCE_ALPHA mangling is not applied
anymore, and ANGLE uses RG textures in absence of GL_ARB_texture_rg for
whatever crazy reasons).
In practice this caused funky colors on OSX with the uyvy422 format,
which is also fixed in this commit.
This uses EGL_ANGLE_stream_producer_d3d_texture_nv12 and related
extensions to map the D3D textures coming from the hardware decoder
directly in GL.
In theory this would be trivial to achieve, but unfortunately ANGLE does
not have a mechanism to "import" D3D textures as GL textures. Instead,
an awkward mechanism via EGL_KHR_stream was implemented, which involves
at least 5 extensions and a lot of glue code. (Even worse than VAAPI EGL
interop, and very far from the simplicity you get on OSX.)
The ANGLE mechanism so far supports only the NV12 texture format, which
means 10 bit won't work. It also does not work in ES3 mode yet. For
these reasons, the "old" ID3D11VideoProcessor code is kept and used as a
fallback.
It forces es2 mode on ANGLE. Only useful for testing. Since the normal
"angle" backend already falls back to es2 if es3 does not work, this new
backend always exit when autoprobing it.
Rename gl_hwdec_driver.map_image to map_frame, and let it fill out a
struct gl_hwdec_frame describing the exact texture layout. This gives
more flexibility to what the hwdec interop can export. In particular, it
can export strange component orders/permutations and textures with
padded size. (The latter originating from cropped video.)
The way gl_hwdec_frame works is in the spirit of the rest of the
vo_opengl video processing code, which tends to put as much information
in immediate state (as part of the dataflow), instead of declaring it
globally. To some degree this duplicates the texplane and img_tex
structs, but until we somehow unify those, it's better to give the hwdec
state its own struct. The fact that changing the hwdec struct would
require changes and testing on at least 4 platform/GPU combinations
makes duplicating it almost a requirement to avoid pain later.
Make gl_hwdec_driver.reinit set the new image format and remove the
gl_hwdec.converted_imgfmt field.
Likewise, gl_hwdec.gl_texture_target is replaced with
gl_hwdec_plane.gl_target.
Split out a init_image_desc function from init_format. The latter is not
called in the hwdec case at all anymore. Setting up most of struct
texplane is also completely separate in the hwdec and normal cases.
video.c does not check whether the hwdec "mapped" image format is
supported. This should not really happen anyway, and if it does, the
hwdec interop backend must fail at creation time, so this is not an
issue.
Older ffmpeg releases don't have ffmpeg git commit
50401f5fb7d778583b03a13bc4440f71063d319d, which fixes ffmpeg's
pkt_timebase check to reject its default "unset" timebase as invalid.
The consequence was that all non-PGS bitmap subtitle timestamps were
forced to 0.
Of course this hit _only_ shitty distros using outdated/badly maintained
ffmpeg releases, so this is not worth working around. I've already
wasted a lot of time on analyzing this dumb issue, and it could be
useful for bisecting, so don't drop pre-3.0 ffmpeg just yet.
Fixes#3109.
libass's ass_process_chunk expects long long int for the timecode and
durations arguments, thus should use llrint instead of lrint.
This does not cause any problems on most platforms, but on cygwin, it
causes strange subtitle behaviour, like subtitles not showing, getting
stuck or old subtitles showing at the same time as new subtitles.
Commit 786f37ae accidentally changed seeking behavior such that
continuous seeking (holding the seek button down) would use the previous
seek target timestamp, instead of the new video timestamp. (This is for
the default mode, seeking to keyframes.)
The result is that the movement on the seekbar is smooth, but the way
the video updates is awkward. Some might actually prefer the new
behavior (and some players effectively show similar bahavior), but I
don't. So restore the old behavior.
This is done in two steps:
First: strictly wait for the entire seek process to finish, which will
effectively make the seeking code pick up the new video timestamp
correctly.
This would play audio immediately, which would result in noise during
continuous seeking, which leads to second: explicitly abort the playback
restarting process if this case is detected, and never play audio.
The main change is with video/hwdec.h. mp_hwdec_info is made opaque (and
renamed to mp_hwdec_devices). Its accessors are mainly thread-safe (or
documented where not), which makes the whole thing saner and cleaner. In
particular, thread-safety rules become less subtle and more obvious.
