When recording or monitoring audio through a computer, the audio interface processes sound in small chunks called buffers — and the buffer size directly determines how much delay (latency) you hear between playing a note and hearing it back.
When recording or monitoring audio through a computer, the audio interface processes sound in small chunks called buffers — and the buffer size directly determines how much delay (latency) you hear between playing a note and hearing it back. Smaller buffers mean lower latency but more strain on your computer's processor.
A buffer size of 256 samples at a 44,100 Hz sample rate. One-Way Latency = (256/44100) × 1000 ≈ 5.8 ms. Round-Trip Latency ≈ 11.6 ms — low enough for most musicians to comfortably play and record in real time.
It's the delay between an audio event (like a keystroke on a MIDI keyboard) and hearing the resulting sound, caused mainly by the audio buffer size.
Latency (ms) = (buffer size in samples ÷ sample rate) × 1000. A larger buffer means higher latency but more processing stability.
Smaller buffers (64-128 samples) minimize latency for real-time monitoring, though they demand more CPU power; larger buffers are fine for mixing where latency doesn't matter.