Buffer size is the number of audio samples your computer collects and processes in one batch before passing them on. It is the main control you have over the trade-off between latency and processing load: a small buffer gives you low latency but stresses the CPU, while a large buffer eases the CPU but adds delay. Set it small while recording and large while mixing.
Quick answer: Use a low buffer size (such as 64 or 128 samples) when tracking so you hear yourself with minimal delay, and a high buffer size (512 or 1024 samples) when mixing so heavy plugin chains do not crackle or drop out.
What buffer size actually does
Your audio interface and computer cannot process audio sample by sample in real time — that would overwhelm the system. Instead, samples are gathered into a small buffer, processed as a block, and then handed off. The buffer size, measured in samples (commonly 32, 64, 128, 256, 512, or 1024), sets how many samples are in each block.
- Smaller buffer = less delay, but the CPU has to work harder and more often.
- Larger buffer = more delay, but the CPU has more breathing room and runs more reliably.
This delay is called latency. For the bigger picture, read what is audio latency.
How buffer size affects latency
Buffer size is the largest factor you control in round-trip latency — the time between making a sound and hearing it back through your DAW. The relationship is direct: at a 44.1 kHz sample rate, a 128-sample buffer corresponds to roughly 2.9 milliseconds per direction, while a 512-sample buffer is closer to 11.6 milliseconds. Real round-trip latency is higher because there are buffers on both input and output plus driver overhead, but the principle holds: halve the buffer and you roughly halve the buffer-related delay. If lag is your main frustration, the buffer is the first lever in a broader playbook for reducing latency when recording.
For a sense of scale, latency under about 10 milliseconds usually feels immediate when monitoring a performance. Higher than that and singers and players start to notice a lag that throws off their timing.
How buffer size affects CPU load
A small buffer forces your computer to process audio in many tiny, frequent chunks, which is harder on the CPU. If the buffer is too small for your project, you will hear clicks, pops, crackles, or dropouts, and your DAW may report an overload. Increasing the buffer size gives the processor more time per block and almost always fixes these glitches. This is why heavy mixing sessions with many plugins need a large buffer, while a glitch-free vocal take with one plugin can run very small.
How buffer size and sample rate work together
Buffer size and sample rate both feed into how much delay you hear, and it helps to keep them separate in your head. The buffer is a number of samples; the sample rate is how many samples make up one second. A fixed buffer therefore represents a shorter slice of time at a higher sample rate. A 256-sample buffer at 44.1 kHz works out to about 5.8 milliseconds per direction, but the same 256 samples at 96 kHz is only around 2.7 milliseconds. So raising the sample rate can lower latency at a given buffer — but it also doubles the data your CPU has to chew through, which eats up the very headroom you were trying to protect. If those numbers are new to you, our guide to sample rate and bit depth explains what each one actually changes.
For most home setups, 44.1 kHz or 48 kHz with a sensible buffer is the practical sweet spot. Pushing to 96 kHz purely to shave milliseconds off monitoring usually costs more in CPU strain than it returns, especially once plugins are loaded. Pick a sample rate to suit the project, then use the buffer as your day-to-day latency control.
How to set buffer size in your DAW
The setting lives in your interface or audio preferences, not usually per-track.
- GarageBand: Open Settings, go to the Audio/MIDI tab, and use Audio Resolution and the driver settings; on macOS the buffer is often managed through the interface’s control panel. Lower it for recording, raise it for mixing.
- Audacity: Go to Preferences, then Devices, and adjust the buffer length under recording/playback settings. Audacity is less focused on live monitoring, so moderate values are fine.
- FL Studio: Open Options, then Audio Settings, and set the Buffer length slider (shown in samples or milliseconds). Use a small value while recording and increase it for mixdown.
Most interfaces also expose buffer size in their own driver control panel on Windows. Always set your levels correctly first, then dial in the buffer.
How to choose the right buffer size
There is no single correct number — the right buffer depends on what you are doing right now and how powerful your machine is. Work through it like this:
- Tracking a live performance: go as low as your system allows without glitches, usually 64 or 128 samples, so the performer hears themselves in time.
- Tracking with direct monitoring: the buffer no longer affects what the performer hears, so leave it comfortably high and avoid CPU stress.
- Programming MIDI or virtual instruments: keep it fairly low (128 to 256) so notes respond promptly as you play the keys, but raise it once the part is recorded.
- Mixing and mastering: push the buffer to 512 or 1024 — you are no longer performing, so the extra delay is irrelevant and stability is everything.
The honest approach is to treat buffer size as a slider you move several times in a session rather than a value you set once and forget.
Common buffer size mistakes
A few habits trip people up repeatedly:
- Leaving the buffer high while recording. If you monitor through the DAW, a large buffer adds noticeable lag that makes performances feel sluggish and pushes players off the beat.
- Leaving the buffer low while mixing. As the plugin count climbs, a tiny buffer is the first thing to crackle and stutter. Raise it before you start troubleshooting anything else.
- Blaming the buffer for every glitch. Background apps, a full disk, unstable USB hubs, and flaky drivers all cause dropouts too. Increase the buffer first, but do not stop there if the problem persists.
- Forgetting direct monitoring exists. Many people fight latency for years without realising their interface can route input straight to the headphones, sidestepping the buffer entirely.
A simple workflow for buffer size
- Recording: set the buffer low (64 or 128 samples) so monitoring feels immediate. If you get glitches, nudge it up one step.
- Mixing: raise the buffer (512 or 1024 samples). You are not performing live, so the added latency does not matter and the CPU stays stable under heavy plugin loads.
- Use direct monitoring if your interface supports it — then you can mix at a high buffer and still hear yourself with no delay while recording. Many interfaces in our audio interfaces hub offer it.
Frequently asked questions
What buffer size should I use for recording vocals?
Start at 128 samples for a comfortable, low-latency feel. If you hear clicks or dropouts, raise it to 256. If your interface has direct monitoring, latency stops mattering and you can use a higher buffer freely.
Does buffer size affect sound quality?
No. Buffer size does not change the recorded audio quality or fidelity — it only affects latency and CPU load. The exported file sounds identical regardless of the buffer used while working.
Why do I get clicks and pops at low buffer sizes?
A buffer that is too small does not give your CPU enough time to process each block, causing dropouts. Increase the buffer size, close background apps, and freeze or bounce heavy tracks to stabilise playback.
Does a bigger buffer make my final export slower?
No. Buffer size governs real-time monitoring while you work, not offline bouncing. When you export or freeze a track, the DAW renders as fast as your CPU allows regardless of the buffer setting, so a large buffer costs you nothing at mixdown.