The new internal API makes it easier to support multiple OpenGL interop
backends. (Although this is not done yet, and it's not clear whether it
ever will.)
This also removes all the API-specific fields from mp_hwdec_ctx and
replaces them with a "ctx" field. For d3d in particular, we drop the
mp_d3d_ctx struct completely, and pass the interfaces directly.
Remove the emulation checks from vaapi.c and vdpau.c; they are
pointless, and the checks that matter are done on the VO layer.
The d3d hardware decoders might slightly change behavior: dxva2-copy
will not use the VO device anymore if the VO supports proper interop.
This pretty much assumes that any in such cases the VO will not use any
form of exclusive mode, which makes using the VO device in copy mode
unnecessary.
This is a big refactor. Some things may be untested and could be broken.
When we receive the wl_shell_surface::configure event, it makes sense
to respect the aspect ratio of the video in windowed mode, but in
fullscreen it forces compositing and wastes resources (until atomic
modesetting is available everywhere and we can stop having
desynchronised planes).
Weston mitigates a resolution mismatch by creating black surfaces and
compositing them around the fullscreen surface, placed at the middle,
while GNOME puts it at the top-left and leaves the rest of the desktop
composited below, both of them producing a subpar experience.
Fixes#3021, #2657.
Add --taskbar-progress command line option and property which controls taskbar
progress indication rendering in Windows 7+. This option is on by default and
can be toggled during playback.
This option does not affect the creation process of ITaskbarList3. When the
option is turned off the progress bar is just hidden with TBPF_NOPROGRESS.
Closes#2535
The X11 error handler is global, and not per-display. If another Xlib
user exists in the process, they can conflict. In theory, it might
happen that e.g. another library sets an error handler (overwriting the
mpv one), and some time after mpv closes its display, restores the error
handler to mpv's one. To mitigate this, check if the error log instance
is actually set, instead of possibly crashing.
The change in vo_x11_uninit() is mostly cosmetic.
Some of this documentation was left woefully inaccurate as color
management in mpv evolved. This commit updates all of the wording and
adds notes and comments where appropriate.
This reportedly makes it work on ODROID-C2. The idea for this hack is
taken from kodi; they unconditionally set some or all of those flags.
I don't trust ALSA enough to hope that setting these flags couldn't
break something else, so we try without them first.
It's not clear whether this is a driver bug or a bug in the ALSA libs.
There is no ALSA bug tracker (the ALSA website has had a dead link to
a deleted bug tracker fo years). There's not much we can do other than
piling up ridiculous hacks. At least I think that at this point invalid
API usage by mpv can be excluded as a cause.
ALSA might be the worst audio API ever.
In order to honor the differences between OpenGL and Direct3D coordinate
systems, ANGLE uses a full FBO copy merely to flip the final frame
vertically. This can be avoided with the EGL_ANGLE_surface_orientation
extension.
I hope that this does what we expect it does: destroy the EGLDisplay
specific to our HDC. (Some implementations will terminate all EGL
contexts in the whole process.)
eglReleaseThread() merely calls eglMakeCurrent(0, 0, 0, 0), which is
not enough.
This commit also fixes the problem fixed with the previous commit,
but I think both changes are needed to make our API usage clean.
If ANGLE was probed before (but rejected), the ANGLE API can remain
"initialized", and eglGetCurrentDisplay() will return a non-NULL
EGLDisplay. Then if a native GL context is used, the ANGLE/EGL API will
then (apparently) keep working alongside native OpenGL API. Since GL
objects are just numbers, they'll simply fail to interact, and OpenGL
will get invalid textures. For some reason this will result in black
textures.
With VAAPI-EGL, something similar could happen in theory, but didn't in
practice.
Options/properties that are choices, and which include "yes" or "no"
values (or both) can now be read and written as MPV_FORMAT_FLAG.
For write access, rejecting flags in these cases was obnoxiously
unintuitive and inconvenient.
For read access, the value of this is less convincing, and actually it's
a major API change. At this point I probably have to admit that the
finer details of the client API are very unstable.
Introduce hwdec-current and hwdec-interop properties.
