mpv/player/audio.c

963 lines
30 KiB
C
Raw Normal View History

/*
* This file is part of mpv.
*
player: change license of most core files to LGPL These files have all in common that they were fully or mostly taken from mplayer.c. (mplayer.c was a huge file that contains almost all of the playback core, until it was split into multiple parts.) This was probably the hardest part to relicense, because so much code was moved around all the time. player/audio.c still does not compile. We'll have to redo audio filtering. Once that is done, we can probably actually provide an actual LGPL configure switch. Here is a relatively detailed list of potential issues: 8d190244: author did not reply, parts were made GPL-only in a previous commit. 7882ea9b: author could not be reached, but the code is gone. wscript still has --datadir switch, but I don't think this is relevant to copyright. f197efd5: unclear origin, but I consider the code gone anyway (replaced with generic OSD mechanisms). 8337d9c2: author did not reply, but only the option still exists (under a different name), other code was removed. d8fd7131: did not reply. Disabled in a previous commit. 05258251: same author as above. Both fields actually seem to have vanished (even when tracking renames), so no action taken. d459e644, 268b2c1a: author did not reply, but we reuse only the options (with different names and slightly or fully different semantics, and completely different implementations), so I don't think this is relevant for copyright. 09e742fe, 17c39c4e: same as above. e8a173de, bff4b3ee: author could not be reached. The commands were reworked to properties, and the code outside of the TV code were moved back to the TV code. So I don't think copyright applies to the current command.c parts (mp_property_tv_color, mp_property_tv_freq, mp_property_tv_scan). The TV parts remain GPL. 0810e427: could not be reached. Disabled in a previous commit. 43744a2d: unknown author, but this was replaced by dynamic alloc (if the change is even copyrightable). 116ca0c7: unknown author; reasoning see input.c relicensing commit. e7e4d1d8: these semantics still exist, but as generic code, and this code was fully removed. f1175cd9: the author of the cited patch is unknown, and upon inspection it turns out that I was only using the idea to pause the player on EOF, so I claim it's not copyright relevant. 25affdcc: author could not be reached (yet) - but it's only a function rename, not copyrightable. 5728504c was committed by Arpi (who agreed), but hints that it might be by a different author. In fact it seems to be mostly this patch: http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/2001-November/002041.html The author did not respond, but it all seems to have been removed later. It's a terrible mess though. Arpi reverted the A-V sync code at first, but left the RTC code for a while. The following commits remove these changes 100%: 14b35442, 7181a091, 31482783, 614f8475, df58e822. cehoyos did explicitly not agree to LGPL, but was involved in the following changes: c99d8fc8: applied a patch and didn't modify it, the original author agreed. 40ac0d31: author could not be reached, but all code is gone anyway. The "af" command has a similar function, but works completely different and actually reuses a mechanism older than this patch. 54350436: applied a patch, but didn't modify it, except for adding a German translation, which was removed later. a2dda036: same situation as above 240b743e: this was made GPL-only in a previous commit 7b25afd7: same as above (for now) kirijua could not be reached, but was a regular patch contributor: c2c997fd: video equalizer code move; probably not copyrightable. Is GPL due to Nick anyway. be54f481: technically, this became the audio track property later. But all what is left is the fact that you pass a track ID to it, so consider the original coypright non-relevant. 2f376d1b: this was rewritten in b7052b43, but for now we can afford to be careful, so this was marked as GPL only in a previous commit. 43844d09: remaining parts in main.c were reverted in a previous commit. anders has mostly disagreed with the LGPL relicensing. Does not want libaf to become LGPL, but made some concessions. In particular, he granted us permission to relicense 4943e9c52c and 242aa6ebd4. We also consider some of his changes remaining in mpv not relevant for copyright (such as 735de602 - we won't remove the this option completely). We will completely remove his other contributions, including the entire audio filter chain. For now, this stuff is marked as GPL only. The remaining question is how much code in player/audio.c (based on the former mplayer.c and dec_audio.c) is under his copyright. I made claims about this in a previous commit. Nick(ols) Kurshev, svn username "nick" and "nickols_k", could not be reached. He had a lot of changes in early MPlayer. It seems all of that was removed, at least in mpv. His main work, like VIDIX or libswscale work, does not exist in mpv anymore, but the changes to mplayer.c and other core parts still deserve attention: a4119f6b, fb927549, ad3529b8, e11b23dc, 5f2178be, 93c371d5: removed in b43d67e0, d1628d12, 24ed01fe, df58e822. 0a83c6ec, 104c125e, 4e067f62, aec5dcc8, b587a3d6, f3de6e6b: DR, VAA, and "tune" stuff was fully removed later on or replaced with other mechanisms. 340183b0: screenshots were redone later (the VOCTRL was even removed, with an independent implementation using the same VOCTRL a few years later), so not relevant anymore. Basically only the 's' shortcut remains (but not its implementation). 92c5c274, bffd4007, 555c6766: for now marked as GPL only in a previous commit. Might contain some trace amounts of "michael"'s copyright, who agreed to LGPL only once the core is relicensed. This will still be respected, but I don't think it matters at this in this case. (Some code touched by him was merged into mplayer.c, and then disappeared after heavy refactoring.) I tried to be as careful and as complete as possible. It can't be excluded that amends to this will be made later. This does not make the player LGPL yet.
