During a resampling operation where 1) user has specified first_pts 2) SWR_FLAG_RESAMPLE is not set initially (directly or otherwise) 3) first_pts has been fulfilled (always using hard compensation) then upon first encountering a delay where a soft compensation is required, swr_set_compensation will lead to another init of swr which will reset outpts to the specified firstpts thus leading to an output frame having its pts = firstpts. When the next input frame is received, swr will see a large delay and inject silence from firstpts to the current frame's pts. This can lead to severe desync and in worst case, loss of audio playback. Parameter firstpts initialized to AV_NOPTS_VALUE in swr_alloc and then checked in swr_init to avoid resetting outpts, thus avoiding reapplication of firstpts. Fixes #4131.
Unified · Split
| @@ -171,6 +171,7 @@ av_cold struct SwrContext *swr_alloc(void){ · 170 unmodified lines | ||
| 171 | 171 | if(s){ |
| 172 | 172 | s->av_class= &av_class; |
| 173 | 173 | av_opt_set_defaults(s); |
| 174 | + s->firstpts = AV_NOPTS_VALUE; | |
| 174 | 175 | } |
| 175 | 176 | return s; |
| 176 | 177 | } |
| @@ -375,8 +375,9 @@ av_cold int swr_init(struct SwrContext *s){ · 374 unmodified lines | ||
| 375 | 375 | if (s->firstpts_in_samples != AV_NOPTS_VALUE) { |
| 376 | 376 | if (!s->async && s->min_compensation >= FLT_MAX/2) |
| 377 | 377 | s->async = 1; |
| 378 | − s->firstpts = | |
| 379 | − s->outpts = s->firstpts_in_samples * s->out_sample_rate; | |
| 378 | + if (s->firstpts == AV_NOPTS_VALUE) | |
| 379 | + s->firstpts = | |
| 380 | + s->outpts = s->firstpts_in_samples * s->out_sample_rate; | |
| 380 | 381 | } else |
| 381 | 382 | s->firstpts = AV_NOPTS_VALUE; |
| 382 | 383 | |
| @@ -1082,6 +1082,9 @@ $(call CROSS_TEST,$(SAMPLERATES_LITE),ARESAMPLE_EXACT_LIN_ASYNC,s32p,s32le,s16) · 1081 unmodified lines | ||
| 1082 | 1082 | $(call CROSS_TEST,$(SAMPLERATES_LITE),ARESAMPLE_EXACT_LIN_ASYNC,fltp,f32le,s16) |
| 1083 | 1083 | $(call CROSS_TEST,$(SAMPLERATES_LITE),ARESAMPLE_EXACT_LIN_ASYNC,dblp,f64le,s16) |
| 1084 | 1084 | |
| 1085 | +FATE_SWR_RESAMPLE-$(call FILTERDEMDEC, ARESAMPLE ASETPTS ATRIM SINE, , PCM_S16LE, LAVFI_INDEV) += fate-swr-async-firstpts | |
| 1086 | +fate-swr-async-firstpts: CMD = framecrc -auto_conversion_filters -copyts -f lavfi -i "sine=r=1000:samples_per_frame=100,asetpts=PTS+S+S*floor(ld(1)/4)+st(1\,ld(1)+1)*0,atrim=end=2" -filter:a aresample=async=300:first_pts=0 | |
| 1087 | + | |
| 1085 | 1088 | FATE_SWR_RESAMPLE-$(call FILTERDEMDECENCMUX, ARESAMPLE, WAV, PCM_S16LE, PCM_S16LE, WAV) += $(FATE_SWR_RESAMPLE) |
| 1086 | 1089 | fate-swr-resample: $(FATE_SWR_RESAMPLE-yes) |
| 1087 | 1090 | FATE_SWR += $(FATE_SWR_RESAMPLE-yes) |
| @@ -0,0 +1,24 @@ | ||
| 1 | +#tb 0: 1/1000 | |
| 2 | +#media_type 0: audio | |
| 3 | +#codec_id 0: pcm_s16le | |
| 4 | +#sample_rate 0: 1000 | |
| 5 | +#channel_layout_name 0: mono | |
| 6 | +0, 0, 0, 132, 264, 0xc2981f45 | |
| 7 | +0, 132, 132, 68, 136, 0xe78e468d | |
| 8 | +0, 200, 200, 100, 200, 0xd55c67d0 | |
| 9 | +0, 300, 300, 100, 200, 0xd55c67d0 | |
| 10 | +0, 400, 400, 100, 200, 0xd55c67d0 | |
| 11 | +0, 500, 500, 93, 186, 0x85ca5db4 | |
| 12 | +0, 593, 593, 110, 220, 0xa2655d0b | |
| 13 | +0, 703, 703, 108, 216, 0x95cb6f01 | |
| 14 | +0, 811, 811, 108, 216, 0xf35668b8 | |
| 15 | +0, 919, 919, 149, 298, 0xc273245f | |
| 16 | +0, 1068, 1068, 136, 272, 0xedeb6e0a | |
| 17 | +0, 1204, 1204, 98, 196, 0xea18668e | |
| 18 | +0, 1302, 1302, 98, 196, 0x412861e7 | |
| 19 | +0, 1400, 1400, 98, 196, 0x7ec361b2 | |
| 20 | +0, 1498, 1498, 110, 220, 0xf3ae6a6a | |
| 21 | +0, 1608, 1608, 108, 216, 0xab2f6c93 | |
| 22 | +0, 1716, 1716, 107, 214, 0x50de6eb9 | |
| 23 | +0, 1823, 1823, 106, 212, 0x67b8656d | |
| 24 | +0, 1929, 1929, 18, 36, 0x2b7911c6 | |