Audio Latency Calculator - Buffer Size and Sample Rate to Milliseconds

Use the free tool below: your data never leaves your browser — all work happens on your device, and nothing is uploaded to any server.

Why your playing feels late

Latency is the delay between playing a note and hearing it back through your computer. It is measured in milliseconds, and small numbers matter more than you might expect: 10 milliseconds feels instant to most players, 20 milliseconds feels like a slight slap-back, and 40 milliseconds or more feels like playing in a different room from your own hands. The calculator above works out latency from the two numbers that control it most directly, buffer size and sample rate, using the simple formula of buffer divided by sample rate times one thousand.

A buffer is a small queue of audio samples that your computer processes in one chunk. A 512-sample buffer at 44,100 Hz holds about 11.6 milliseconds of audio, so every chunk takes that long to fill before it can be processed and played. Smaller buffers mean shorter waits but more work for the computer, because it must wake up and process audio more often. If the computer cannot keep up, you hear clicks, pops and dropouts, which are worse than a few extra milliseconds of delay. Latency tuning is therefore a trade-off, not a race to zero.

How to use the calculator

Pick your buffer size from the list, or read it from your audio interface settings.
Pick your sample rate, usually 44,100 or 48,000 Hz.
Choose how many buffers are in the round trip: one for a one-way trip, two for input plus output, which is the usual real-world case.
Read the one-buffer time and the total, and compare with the guide: under 10 milliseconds is great for live monitoring, under 20 is usable, above that most players struggle.

A worked example

Take a 256-sample buffer at 48,000 Hz. One buffer takes 5.33 milliseconds. A round trip through input and output buffers doubles that to 10.67 milliseconds before anything else is counted. Add converter delay, driver overhead and a couple of plug-ins, and a real system often lands between 12 and 18 milliseconds at these settings. Drop the buffer to 128 samples and the buffer part falls to 5.33 milliseconds round trip, but the CPU must now process twice as many chunks per second, so a modest laptop may start crackling on a heavy session. Raise the buffer to 1024 samples and the round trip buffer time jumps to 42.67 milliseconds, which is fine for mixing but hopeless for tracking a vocal against a backing track.

Sample rate plays a supporting role here: at a higher sample rate, the same buffer holds less time, so latency falls. A 256-sample buffer at 96,000 Hz takes only 2.67 milliseconds. That is one reason some players track at high sample rates, but the CPU cost rises with the rate too, so it is not a free lunch. The calculator lets you test these trade-offs instantly instead of changing settings and re-testing by ear each time.

What the calculator does not count

The numbers here cover buffer time only, and real latency is always larger. Analogue-to-digital and digital-to-analogue converters add roughly a millisecond each, drivers add their own hidden buffers, and every plug-in in the monitoring chain adds its look-ahead or processing delay. Some plug-ins, like linear-phase equalisers and limiters, add tens of milliseconds on their own. The honest way to use this tool is as a floor: your real latency is at least the number shown, and the gap between the shown number and your measured round-trip latency tells you how much your drivers and plug-ins are adding. Most interfaces report a measured round-trip figure in their control panel, and comparing the two is a useful diagnostic.

Practical ways to cut latency

Use a dedicated audio interface with proper drivers instead of built-in sound: the driver is usually the biggest single win.
Monitor through the interface’s direct monitoring path, which routes your input to your headphones without passing through the computer at all.
Freeze or bypass heavy plug-ins while tracking, and bring them back for mixing when latency no longer matters.
Raise the buffer for mixing and lower it only for tracking: there is no rule that one setting must serve both jobs.

Who uses this calculation

Musicians use it when a new interface feels sluggish, podcasters use it when their own voice echoes in their headphones, and live performers use it to decide whether a laptop rig is safe on stage. If your playing feels late, do not practise harder against the delay: measure it first, because a settings change that takes thirty seconds often fixes what hours of practice cannot.

Audio Latency Calculator – Buffer Size and Sample Rate to Milliseconds