2017-06-23 13:53:41 +00:00
* mpv is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* mpv is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
player: change license of most core files to LGPL These files have all in common that they were fully or mostly taken from mplayer.c. (mplayer.c was a huge file that contains almost all of the playback core, until it was split into multiple parts.) This was probably the hardest part to relicense, because so much code was moved around all the time. player/audio.c still does not compile. We'll have to redo audio filtering. Once that is done, we can probably actually provide an actual LGPL configure switch. Here is a relatively detailed list of potential issues: 8d190244: author did not reply, parts were made GPL-only in a previous commit. 7882ea9b: author could not be reached, but the code is gone. wscript still has --datadir switch, but I don't think this is relevant to copyright. f197efd5: unclear origin, but I consider the code gone anyway (replaced with generic OSD mechanisms). 8337d9c2: author did not reply, but only the option still exists (under a different name), other code was removed. d8fd7131: did not reply. Disabled in a previous commit. 05258251: same author as above. Both fields actually seem to have vanished (even when tracking renames), so no action taken. d459e644, 268b2c1a: author did not reply, but we reuse only the options (with different names and slightly or fully different semantics, and completely different implementations), so I don't think this is relevant for copyright. 09e742fe, 17c39c4e: same as above. e8a173de, bff4b3ee: author could not be reached. The commands were reworked to properties, and the code outside of the TV code were moved back to the TV code. So I don't think copyright applies to the current command.c parts (mp_property_tv_color, mp_property_tv_freq, mp_property_tv_scan). The TV parts remain GPL. 0810e427: could not be reached. Disabled in a previous commit. 43744a2d: unknown author, but this was replaced by dynamic alloc (if the change is even copyrightable). 116ca0c7: unknown author; reasoning see input.c relicensing commit. e7e4d1d8: these semantics still exist, but as generic code, and this code was fully removed. f1175cd9: the author of the cited patch is unknown, and upon inspection it turns out that I was only using the idea to pause the player on EOF, so I claim it's not copyright relevant. 25affdcc: author could not be reached (yet) - but it's only a function rename, not copyrightable. 5728504c was committed by Arpi (who agreed), but hints that it might be by a different author. In fact it seems to be mostly this patch: http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/2001-November/002041.html The author did not respond, but it all seems to have been removed later. It's a terrible mess though. Arpi reverted the A-V sync code at first, but left the RTC code for a while. The following commits remove these changes 100%: 14b35442, 7181a091, 31482783, 614f8475, df58e822. cehoyos did explicitly not agree to LGPL, but was involved in the following changes: c99d8fc8: applied a patch and didn't modify it, the original author agreed. 40ac0d31: author could not be reached, but all code is gone anyway. The "af" command has a similar function, but works completely different and actually reuses a mechanism older than this patch. 54350436: applied a patch, but didn't modify it, except for adding a German translation, which was removed later. a2dda036: same situation as above 240b743e: this was made GPL-only in a previous commit 7b25afd7: same as above (for now) kirijua could not be reached, but was a regular patch contributor: c2c997fd: video equalizer code move; probably not copyrightable. Is GPL due to Nick anyway. be54f481: technically, this became the audio track property later. But all what is left is the fact that you pass a track ID to it, so consider the original coypright non-relevant. 2f376d1b: this was rewritten in b7052b43, but for now we can afford to be careful, so this was marked as GPL only in a previous commit. 43844d09: remaining parts in main.c were reverted in a previous commit. anders has mostly disagreed with the LGPL relicensing. Does not want libaf to become LGPL, but made some concessions. In particular, he granted us permission to relicense 4943e9c52c and 242aa6ebd4. We also consider some of his changes remaining in mpv not relevant for copyright (such as 735de602 - we won't remove the this option completely). We will completely remove his other contributions, including the entire audio filter chain. For now, this stuff is marked as GPL only. The remaining question is how much code in player/audio.c (based on the former mplayer.c and dec_audio.c) is under his copyright. I made claims about this in a previous commit. Nick(ols) Kurshev, svn username "nick" and "nickols_k", could not be reached. He had a lot of changes in early MPlayer. It seems all of that was removed, at least in mpv. His main work, like VIDIX or libswscale work, does not exist in mpv anymore, but the changes to mplayer.c and other core parts still deserve attention: a4119f6b, fb927549, ad3529b8, e11b23dc, 5f2178be, 93c371d5: removed in b43d67e0, d1628d12, 24ed01fe, df58e822. 0a83c6ec, 104c125e, 4e067f62, aec5dcc8, b587a3d6, f3de6e6b: DR, VAA, and "tune" stuff was fully removed later on or replaced with other mechanisms. 340183b0: screenshots were redone later (the VOCTRL was even removed, with an independent implementation using the same VOCTRL a few years later), so not relevant anymore. Basically only the 's' shortcut remains (but not its implementation). 92c5c274, bffd4007, 555c6766: for now marked as GPL only in a previous commit. Might contain some trace amounts of "michael"'s copyright, who agreed to LGPL only once the core is relicensed. This will still be respected, but I don't think it matters at this in this case. (Some code touched by him was merged into mplayer.c, and then disappeared after heavy refactoring.) I tried to be as careful and as complete as possible. It can't be excluded that amends to this will be made later. This does not make the player LGPL yet.
2017-06-23 13:53:41 +00:00
* GNU Lesser General Public License for more details.
*
player: change license of most core files to LGPL These files have all in common that they were fully or mostly taken from mplayer.c. (mplayer.c was a huge file that contains almost all of the playback core, until it was split into multiple parts.) This was probably the hardest part to relicense, because so much code was moved around all the time. player/audio.c still does not compile. We'll have to redo audio filtering. Once that is done, we can probably actually provide an actual LGPL configure switch. Here is a relatively detailed list of potential issues: 8d190244: author did not reply, parts were made GPL-only in a previous commit. 7882ea9b: author could not be reached, but the code is gone. wscript still has --datadir switch, but I don't think this is relevant to copyright. f197efd5: unclear origin, but I consider the code gone anyway (replaced with generic OSD mechanisms). 8337d9c2: author did not reply, but only the option still exists (under a different name), other code was removed. d8fd7131: did not reply. Disabled in a previous commit. 05258251: same author as above. Both fields actually seem to have vanished (even when tracking renames), so no action taken. d459e644, 268b2c1a: author did not reply, but we reuse only the options (with different names and slightly or fully different semantics, and completely different implementations), so I don't think this is relevant for copyright. 09e742fe, 17c39c4e: same as above. e8a173de, bff4b3ee: author could not be reached. The commands were reworked to properties, and the code outside of the TV code were moved back to the TV code. So I don't think copyright applies to the current command.c parts (mp_property_tv_color, mp_property_tv_freq, mp_property_tv_scan). The TV parts remain GPL. 0810e427: could not be reached. Disabled in a previous commit. 