Deprecate hwdec-detected, which never made a lot of sense, and which is
replaced by the new properties. hwdec-active also becomes useless, as
hwdec-current is a superset, so it's deprecated too (for now).
Cache misses are a normal and expected part of the operation of a cache.
It doesn't really make sense to show a user-visible warning for them.
To work-around this, just skip trying to open the cache if it doesn't
exist yet.
First of all, black point compensation is now on by default. This is
really rather harmless and only improves the result (where "improvement"
means "less black clipping").
Second, this adds an option to limit the ICC profile's contrast, which
helps for untagged matrix profiles that are implicitly black scaled even
in colorimetric intent. (Note that this relies on BPC being enabled to
work properly, which is why the two changes are tied together)
Third, this uses the LittleCMS built in black point estimator instead of
relying on the presence of accurate A2B tables. This also checks tags
and does some amounts of noise elimination.
If the option is unspecified and the profile is missing black point
information, print a warning instructing the user to set the option, and
fall back to 1000 otherwise.
For "current" markers on OSD properties like chapter-list. The marker is
now an actual arrow instead of "> ", and non-current entries will have
the same indentation as the current entry.
While I'm not entirely sure about the new look of those lists, it's a
bit better than the visual mess that was before.
The vdpau_mixer could fail to be recreated properly if preemption
occured at some point before playback initialization (like when using
--hwdec-preload and the opengl-cb API).
Normally, the vdpau_mixer was supposed to be marked invalid when the
components using it detect a preemption, e.g. in hwdec_vdpau.c. This one
didn't mark the vdpau_mixer as invalid if preemption was detected in
reinit(), only in map_image().
It's cleaner to detect preemption directly in the vdpau_mixer, which
ensures it's always recreated correctly.
The "fs-only" choice sets the _NET_WM_BYPASS_COMPOSITOR to 1 if the
window is fullscreened, and 0 otherwise. (0 is specified to be the
implicit default - i.e. no change is requested in windowed mode.)
In particular, change the default to "fs-only".
Fixes#2582.
Including initguid.h at the top of a file that uses references to GUIDs
causes the GUIDs to be declared globally with __declspec(selectany). The
'selectany' attribute tells the linker to consolidate multiple
definitions of each GUID, which would be great except that, in Cygwin
and MinGW GCC 6.1, this method of linking makes the GUIDs conflict with
the ones declared in libuuid.a.
Since initguid.h obsoletes libuuid.a in modern compilers that support
__declspec(selectany), add initguid.h to all files that use GUIDs and
remove libuuid.a from the build.
Fixes#3097
Flag that is set by default. Reseting it will result in mpv trying to fit
client area with video instead of the whole window with border and
decorations on the screen.
Marked as (Windows only) for now until it's implemented on other platforms.
Substraction of 1 is not necessary due to .right and .bottom values of RECT
struct describing a point one pixel outside of the rectangle.
Fixes#2935. (2.b)
--sub-ass=no / --ass=no still work, but --ass-style-override=strip is
preferred now. With this change, --ass-style-override can control all
the types of style overriding.
In particular, this moves the depth test to common code.
Should be functionally equivalent, except that for DXVA2, the
IDirectXVideoDecoderService_GetDecoderRenderTargets API is called
more often potentially.
Because it's annoying and feels unnatural.
If the B point is set while paused, don't seek. If not paused, it should
properly loop immediately.
In theory there's a chance that it will show at least 1 frame after the
loop point when setting the B point. But let's not care about that.
This gives us 16 bit fixed-point integer texture formats, including
ability to sample from them with linear filtering, and using them as FBO
attachments.
The integer texture format path is still there for the sake of ANGLE,
which does not support GL_EXT_texture_norm16 yet.
The change to pass_dither() is needed, because the code path using
GL_R16 for the dither texture relies on glTexImage2D being able to
convert from GL_FLOAT to GL_R16. GLES does not allow this. This could be
trivially fixed by doing the conversion ourselves, but I'm too lazy to
do this now.
This fixes backstepping getting "stuck" when e.g. holding down a key
bound to the backstep command. The reason is that even if the backstep
itself is finished, the next backstep might not take the new video PTS
as reference if the hr-seek itself isn't finished yet.
