Audio Buffer Latency Calculator | Turn Buffer Size and Sample Rate into Milliseconds

Convert any buffer size and sample rate into the latency your DAW or audio engine adds, in milliseconds. The result also shows the distance sound travels in that time (speed of sound 343 m/s), the 60fps frame equivalent, and a live / studio / offline verdict.

💡 About this tool

Dropping your DAW buffer from 256 to 128 samples cuts latency, but by how many milliseconds? Most people work by feel. Latency is just "buffer samples / sample rate × 1000," which means the same 256-sample buffer behaves very differently at 44100 Hz versus 96000 Hz — nearly a 2× difference in delay.

This calculator lets you swap numbers and see the delay gap in tangible units. Beyond raw milliseconds, it prints how far sound travels in that time and how many frames of a 60fps timeline the delay spans. "3 ms ≈ standing 1 metre away" reframes an abstract figure as something physical. The signal-path selector (input only / input + output round-trip / monitor chain) applies a multiplier so you can rough out real-world round-trip delay, not just the one-way figure.

🧐 Frequently Asked Questions

How low should latency be? A round-trip under roughly 10 ms feels immediate, like playing an acoustic instrument. Many engineers track comfortably up to about 12 ms, and discomfort creeps in past 20 ms. This tool flags under 10 ms as live-capable, 10-30 ms as studio recording, and over 30 ms as offline only.

Is a smaller buffer always better? No. Lower buffers cut latency but raise CPU load, which causes clicks and pops. The standard move is 64-256 samples while tracking, then 512-1024 while mixing for stability. Compare both figures here before you commit to one.

Does a higher sample rate reduce latency? For the same buffer size, yes. The time to process a fixed sample count is inversely proportional to the sample rate, so 256 samples clears faster at 96000 Hz than at 44100 Hz. The trade-off is higher CPU and disk load.

Will my interface actually hit this number? No — this is the buffer-only theoretical figure. Real round-trip latency adds ASIO / Core Audio driver buffers, OS scheduling, and A/D-D/A conversion. The round-trip and monitor-chain options approximate that with a multiplier, but expect the measured value to run higher.

Does direct monitoring make latency zero? Effectively yes. Direct monitoring on an audio interface routes the input straight to the outputs, bypassing the buffer entirely, so you hear yourself with near-zero delay. Buffer latency only matters when you need to monitor the signal through plugins.

📚 Why 10 ms is the magic number

Human hearing can't really detect delays under about 3 ms, and the 3-10 ms band is where most people start noticing. That window maps onto roughly 1 to 3.4 metres of air travel. When you hold an acoustic guitar, the strings sit a few dozen centimetres from your ears — you're constantly hearing a "natural" delay of about a millisecond. Monitoring under 10 ms feels like a real instrument precisely because it lands inside that everyday range. The distance readout in this tool makes the idea concrete: a 2048-sample buffer at 48000 Hz is about 42 ms, which converts to roughly 14.6 metres — close to how long a drummer's hit takes to reach you from the far side of a stage.