43744a2d: unknown author, but this was replaced by dynamic alloc (if the change is even copyrightable). 116ca0c7: unknown author; reasoning see input.c relicensing commit. e7e4d1d8: these semantics still exist, but as generic code, and this code was fully removed. f1175cd9: the author of the cited patch is unknown, and upon inspection it turns out that I was only using the idea to pause the player on EOF, so I claim it's not copyright relevant. 25affdcc: author could not be reached (yet) - but it's only a function rename, not copyrightable. 5728504c was committed by Arpi (who agreed), but hints that it might be by a different author. In fact it seems to be mostly this patch: http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/2001-November/002041.html The author did not respond, but it all seems to have been removed later. It's a terrible mess though. Arpi reverted the A-V sync code at first, but left the RTC code for a while. The following commits remove these changes 100%: 14b35442, 7181a091, 31482783, 614f8475, df58e822. cehoyos did explicitly not agree to LGPL, but was involved in the following changes: c99d8fc8: applied a patch and didn't modify it, the original author agreed. 40ac0d31: author could not be reached, but all code is gone anyway. The "af" command has a similar function, but works completely different and actually reuses a mechanism older than this patch. 54350436: applied a patch, but didn't modify it, except for adding a German translation, which was removed later. a2dda036: same situation as above 240b743e: this was made GPL-only in a previous commit 7b25afd7: same as above (for now) kirijua could not be reached, but was a regular patch contributor: c2c997fd: video equalizer code move; probably not copyrightable. Is GPL due to Nick anyway. be54f481: technically, this became the audio track property later. But all what is left is the fact that you pass a track ID to it, so consider the original coypright non-relevant. 2f376d1b: this was rewritten in b7052b43, but for now we can afford to be careful, so this was marked as GPL only in a previous commit. 43844d09: remaining parts in main.c were reverted in a previous commit. anders has mostly disagreed with the LGPL relicensing. Does not want libaf to become LGPL, but made some concessions. In particular, he granted us permission to relicense 4943e9c52c and 242aa6ebd4. We also consider some of his changes remaining in mpv not relevant for copyright (such as 735de602 - we won't remove the this option completely). We will completely remove his other contributions, including the entire audio filter chain. For now, this stuff is marked as GPL only. The remaining question is how much code in player/audio.c (based on the former mplayer.c and dec_audio.c) is under his copyright. I made claims about this in a previous commit. Nick(ols) Kurshev, svn username "nick" and "nickols_k", could not be reached. He had a lot of changes in early MPlayer. It seems all of that was removed, at least in mpv. His main work, like VIDIX or libswscale work, does not exist in mpv anymore, but the changes to mplayer.c and other core parts still deserve attention: a4119f6b, fb927549, ad3529b8, e11b23dc, 5f2178be, 93c371d5: removed in b43d67e0, d1628d12, 24ed01fe, df58e822. 0a83c6ec, 104c125e, 4e067f62, aec5dcc8, b587a3d6, f3de6e6b: DR, VAA, and "tune" stuff was fully removed later on or replaced with other mechanisms. 340183b0: screenshots were redone later (the VOCTRL was even removed, with an independent implementation using the same VOCTRL a few years later), so not relevant anymore. Basically only the 's' shortcut remains (but not its implementation). 92c5c274, bffd4007, 555c6766: for now marked as GPL only in a previous commit. Might contain some trace amounts of "michael"'s copyright, who agreed to LGPL only once the core is relicensed. This will still be respected, but I don't think it matters at this in this case. (Some code touched by him was merged into mplayer.c, and then disappeared after heavy refactoring.) I tried to be as careful and as complete as possible. It can't be excluded that amends to this will be made later. This does not make the player LGPL yet.
2017-06-23 13:53:41 +00:00
* You should have received a copy of the GNU Lesser General Public
* License along with mpv. If not, see <http://www.gnu.org/licenses/>.
*/
#include <stddef.h>
#include <stdbool.h>
#include <inttypes.h>
#include <limits.h>
#include <math.h>
#include <assert.h>
#include "config.h"
#include "mpv_talloc.h"
#include "common/msg.h"
#include "common/encode.h"
#include "options/options.h"
#include "common/common.h"
#include "osdep/timer.h"
#include "audio/format.h"
#include "audio/out/ao.h"
#include "demux/demux.h"
#include "filters/f_async_queue.h"
#include "filters/f_decoder_wrapper.h"
#include "filters/filter_internal.h"
#include "core.h"
#include "command.h"
enum {
AD_OK = 0,
AD_EOF = -2,
AD_WAIT = -4,
};
static void ao_process(struct mp_filter *f);
static void update_speed_filters(struct MPContext *mpctx)
{
struct ao_chain *ao_c = mpctx->ao_chain;
if (!ao_c)
return;
double speed = mpctx->opts->playback_speed;
double resample = mpctx->speed_factor_a;
double drop = 1.0;
if (!mpctx->opts->pitch_correction) {
resample *= speed;
speed = 1.0;
}
if (mpctx->display_sync_active && mpctx->opts->video_sync == VS_DISP_ADROP) {
drop *= speed * resample;
resample = speed = 1.0;
}
mp_output_chain_set_audio_speed(ao_c->filter, speed, resample, drop);
}
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
static int recreate_audio_filters(struct MPContext *mpctx)
{
struct ao_chain *ao_c = mpctx->ao_chain;
assert(ao_c);
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
if (!mp_output_chain_update_filters(ao_c->filter, mpctx->opts->af_settings))
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
goto fail;
update_speed_filters(mpctx);
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
mp_notify(mpctx, MPV_EVENT_AUDIO_RECONFIG, NULL);
return 0;
fail:
MP_ERR(mpctx, "Audio filter initialized failed!\n");
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
return -1;
}
int reinit_audio_filters(struct MPContext *mpctx)
{
struct ao_chain *ao_c = mpctx->ao_chain;
if (!ao_c)
return 0;
double delay = mp_output_get_measured_total_delay(ao_c->filter);
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
if (recreate_audio_filters(mpctx) < 0)
return -1;
double ndelay = mp_output_get_measured_total_delay(ao_c->filter);
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
// Only force refresh if the amount of dropped buffered data is going to
// cause "issues" for the A/V sync logic.
if (mpctx->audio_status == STATUS_PLAYING && delay - ndelay >= 0.2)
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
issue_refresh_seek(mpctx, MPSEEK_EXACT);
return 1;
}
static double db_gain(double db)
{
return pow(10.0, db/20.0);
}
static float compute_replaygain(struct MPContext *mpctx)
{
struct MPOpts *opts = mpctx->opts;
float rgain = 1.0;
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
struct replaygain_data *rg = NULL;
struct track *track = mpctx->current_track[0][STREAM_AUDIO];
if (track)
rg = track->stream->codec->replaygain_data;
if (opts->rgain_mode && rg) {
MP_VERBOSE(mpctx, "Replaygain: Track=%f/%f Album=%f/%f\n",
rg->track_gain, rg->track_peak,
rg->album_gain, rg->album_peak);
float gain, peak;
if (opts->rgain_mode == 1) {
gain = rg->track_gain;
peak = rg->track_peak;
} else {
gain = rg->album_gain;
peak = rg->album_peak;
}
gain += opts->rgain_preamp;
rgain = db_gain(gain);
MP_VERBOSE(mpctx, "Applying replay-gain: %f\n", rgain);
if (!opts->rgain_clip) { // clipping prevention
rgain = MPMIN(rgain, 1.0 / peak);
MP_VERBOSE(mpctx, "...with clipping prevention: %f\n", rgain);
}
} else if (opts->rgain_fallback) {
rgain = db_gain(opts->rgain_fallback);
MP_VERBOSE(mpctx, "Applying fallback gain: %f\n", rgain);
}
return rgain;
}
// Called when opts->softvol_volume or opts->softvol_mute were changed.
