with Audio Stretching disabled:
Cubeb suffers from audio popping and doesn't let you adjust latency. Persists with Audio Stretching enabled.
OpenAL is similar to Cubeb when set to 100ms latency and is super crackly at low latency (30~ms) values. Audio is obviously very laggy at 100ms.
WASAPI (Exclusive) sounds best at very low latency (below 20ms) and gets progressively stuttery at 30+ms (extremely stuttery at 100ms). Seems to sound perfect at 2ms (two milliseconds, in case you think I wrote a typo). I tried 1ms but there was no audio. Audio stretching not required for good sound.
So... yeah, why does lowering Latency for WASAPI improve the situation? Is it not an audio buffer? Shouldn't increasing the buffer be how you improve the situation? It really caught me off guard that lowering it improved the situation.
Edit: Huh, seems to be pretty temperamental, I get no audio when set to a latency of 1, 3 and 5ms but I do get audio at 2 and 4ms.
Edit 2: Someone explained to me that maybe WASAPI is treating the soundcard like a DAW, in which case the lower you set the latency the higher the CPU usage but the more likely the CPU will finish processing the data before the audio data is overwritten with new data.
Cubeb suffers from audio popping and doesn't let you adjust latency. Persists with Audio Stretching enabled.
OpenAL is similar to Cubeb when set to 100ms latency and is super crackly at low latency (30~ms) values. Audio is obviously very laggy at 100ms.
WASAPI (Exclusive) sounds best at very low latency (below 20ms) and gets progressively stuttery at 30+ms (extremely stuttery at 100ms). Seems to sound perfect at 2ms (two milliseconds, in case you think I wrote a typo). I tried 1ms but there was no audio. Audio stretching not required for good sound.
So... yeah, why does lowering Latency for WASAPI improve the situation? Is it not an audio buffer? Shouldn't increasing the buffer be how you improve the situation? It really caught me off guard that lowering it improved the situation.
Edit: Huh, seems to be pretty temperamental, I get no audio when set to a latency of 1, 3 and 5ms but I do get audio at 2 and 4ms.
Edit 2: Someone explained to me that maybe WASAPI is treating the soundcard like a DAW, in which case the lower you set the latency the higher the CPU usage but the more likely the CPU will finish processing the data before the audio data is overwritten with new data.
Computer specifications:
Windows 10 Pro x64 | Ryzen 3700X | ASUS Crosshair VIII Hero WiFi Motherboard | 16GB 3600MHz | Geforce 1070Ti | Corsair AX760 PSU
Windows 10 Pro x64 | Ryzen 3700X | ASUS Crosshair VIII Hero WiFi Motherboard | 16GB 3600MHz | Geforce 1070Ti | Corsair AX760 PSU