diff options
author | Chris Robinson <[email protected]> | 2017-05-02 03:57:17 -0700 |
---|---|---|
committer | Chris Robinson <[email protected]> | 2017-05-02 03:58:18 -0700 |
commit | afe2065d60d211a8009909abc327eea8830c10f5 (patch) | |
tree | c09c5f42fd0a07ba39bad17b16a9dc346e16e2eb /OpenAL32 | |
parent | f1a249b47a31eb7d7501cdc443611f6973159893 (diff) |
Set a voice as 'moving' if it starts/resumes at an offset
Diffstat (limited to 'OpenAL32')
-rw-r--r-- | OpenAL32/alSource.c | 29 |
1 files changed, 21 insertions, 8 deletions
diff --git a/OpenAL32/alSource.c b/OpenAL32/alSource.c index f2c4908f..3720ce49 100644 --- a/OpenAL32/alSource.c +++ b/OpenAL32/alSource.c @@ -2438,6 +2438,8 @@ AL_API ALvoid AL_APIENTRY alSourcePlayv(ALsizei n, const ALuint *sources) { ALbufferlistitem *BufferList; ALbuffer *buffer = NULL; + bool start_moving = false; + ALsizei s; source = LookupSource(context, sources[i]); WriteLock(&source->queue_lock); @@ -2478,11 +2480,17 @@ AL_API ALvoid AL_APIENTRY alSourcePlayv(ALsizei n, const ALuint *sources) case AL_PAUSED: assert(voice != NULL); - /* A source that's paused simply resumes. Make sure it uses the - * volume last specified; there's no reason to fade from where - * it stopped at. + /* A source that's paused simply resumes. Clear its mixing + * parameters and mark it as 'moving' so it fades in from + * silence. */ - voice->Flags &= ~VOICE_IS_MOVING; + voice->Step = 0; + voice->Flags |= VOICE_IS_MOVING; + memset(voice->Direct.Params, 0, sizeof(voice->Direct.Params[0])* + voice->NumChannels); + for(s = 0;s < device->NumAuxSends;s++) + memset(voice->Send[s].Params, 0, sizeof(voice->Send[s].Params[0])* + voice->NumChannels); ATOMIC_STORE(&voice->Playing, true, almemory_order_release); ATOMIC_STORE(&source->state, AL_PLAYING, almemory_order_release); goto finish_play; @@ -2521,7 +2529,12 @@ AL_API ALvoid AL_APIENTRY alSourcePlayv(ALsizei n, const ALuint *sources) ATOMIC_STORE(&voice->position, 0, almemory_order_relaxed); ATOMIC_STORE(&voice->position_fraction, 0, almemory_order_relaxed); if(source->OffsetType != AL_NONE) + { ApplyOffset(source, voice); + start_moving = ATOMIC_LOAD(&voice->position, almemory_order_relaxed) != 0 || + ATOMIC_LOAD(&voice->position_fraction, almemory_order_relaxed) != 0 || + ATOMIC_LOAD(&voice->current_buffer, almemory_order_relaxed) != BufferList; + } voice->NumChannels = ChannelsFromFmt(buffer->FmtChannels); voice->SampleSize = BytesFromFmt(buffer->FmtType); @@ -2534,10 +2547,10 @@ AL_API ALvoid AL_APIENTRY alSourcePlayv(ALsizei n, const ALuint *sources) */ voice->Step = 0; - voice->Flags = 0; - for(j = 0;j < voice->NumChannels;j++) - memset(&voice->Direct.Params[j].Hrtf.State, 0, - sizeof(voice->Direct.Params[j].Hrtf.State)); + voice->Flags = start_moving ? VOICE_IS_MOVING : 0; + memset(voice->Direct.Params, 0, sizeof(voice->Direct.Params[0])*voice->NumChannels); + for(s = 0;s < device->NumAuxSends;s++) + memset(voice->Send[s].Params, 0, sizeof(voice->Send[s].Params[0])*voice->NumChannels); if(device->AvgSpeakerDist > 0.0f) { ALfloat w1 = SPEEDOFSOUNDMETRESPERSEC / |