void audio_update_volume(struct MPContext *mpctx)
{
struct MPOpts *opts = mpctx->opts;
struct ao_chain *ao_c = mpctx->ao_chain;
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
if (!ao_c || !ao_c->ao)
return;
float gain = MPMAX(opts->softvol_volume / 100.0, 0);
gain = pow(gain, 3);
gain *= compute_replaygain(mpctx);
if (opts->softvol_mute == 1)
gain = 0.0;
audio: add audio softvol processing to AO This does what af_volume used to do. Since we couldn't relicense it, just rewrite it. Since we don't have a new filter mechanism yet, and the libavfilter is too inconvenient, do applying the volume gain in ao.c directly. This is done before handling the audio data to the driver. Since push.c runs a separate thread, and pull.c is called asynchronously from the audio driver's thread, the volume value needs to be synchronized. There's no existing central mutex, so do some shit with atomics. Since there's no atomic_float type predefined (which is at least needed when using the legacy wrapper), do some nonsense about reinterpret casting the float value to an int for the purpose of atomic access. Not sure if using memcpy() is undefined behavior, but for now I don't care. The advantage of not using a filter is lower complexity (no filter auto insertion), and lower latency (gain processing is done after our internal audio buffer of at least 200ms). Disavdantages include inability to use native volume control _before_ other filters with custom filter chains, and the need to add new processing for each new sample type. Since this doesn't reuse any of the old GPL code, nor does indirectly rely on it, volume and replaygain handling now works in LGPL mode. How to process the gain is inspired by libavfilter's af_volume (LGPL). In particular, we use exactly the same rounding, and we quantize processing for integer sample types by 256 steps. Some of libavfilter's copyright may or may not apply, but I think not, and it's the same license anyway.
2017-11-29 20:30:10 +00:00
ao_set_gain(ao_c->ao, gain);
}
// Call this if opts->playback_speed or mpctx->speed_factor_* change.
void update_playback_speed(struct MPContext *mpctx)
{
mpctx->audio_speed = mpctx->opts->playback_speed * mpctx->speed_factor_a;
mpctx->video_speed = mpctx->opts->playback_speed * mpctx->speed_factor_v;
update_speed_filters(mpctx);
}
static void ao_chain_reset_state(struct ao_chain *ao_c)
{
ao_c->last_out_pts = MP_NOPTS_VALUE;
ao_c->out_eof = false;
player: partially rework --cache-pause The --cache-pause feature (enabled by default) will pause playback for a while if network runs out of data. If this is not done, then playback will go on frame-wise (as packets are slowly read from the network and then instantly decoded and displayed). This feature is actually useless, as you won't get nice playback no matter what if network is too slow, but I guess I still prefer this behavior for some reason. This commit changes this behavior from using the demuxer cache state only, to trying to use underrun information from the AO/VO. This means if you have a very large audio buffer, then cache-pausing will trigger once that buffer is depleted, which will be some time _after_ the demuxer cache has run out. This requires explicit support from the AO. Otherwise, the behavior should be mostly the same as before this commit. This does not care about the AO buffer. In theory, the AO may underrun, then the player will write some data to the AO buffer, then the AO will recover and play this bit of data, then the player will probably trigger the cache-pause behavior. The probability of this happening should be pretty low, so I will hold off fixing this until the next refactor of the AO chain (if ever). The VO underflow detection was devised and tested in 5 minutes, and may not be correct. At least I'm fairly sure that the combination of all the factors should make incorrect behavior relatively unlikely, but problems are possible. Also, the demux_reader_state.underrun field may be inaccurate. It's only the present state at the time demux_get_reader_state() was called, and may exclude past underruns. In theory, this could cause "close" cases to be missed. Then you might get an audio underrun without cache-pausing acting on it. If the stars align, this could happen multiple times in the row, effectively making this feature not work. The most user-visible consequence of this change is that the user will now see an AO underrun warning every time the cache runs out. Maybe this cache-pause feature should just be removed...
2019-10-11 17:34:04 +00:00
ao_c->underrun = false;
ao_c->start_pts_known = false;
ao_c->start_pts = MP_NOPTS_VALUE;
ao_c->untimed_throttle = false;
ao_c->underrun = false;
}
void reset_audio_state(struct MPContext *mpctx)
{
if (mpctx->ao_chain) {
ao_chain_reset_state(mpctx->ao_chain);
struct track *t = mpctx->ao_chain->track;
if (t && t->dec)
mp_decoder_wrapper_set_play_dir(t->dec, mpctx->play_dir);
}
mpctx->audio_status = mpctx->ao_chain ? STATUS_SYNCING : STATUS_EOF;
mpctx->delay = 0;
mpctx->logged_async_diff = -1;
}
void uninit_audio_out(struct MPContext *mpctx)
{
struct ao_chain *ao_c = mpctx->ao_chain;
if (ao_c) {
ao_c->ao_queue = NULL;
TA_FREEP(&ao_c->queue_filter);
ao_c->ao = NULL;
}
if (mpctx->ao) {
// Note: with gapless_audio, stop_play is not correctly set
if ((mpctx->opts->gapless_audio || mpctx->stop_play == AT_END_OF_FILE) &&
ao_is_playing(mpctx->ao) && !get_internal_paused(mpctx))
{
MP_VERBOSE(mpctx, "draining left over audio\n");
ao_drain(mpctx->ao);
}
ao_uninit(mpctx->ao);
mp_notify(mpctx, MPV_EVENT_AUDIO_RECONFIG, NULL);
}
mpctx->ao = NULL;
TA_FREEP(&mpctx->ao_filter_fmt);
}
static void ao_chain_uninit(struct ao_chain *ao_c)
{
struct track *track = ao_c->track;
if (track) {
assert(track->ao_c == ao_c);
track->ao_c = NULL;
if (ao_c->dec_src)
assert(track->dec->f->pins[0] == ao_c->dec_src);
talloc_free(track->dec->f);
track->dec = NULL;
}
if (ao_c->filter_src)
mp_pin_disconnect(ao_c->filter_src);
talloc_free(ao_c->filter->f);
talloc_free(ao_c->ao_filter);
talloc_free(ao_c);
}
void uninit_audio_chain(struct MPContext *mpctx)
{
if (mpctx->ao_chain) {
ao_chain_uninit(mpctx->ao_chain);
mpctx->ao_chain = NULL;
mpctx->audio_status = STATUS_EOF;
mp_notify(mpctx, MPV_EVENT_AUDIO_RECONFIG, NULL);
}
}
static char *audio_config_to_str_buf(char *buf, size_t buf_sz, int rate,
int format, struct mp_chmap channels)
audio: introduce a new type to hold audio frames This is pretty pointless, but I believe it allows us to claim that the new code is not affected by the copyright of the old code. This is needed, because the original mp_audio struct was written by someone who has disagreed with LGPL relicensing (it was called af_data at the time, and was defined in af.h). The "GPL'ed" struct contents that surive are pretty trivial: just the data pointer, and some metadata like the format, samplerate, etc. - but at least in this case, any new code would be extremely similar anyway, and I'm not really sure whether it's OK to claim different copyright. So what we do is we just use AVFrame (which of course is LGPL with 100% certainty), and add some accessors around it to adapt it to mpv conventions. Also, this gets rid of some annoying conventions of mp_audio, like the struct fields that require using an accessor to write to them anyway. For the most part, this change is only dumb replacements of mp_audio related functions and fields. One minor actual change is that you can't allocate the new type on the stack anymore. Some code still uses mp_audio. All audio filter code will be deleted, so it makes no sense to convert this code. (Audio filters which are LGPL and which we keep will have to be ported to a new filter infrastructure anyway.) player/audio.c uses it because it interacts with the old filter code. push.c has some complex use of mp_audio and mp_audio_buffer, but this and pull.c will most likely be rewritten to do something else.