The intention of not waiting for the hr-seek to finish was faster
backstepping by possibly skipping audio decoding. But it probably
doesn't matter enough to make the rest of the code more complex.
Basically this gets rid of the need for the accessors in d3d11va.h, and
the code can be cleaned up a little bit.
Note that libavcodec only defines a ID3D11VideoDecoderOutputView pointer
in the last plane pointers, but it tolerates/passes through the other
plane pointers we set.
This uses ID3D11VideoProcessor to convert the video to a RGBA surface,
which is then bound to ANGLE. Currently ANGLE does not provide any way
to bind nv12 surfaces directly, so this will have to do.
ID3D11VideoContext1 would give us slightly more control about the
colorspace conversion, though it's still not good, and not available
in MinGW headers yet.
The video processor is created lazily, because we need to have the coded
frame size, of which AVFrame and mp_image have no concept of. Doing the
creation lazily is less of a pain than somehow hacking the coded frame
size into mp_image.
I'm not really sure how ID3D11VideoProcessorInputView is supposed to
work. We recreate it on every frame, which is simple and hopefully
doesn't affect performance.
As a positive side-effect, this also errors out gracefully for the
extremely unlikely but possible case certain builtin filters are not
available. (This could happen only with crippled libavfilter builds that
can't be used by anything using its public API.)
Setting this here is a race condition. It's called from a CoreAudio
callbacks, and there are no locks. It's a string, so this can be
potentially severe.
It's hard to fix and only CoreAudio supported it, so remove it.
This causes the "audio-out-detected-device" property to return nothing
on all platforms.
Sucks, but better than freezing forever given the (to me) unpredictable
RPI behavior. This will be good enough to drop out of vsync timing mode,
or to abort playback.
Recreate all dispmanx objects after mode changes signalled by the TV
callback. This is needed since dispmanx objects are marked as invalid
and cease working.
One important point is that the vsync callbacks will stop coming when
this happens, so restoring the callback is important.
Note that the MMAL renderer itself does not get trashed by the firmware
on such events, but we completely reconfigure it anyway when it happens.
For Mediacodec in particular we don't care about the format. It can just
decode to whatever it wants. The only case we would care about is it not
returning an opaque format if we don't have proper interop, but
libavcodec always returns non-opaque formats by default.
Use the recently added lavc_suffix mechanism to select the wrapper
decoder.
With all hwdec callbacks being optional, and RPI/Mediacodec having only
dummy callbacks, all the callbacks can be removed as well.
The result is that the vd_lavc_hwdec struct for both of them is tiny.
It's better to move them to vd_lavc.c directly, because they are so
trivial and small.
This is intended for cases when --hwdec needs to override the decoder
implementation in use, like for example on the RPI.
It does two things:
1. Allow the hwdec to indicate a decoder suffix. libavcodec by
convention adds a suffix to all wrapper decoders, and here we start
relying on it. While not necessarily the best idea, it's the only
thing we got. libavcodec's hwaccel list is useless, because it only
has the codec ID, not the associated decoder's name.
2. Make --hwdec=auto work properly. It shouldn't fail anymore, and hwdec
probing should reliably work, even if a different decoder is selected
with --vd. The semantics of --hwdec should dictate that it overrides
the default decoder.
A minor simplification. Most callers don't need this, and there's no
good reason why the caller should provide an "initializer" like this.
(This function calls mp_image_new_dummy_ref(), which has no reason
for an initializer either.)
Another crappy fix for timestamp reset issues. This time, we try to fix
files which have very weird but legitimate frame durations, such as
cdgraphics. It can have many short frames, but once in a while there are
potentially very long frames.
Fixes#3027.
The active texture and some pixelstore parameters are now always reset
to defaults when entering and leaving the renderer. Could be important
for libmpv.
Commit 382bafcb changed the behavior for ab-loop-a. This commit changes
ab-loop-b so that the behavior is symmetric.
Adjust the OSD rendering accordingly to the two changes.
Also fix mentions of the "ab_loop" command to the now preferred
"ab-loop".
The check whether video is ready yet was done only in STATUS_FILLING.
But it also switched to STATUS_READY, which means the next time
fill_audio_out_buffers() was called, audio would actually be started
before video.
