diff options
author | Chris Robinson <[email protected]> | 2017-02-18 16:55:48 -0800 |
---|---|---|
committer | Chris Robinson <[email protected]> | 2017-02-18 16:55:48 -0800 |
commit | 2448f88e70f6207ad5743f0a55eaa5de7cbce737 (patch) | |
tree | 135d60ee0a486d70ab5482011041d4a363c78537 /Alc/backends | |
parent | d8c42918f46e2439c81c4d4562df8fa7ad2e01fd (diff) |
Return some device latency by default
A device will never have 0 latency. OpenAL Soft itself uses a sample buffer
length of UpdateSize*NumUpdates, and during playback will have about
(NumUpdates-1) periods filled, more or less. Without a more accurate
measurement from the playback system, this is better than reporting 0.
Diffstat (limited to 'Alc/backends')
-rw-r--r-- | Alc/backends/base.c | 9 |
1 files changed, 7 insertions, 2 deletions
diff --git a/Alc/backends/base.c b/Alc/backends/base.c index ff808f53..e4305653 100644 --- a/Alc/backends/base.c +++ b/Alc/backends/base.c @@ -4,6 +4,7 @@ #include <stdlib.h> #include "alMain.h" +#include "alu.h" #include "backends/base.h" @@ -45,8 +46,12 @@ ClockLatency ALCbackend_getClockLatency(ALCbackend *self) almtx_lock(&self->mMutex); ret.ClockTime = GetDeviceClockTime(device); - // TODO: Perhaps should be NumUpdates-1 worth of UpdateSize? - ret.Latency = 0; + /* NOTE: The device will generally have about all but one periods filled at + * any given time during playback. Without a more accurate measurement from + * the output, this is an okay approximation. + */ + ret.Latency = device->UpdateSize * DEVICE_CLOCK_RES / device->Frequency * + maxu(device->NumUpdates-1, 1); almtx_unlock(&self->mMutex); return ret; |