2017-08-16 19:00:20 +00:00
{
char ch[128];
mp_chmap_to_str_buf(ch, sizeof(ch), &channels);
char *hr_ch = mp_chmap_to_str_hr(&channels);
if (strcmp(hr_ch, ch) != 0)
mp_snprintf_cat(ch, sizeof(ch), " (%s)", hr_ch);
snprintf(buf, buf_sz, "%dHz %s %dch %s", rate,
ch, channels.num, af_fmt_to_str(format));
return buf;
audio: introduce a new type to hold audio frames This is pretty pointless, but I believe it allows us to claim that the new code is not affected by the copyright of the old code. This is needed, because the original mp_audio struct was written by someone who has disagreed with LGPL relicensing (it was called af_data at the time, and was defined in af.h). The "GPL'ed" struct contents that surive are pretty trivial: just the data pointer, and some metadata like the format, samplerate, etc. - but at least in this case, any new code would be extremely similar anyway, and I'm not really sure whether it's OK to claim different copyright. So what we do is we just use AVFrame (which of course is LGPL with 100% certainty), and add some accessors around it to adapt it to mpv conventions. Also, this gets rid of some annoying conventions of mp_audio, like the struct fields that require using an accessor to write to them anyway. For the most part, this change is only dumb replacements of mp_audio related functions and fields. One minor actual change is that you can't allocate the new type on the stack anymore. Some code still uses mp_audio. All audio filter code will be deleted, so it makes no sense to convert this code. (Audio filters which are LGPL and which we keep will have to be ported to a new filter infrastructure anyway.) player/audio.c uses it because it interacts with the old filter code. push.c has some complex use of mp_audio and mp_audio_buffer, but this and pull.c will most likely be rewritten to do something else.
2017-08-16 19:00:20 +00:00
}
// Decide whether on a format change, we should reinit the AO.
static bool keep_weak_gapless_format(struct mp_aframe *old, struct mp_aframe* new)
{
bool res = false;
struct mp_aframe *new_mod = mp_aframe_new_ref(new);
if (!new_mod)
abort();
// If the sample formats are compatible (== libswresample generally can
// convert them), keep the AO. On other changes, recreate it.
int old_fmt = mp_aframe_get_format(old);
int new_fmt = mp_aframe_get_format(new);
if (af_format_conversion_score(old_fmt, new_fmt) == INT_MIN)
goto done; // completely incompatible formats
if (!mp_aframe_set_format(new_mod, old_fmt))
goto done;
res = mp_aframe_config_equals(old, new_mod);
done:
talloc_free(new_mod);
return res;
}
static void ao_chain_set_ao(struct ao_chain *ao_c, struct ao *ao)
{
if (ao_c->ao != ao) {
assert(!ao_c->ao);
ao_c->ao = ao;
ao_c->ao_queue = ao_get_queue(ao_c->ao);
ao_c->queue_filter = mp_async_queue_create_filter(ao_c->ao_filter,
MP_PIN_IN, ao_c->ao_queue);
mp_async_queue_set_notifier(ao_c->queue_filter, ao_c->ao_filter);
// Make sure filtering never stops with frames stuck in access filter.
mp_filter_set_high_priority(ao_c->queue_filter, true);
audio_update_volume(ao_c->mpctx);
}
if (ao_c->filter->ao_needs_update)
mp_output_chain_set_ao(ao_c->filter, ao_c->ao);
mp_filter_wakeup(ao_c->ao_filter);
}
static int reinit_audio_filters_and_output(struct MPContext *mpctx)
{
struct MPOpts *opts = mpctx->opts;
struct ao_chain *ao_c = mpctx->ao_chain;
assert(ao_c);
struct track *track = ao_c->track;
assert(ao_c->filter->ao_needs_update);
// The "ideal" filter output format
struct mp_aframe *out_fmt = mp_aframe_new_ref(ao_c->filter->output_aformat);
if (!out_fmt)
abort();
if (!mp_aframe_config_is_valid(out_fmt)) {
talloc_free(out_fmt);
goto init_error;
}
if (af_fmt_is_pcm(mp_aframe_get_format(out_fmt))) {
if (opts->force_srate)
mp_aframe_set_rate(out_fmt, opts->force_srate);
if (opts->audio_output_format)
mp_aframe_set_format(out_fmt, opts->audio_output_format);
if (opts->audio_output_channels.num_chmaps == 1)
mp_aframe_set_chmap(out_fmt, &opts->audio_output_channels.chmaps[0]);
}
// Weak gapless audio: if the filter output format is the same as the
// previous one, keep the AO and don't reinit anything.
// Strong gapless: always keep the AO
if ((mpctx->ao_filter_fmt && mpctx->ao && opts->gapless_audio < 0 &&
keep_weak_gapless_format(mpctx->ao_filter_fmt, out_fmt)) ||
(mpctx->ao && opts->gapless_audio > 0))
{
ao_chain_set_ao(ao_c, mpctx->ao);
talloc_free(out_fmt);
return 0;
}
// Wait until all played.
if (mpctx->ao && ao_is_playing(mpctx->ao)) {
talloc_free(out_fmt);
return 0;
}
// Format change during syncing. Force playback start early, then wait.