In most situations, this bug didn't show up, because it was only
triggered if the demuxer didn't provide video packets quickly enough,
but did for audio packets.
Also log when audio is started.
(I hate fill_audio_out_buffers(), why did I write it?)
Strictly schedule an update in regular intervals as long as either
stream cache or demuxer are prefetching. Don't update just always
because the stream cache is enabled ("idle != -1") or cache-related
properties are observed (mp_client_event_is_registered()).
Also, the "idle" variable was awkard; get rid of it with equivalent
code.
Calculate the buffering percentage in the same code which determines
whether the player is or should be buffering. In particular it can't
happen that percentage and buffering state are slightly out of sync due
to calling DEMUXER_CTRL_GET_READER_STATE and reusing it with the
previously determined buffering state.
Now it's also easier to guarantee that the buffering state is updated
properly.
Add some more verbose output as well.
(Damn I hate this code, why did I write it?)
demux_playlist.c recognizes if the source stream points to a directory,
and adds its directory entries. Until now, only 1 level was added.
During playback, further directory entries could be resolved as
directory paths were "played".
While this worked fine, it lead to frequent user confusion, because
playlist resuming and other things didn't work as expected. So just
recursively scan everything.
I'm unsure whether it's a good fix, but at least it gets rid of the
complaints. (And probably will add others.)
Commit a9bd4535 generally changed properties are set to string values.
This actually broke the fallback for non-string properties, because the
set string action was redirected directly to the property, instead of
the generic handler and its fallback code.
As of ffmpeg git master, only the libavdevice decklink wrapper supports
this. Everything else has dropped support.
You're now supposed to use AV_CODEC_ID_WRAPPED_AVFRAME, which works a
bit differently. Normal AVFrames should still work for these encoders.
And remove the same thing from the client API code.
The command.c code has to deal with many specialized M_PROPERTY_SET_*
actions, and we bother with a subset only.
If a mpv_node wrapped a string, the behavior was different from calling
mpv_set_property() with MPV_FORMAT_STRING directly. Change this.
The original intention was to be strict about types if MPV_FORMAT_NODE
is used. But I think the result was less than ideal, and the same change
towards less strict behavior was made to mpv_set_option() ages ago.
Since what we're doing is a linear blend of the four colors, we can just
do it for free by using GPU sampling.
This requires significantly fewer texture fetches and calculations to
compute the final color, making it much more efficient. The code is also
much shorter and simpler.
Now it will always be able to seek back to the start, even if the index
is sparse or misses the first entry.
This can be achieved by reusing the logic for incremental index
generation (for files with no index), and start time probing (for making
sure the first block is always indexed).
Until now, we have made the assumption that a driver will use only 1
hardware surface format. the format is dictated by the driver (you
don't create surfaces with a specific format - you just pass a
rt_format and get a surface that will be in a specific driver-chosen
format).
In particular, the renderer created a dummy surface to probe the format,
and hoped the decoder would produce the same format. Due to a driver
bug this required a workaround to actually get the same format as the
driver did.
Change this so that the format is determined in the decoder. The format
is then passed down as hw_subfmt, which allows the renderer to configure
itself with the correct format. If the hardware surface changes its
format midstream, the renderer can be reconfigured using the normal
mechanisms.
This calls va_surface_init_subformat() each time after the decoder
returns a surface. Since libavcodec/AVFrame has no concept of sub-
formats, this is unavoidable. It creates and destroys a derived
VAImage, but this shouldn't have any bad performance effects (at
least I didn't notice any measurable effects).
Note that vaDeriveImage() failures are silently ignored as some
drivers (the vdpau wrapper) support neither vaDeriveImage, nor EGL
interop. In addition, we still probe whether we can map an image
in the EGL interop code. This is important as it's the only way
to determine whether EGL interop is supported at all. With respect
to the driver bug mentioned above, it doesn't matter which format
the test surface has.
In vf_vavpp, also remove the rt_format guessing business. I think the
existing logic was a bit meaningless anyway. It's not even a given
that vavpp produces the same rt_format for output.
Loaded 100 of 400 files, more files were not shown because too many files have changed in this diff.
Show more
Reference in new issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Martin Herkt