if (ao_c->ao_queue && mp_async_queue_get_frames(ao_c->ao_queue) &&
mpctx->audio_status == STATUS_SYNCING)
{
mpctx->audio_status = STATUS_READY;
mp_wakeup_core(mpctx);
talloc_free(out_fmt);
return 0;
}
if (mpctx->audio_status == STATUS_READY) {
talloc_free(out_fmt);
return 0;
}
uninit_audio_out(mpctx);
int out_rate = mp_aframe_get_rate(out_fmt);
int out_format = mp_aframe_get_format(out_fmt);
struct mp_chmap out_channels = {0};
mp_aframe_get_chmap(out_fmt, &out_channels);
int ao_flags = 0;
bool spdif_fallback = af_fmt_is_spdif(out_format) &&
ao_c->spdif_passthrough;
if (opts->ao_null_fallback && !spdif_fallback)
ao_flags |= AO_INIT_NULL_FALLBACK;
if (opts->audio_stream_silence)
ao_flags |= AO_INIT_STREAM_SILENCE;
if (opts->audio_exclusive)
ao_flags |= AO_INIT_EXCLUSIVE;
if (af_fmt_is_pcm(out_format)) {
if (!opts->audio_output_channels.set ||
opts->audio_output_channels.auto_safe)
ao_flags |= AO_INIT_SAFE_MULTICHANNEL_ONLY;
mp_chmap_sel_list(&out_channels,
opts->audio_output_channels.chmaps,
opts->audio_output_channels.num_chmaps);
}
mpctx->ao_filter_fmt = out_fmt;
mpctx->ao = ao_init_best(mpctx->global, ao_flags, mp_wakeup_core_cb,
mpctx, mpctx->encode_lavc_ctx, out_rate,
out_format, out_channels);
int ao_rate = 0;
int ao_format = 0;
struct mp_chmap ao_channels = {0};
if (mpctx->ao)
ao_get_format(mpctx->ao, &ao_rate, &ao_format, &ao_channels);
// Verify passthrough format was not changed.
if (mpctx->ao && af_fmt_is_spdif(out_format)) {
if (out_rate != ao_rate || out_format != ao_format ||
!mp_chmap_equals(&out_channels, &ao_channels))
{
MP_ERR(mpctx, "Passthrough format unsupported.\n");
ao_uninit(mpctx->ao);
mpctx->ao = NULL;
}
}
if (!mpctx->ao) {
// If spdif was used, try to fallback to PCM.
if (spdif_fallback && ao_c->track && ao_c->track->dec) {
MP_VERBOSE(mpctx, "Falling back to PCM output.\n");
ao_c->spdif_passthrough = false;
ao_c->spdif_failed = true;
mp_decoder_wrapper_set_spdif_flag(ao_c->track->dec, false);
if (!mp_decoder_wrapper_reinit(ao_c->track->dec))
goto init_error;
reset_audio_state(mpctx);
mp_output_chain_reset_harder(ao_c->filter);
mp_wakeup_core(mpctx); // reinit with new format next time
return 0;
}
MP_ERR(mpctx, "Could not open/initialize audio device -> no sound.\n");
mpctx->error_playing = MPV_ERROR_AO_INIT_FAILED;
goto init_error;
}
char tmp[192];
MP_INFO(mpctx, "AO: [%s] %s\n", ao_get_name(mpctx->ao),
audio_config_to_str_buf(tmp, sizeof(tmp), ao_rate, ao_format,
ao_channels));
MP_VERBOSE(mpctx, "AO: Description: %s\n", ao_get_description(mpctx->ao));
update_window_title(mpctx, true);
ao_c->ao_resume_time =
opts->audio_wait_open > 0 ? mp_time_sec() + opts->audio_wait_open : 0;
ao_set_paused(mpctx->ao, get_internal_paused(mpctx));
ao_chain_set_ao(ao_c, mpctx->ao);
audio_update_volume(mpctx);
// Almost nonsensical hack to deal with certain format change scenarios.
if (mpctx->audio_status == STATUS_PLAYING)
ao_start(mpctx->ao);
mp_wakeup_core(mpctx);
mp_notify(mpctx, MPV_EVENT_AUDIO_RECONFIG, NULL);
return 0;
init_error:
uninit_audio_chain(mpctx);
uninit_audio_out(mpctx);
error_on_track(mpctx, track);
return -1;
}
int init_audio_decoder(struct MPContext *mpctx, struct track *track)
{
assert(!track->dec);
if (!track->stream)
goto init_error;
track->dec = mp_decoder_wrapper_create(mpctx->filter_root, track->stream);
if (!track->dec)
goto init_error;
if (track->ao_c)
mp_decoder_wrapper_set_spdif_flag(track->dec, true);
if (!mp_decoder_wrapper_reinit(track->dec))
goto init_error;
return 1;
init_error:
if (track->sink)
mp_pin_disconnect(track->sink);
track->sink = NULL;
error_on_track(mpctx, track);
return 0;
}
void reinit_audio_chain(struct MPContext *mpctx)
{
struct track *track = NULL;
track = mpctx->current_track[0][STREAM_AUDIO];
if (!track || !track->stream) {
if (!mpctx->encode_lavc_ctx)
uninit_audio_out(mpctx);
error_on_track(mpctx, track);
return;
}
reinit_audio_chain_src(mpctx, track);
}
static const struct mp_filter_info ao_filter = {
.name = "ao",
.process = ao_process,
};
// (track=NULL creates a blank chain, used for lavfi-complex)
void reinit_audio_chain_src(struct MPContext *mpctx, struct track *track)
{
assert(!mpctx->ao_chain);
mp_notify(mpctx, MPV_EVENT_AUDIO_RECONFIG, NULL);
struct ao_chain *ao_c = talloc_zero(NULL, struct ao_chain);
mpctx->ao_chain = ao_c;
ao_c->mpctx = mpctx;
ao_c->log = mpctx->log;
ao_c->filter =
mp_output_chain_create(mpctx->filter_root, MP_OUTPUT_CHAIN_AUDIO);
ao_c->spdif_passthrough = true;
ao_c->last_out_pts = MP_NOPTS_VALUE;
ao_c->delay = mpctx->opts->audio_delay;
ao_c->ao_filter = mp_filter_create(mpctx->filter_root, &ao_filter);
if (!ao_c->filter || !ao_c->ao_filter)
goto init_error;
ao_c->ao_filter->priv = ao_c;
mp_filter_add_pin(ao_c->ao_filter, MP_PIN_IN, "in");
mp_pin_connect(ao_c->ao_filter->pins[0], ao_c->filter->f->pins[1]);
if (track) {
ao_c->track = track;
track->ao_c = ao_c;
if (!init_audio_decoder(mpctx, track))
goto init_error;
ao_c->dec_src = track->dec->f->pins[0];
mp_pin_connect(ao_c->filter->f->pins[0], ao_c->dec_src);
}
reset_audio_state(mpctx);
if (recreate_audio_filters(mpctx) < 0)
goto init_error;
if (mpctx->ao)
audio_update_volume(mpctx);
mp_wakeup_core(mpctx);
return;
init_error:
uninit_audio_chain(mpctx);
uninit_audio_out(mpctx);
error_on_track(mpctx, track);
}
// Return pts value corresponding to the start point of audio written to the
// ao queue so far.
double written_audio_pts(struct MPContext *mpctx)
{
return mpctx->ao_chain ? mpctx->ao_chain->last_out_pts : MP_NOPTS_VALUE;
}
// Return pts value corresponding to currently playing audio.
double playing_audio_pts(struct MPContext *mpctx)
{
double pts = written_audio_pts(mpctx);
if (pts == MP_NOPTS_VALUE || !mpctx->ao)
return pts;
return pts - mpctx->audio_speed * ao_get_delay(mpctx->ao);
}
// This garbage is needed for untimed AOs. These consume audio infinitely fast,
// so try keeping approximate A/V sync by blocking audio transfer as needed.
static void update_throttle(struct MPContext *mpctx)
{
struct ao_chain *ao_c = mpctx->ao_chain;
bool new_throttle = mpctx->audio_status == STATUS_PLAYING &&
mpctx->delay > 0 && ao_c && ao_c->ao &&
ao_untimed(ao_c->ao) &&
mpctx->video_status != STATUS_EOF;
if (ao_c && new_throttle != ao_c->untimed_throttle) {
ao_c->untimed_throttle = new_throttle;
mp_wakeup_core(mpctx);
mp_filter_wakeup(ao_c->ao_filter);
}
}
static void ao_process(struct mp_filter *f)
{
struct ao_chain *ao_c = f->priv;
struct MPContext *mpctx = ao_c->mpctx;
if (!ao_c->queue_filter) {
// This will eventually lead to the creation of the AO + queue, due
// to how f_output_chain and AO management works.
mp_pin_out_request_data(f->ppins[0]);
// Check for EOF with no data case, which is a mess because everything
// hates us.
struct mp_frame frame = mp_pin_out_read(f->ppins[0]);
if (frame.type == MP_FRAME_EOF) {
MP_VERBOSE(mpctx, "got EOF with no data before it\n");
ao_c->out_eof = true;
mpctx->audio_status = STATUS_DRAINING;
mp_wakeup_core(mpctx);
} else if (frame.type) {
mp_pin_out_unread(f->ppins[0], frame);
}
return;
}
// Due to mp_async_queue_set_notifier() thhis function is called when the
// queue becomes full. This affects state changes in the normal playloop,
// so wake it up. But avoid redundant wakeups during normal playback.
if (mpctx->audio_status != STATUS_PLAYING &&
mp_async_queue_is_full(ao_c->ao_queue))
mp_wakeup_core(mpctx);
2019-11-17 00:17:34 +00:00
if (mpctx->audio_status == STATUS_SYNCING && !ao_c->start_pts_known)
return;
if (ao_c->untimed_throttle)
return;
if (!mp_pin_can_transfer_data(ao_c->queue_filter->pins[0], f->ppins[0]))
return;
struct mp_frame frame = mp_pin_out_read(f->ppins[0]);
if (frame.type == MP_FRAME_AUDIO) {
struct mp_aframe *af = frame.data;
double endpts = get_play_end_pts(mpctx);
if (endpts != MP_NOPTS_VALUE) {
endpts *= mpctx->play_dir;
// Avoid decoding and discarding the entire rest of the file.
if (mp_aframe_get_pts(af) >= endpts) {
mp_pin_out_unread(f->ppins[0], frame);
if (!ao_c->out_eof) {
ao_c->out_eof = true;
mp_pin_in_write(ao_c->queue_filter->pins[0], MP_EOF_FRAME);
}
return;
}
}
double startpts = mpctx->audio_status == STATUS_SYNCING ?
ao_c->start_pts : MP_NOPTS_VALUE;
mp_aframe_clip_timestamps(af, startpts, endpts);
int samples = mp_aframe_get_size(af);
if (!samples) {
mp_filter_internal_mark_progress(f);
mp_frame_unref(&frame);
return;
}
ao_c->out_eof = false;
if (mpctx->audio_status == STATUS_DRAINING ||
mpctx->audio_status == STATUS_EOF)
{
// If a new frame comes decoder/filter EOF, we should preferably
// call get_sync_pts() again, which (at least in obscure situations)
// may require us to wait a while until the sync PTS is known. Our
// code sucks and can't deal with that, so jump through a hoop to
// get things done in the correct order.
mp_pin_out_unread(f->ppins[0], frame);
ao_c->start_pts_known = false;
mpctx->audio_status = STATUS_SYNCING;
mp_wakeup_core(mpctx);
MP_VERBOSE(mpctx, "new audio frame after EOF\n");
return;
}
mpctx->shown_aframes += samples;
double real_samplerate = mp_aframe_get_rate(af) / mpctx->audio_speed;
mpctx->delay += samples / real_samplerate;
ao_c->last_out_pts = mp_aframe_end_pts(af);
update_throttle(mpctx);
// Gapless case: the AO is still playing from previous file. It makes
// no sense to wait, and in fact the "full queue" event we're waiting
// for may never happen, so start immediately.
// If the new audio starts "later" (big video sync offset), transfer
// of data is stopped somewhere else.
if (mpctx->audio_status == STATUS_SYNCING && ao_is_playing(ao_c->ao)) {
mpctx->audio_status = STATUS_READY;
mp_wakeup_core(mpctx);
MP_VERBOSE(mpctx, "previous audio still playing; continuing\n");
}
mp_pin_in_write(ao_c->queue_filter->pins[0], frame);
} else if (frame.type == MP_FRAME_EOF) {
MP_VERBOSE(mpctx, "audio filter EOF\n");
ao_c->out_eof = true;
mp_wakeup_core(mpctx);
mp_pin_in_write(ao_c->queue_filter->pins[0], frame);
mp_filter_internal_mark_progress(f);
} else {
mp_frame_unref(&frame);
}
}
void reload_audio_output(struct MPContext *mpctx)
{
if (!mpctx->ao)
return;
ao_reset(mpctx->ao);
uninit_audio_out(mpctx);
reinit_audio_filters(mpctx); // mostly to issue refresh seek
struct ao_chain *ao_c = mpctx->ao_chain;
if (ao_c) {
reset_audio_state(mpctx);
mp_output_chain_reset_harder(ao_c->filter);
}
// Whether we can use spdif might have changed. If we failed to use spdif
// in the previous initialization, try it with spdif again (we'll fallback
// to PCM again if necessary).
if (ao_c && ao_c->track) {
struct mp_decoder_wrapper *dec = ao_c->track->dec;
if (dec && ao_c->spdif_failed) {
ao_c->spdif_passthrough = true;
ao_c->spdif_failed = false;
mp_decoder_wrapper_set_spdif_flag(ao_c->track->dec, true);
if (!mp_decoder_wrapper_reinit(dec)) {
MP_ERR(mpctx, "Error reinitializing audio.\n");
error_on_track(mpctx, ao_c->track);
}
}
}
mp_wakeup_core(mpctx);
}
// Returns audio start pts for seeking or video sync.
// Returns false if PTS is not known yet.
static bool get_sync_pts(struct MPContext *mpctx, double *pts)
{
struct MPOpts *opts = mpctx->opts;
*pts = MP_NOPTS_VALUE;
if (!opts->initial_audio_sync)
return true;
bool sync_to_video = mpctx->vo_chain && mpctx->video_status != STATUS_EOF &&
!mpctx->vo_chain->is_sparse;
if (sync_to_video) {
if (mpctx->video_status < STATUS_READY)
return false; // wait until we know a video PTS
if (mpctx->video_pts != MP_NOPTS_VALUE)
*pts = mpctx->video_pts - opts->audio_delay;
} else if (mpctx->hrseek_active) {
*pts = mpctx->hrseek_pts;
} else {
// If audio-only is enabled mid-stream during playback, sync accordingly.
*pts = mpctx->playback_pts;
}
return true;
}
// Look whether audio can be started yet - if audio has to start some time
// after video.
// Caller needs to ensure mpctx->restart_complete is OK
void audio_start_ao(struct MPContext *mpctx)
{
struct ao_chain *ao_c = mpctx->ao_chain;
if (!ao_c || !ao_c->ao || mpctx->audio_status != STATUS_READY)
return;
double pts = MP_NOPTS_VALUE;
if (!get_sync_pts(mpctx, &pts))
return;
double apts = playing_audio_pts(mpctx); // (basically including mpctx->delay)
if (pts != MP_NOPTS_VALUE && apts != MP_NOPTS_VALUE && pts < apts &&
mpctx->video_status != STATUS_EOF)
{
double diff = (apts - pts) / mpctx->opts->playback_speed;
if (!get_internal_paused(mpctx))
mp_set_timeout(mpctx, diff);
if (mpctx->logged_async_diff != diff) {
MP_VERBOSE(mpctx, "delaying audio start %f vs. %f, diff=%f\n",
apts, pts, diff);
mpctx->logged_async_diff = diff;
}
return;
}
MP_VERBOSE(mpctx, "starting audio playback\n");
ao_start(ao_c->ao);
mpctx->audio_status = STATUS_PLAYING;
if (ao_c->out_eof) {
mpctx->audio_status = STATUS_DRAINING;
MP_VERBOSE(mpctx, "audio draining\n");
}
ao_c->underrun = false;
mpctx->logged_async_diff = -1;
mp_wakeup_core(mpctx);
}
void fill_audio_out_buffers(struct MPContext *mpctx)
{
struct MPOpts *opts = mpctx->opts;
if (mpctx->ao && ao_query_and_reset_events(mpctx->ao, AO_EVENT_RELOAD))
reload_audio_output(mpctx);
if (mpctx->ao && ao_query_and_reset_events(mpctx->ao,
AO_EVENT_INITIAL_UNBLOCK))
ao_unblock(mpctx->ao);
update_throttle(mpctx);
struct ao_chain *ao_c = mpctx->ao_chain;
if (!ao_c)
return;
if (ao_c->filter->failed_output_conversion) {
error_on_track(mpctx, ao_c->track);
return;
}
if (ao_c->filter->ao_needs_update) {
if (reinit_audio_filters_and_output(mpctx) < 0)
return;
}
if (mpctx->vo_chain && ao_c->track && ao_c->track->dec &&
mp_decoder_wrapper_get_pts_reset(ao_c->track->dec))
{
MP_WARN(mpctx, "Reset playback due to audio timestamp reset.\n");
reset_playback_state(mpctx);
mp_wakeup_core(mpctx);
}
if (mpctx->audio_status == STATUS_SYNCING) {
double pts;
bool ok = get_sync_pts(mpctx, &pts);
// If the AO is still playing from the previous file (due to gapless),
// but if video is active, this may not work if audio starts later than
// video, and gapless has no advantages anyway. So block doing anything
// until the old audio is fully played.
// (Buggy if AO underruns.)
if (mpctx->ao && ao_is_playing(mpctx->ao) &&
mpctx->video_status != STATUS_EOF) {
MP_VERBOSE(mpctx, "blocked, waiting for old audio to play\n");
ok = false;
}
if (ao_c->start_pts_known != ok || ao_c->start_pts != pts) {
ao_c->start_pts_known = ok;
ao_c->start_pts = pts;
mp_filter_wakeup(ao_c->ao_filter);
}
if (ao_c->ao && mp_async_queue_is_full(ao_c->ao_queue)) {
mpctx->audio_status = STATUS_READY;
mp_wakeup_core(mpctx);
MP_VERBOSE(mpctx, "audio ready\n");
} else if (ao_c->out_eof) {
// Force playback start early.
mpctx->audio_status = STATUS_READY;
mp_wakeup_core(mpctx);
MP_VERBOSE(mpctx, "audio ready (and EOF)\n");
}
}
if (ao_c->ao && !ao_is_playing(ao_c->ao) && !ao_c->underrun &&
(mpctx->audio_status == STATUS_PLAYING ||
mpctx->audio_status == STATUS_DRAINING))
{
// Should be playing, but somehow isn't.
if (ao_c->out_eof && !mp_async_queue_get_frames(ao_c->ao_queue)) {
MP_VERBOSE(mpctx, "AO signaled EOF (while in state %s)\n",
mp_status_str(mpctx->audio_status));
mpctx->audio_status = STATUS_EOF;
mp_wakeup_core(mpctx);
// stops untimed AOs, stops pull AOs from streaming silence
ao_reset(ao_c->ao);
} else {
if (!ao_c->ao_underrun) {
MP_WARN(mpctx, "Audio device underrun detected.\n");
ao_c->ao_underrun = true;
mp_wakeup_core(mpctx);
ao_c->underrun = true;
}
// Wait until buffers are filled before recovering underrun.
if (ao_c->out_eof || mp_async_queue_is_full(ao_c->ao_queue)) {
MP_VERBOSE(mpctx, "restarting audio after underrun\n");
ao_start(mpctx->ao_chain->ao);
ao_c->ao_underrun = false;
ao_c->underrun = false;
mp_wakeup_core(mpctx);
}
}
}
if (mpctx->audio_status == STATUS_PLAYING && ao_c->out_eof) {
mpctx->audio_status = STATUS_DRAINING;
MP_VERBOSE(mpctx, "audio draining\n");
mp_wakeup_core(mpctx);
}
if (mpctx->audio_status == STATUS_DRAINING) {
// Wait until the AO has played all queued data. In the gapless case,
// we trigger EOF immediately, and let it play asynchronously.
if (!ao_c->ao || (!ao_is_playing(ao_c->ao) ||
(opts->gapless_audio && !ao_untimed(ao_c->ao))))
{
MP_VERBOSE(mpctx, "audio EOF reached\n");
mpctx->audio_status = STATUS_EOF;
mp_wakeup_core(mpctx);
}
}
if (mpctx->restart_complete)
audio_start_ao(mpctx); // in case it got delayed
}
// Drop data queued for output, or which the AO is currently outputting.
void clear_audio_output_buffers(struct MPContext *mpctx)
{
if (mpctx->ao)
ao_reset(mpctx->ao);
}