OBS audio bitrate: what to set and why

Open Settings and go to the Output section. What you see next depends on which output mode you are in, and this is where most confusion starts.

In Simple output mode there is a single Audio Bitrate dropdown that applies to everything. In Advanced mode there is a separate Audio tab, where each of the six audio tracks has its own bitrate. If you change the bitrate in Simple mode and hear no difference in your recording, check whether you were actually in Advanced mode all along.

The choices are the same either way: 64, 96, 128, 160, 192, 224, 256, 288 and 320 kbps. OBS defaults to 160.

Audio bitrate is how much data is spent describing each second of sound. More data means the encoder can represent the original more faithfully; less data means it discards more, and at some point you start hearing what was discarded.

It has nothing to do with how loud your stream is. Raising the bitrate will not fix a quiet microphone, and lowering it will not stop distortion. Those are gain problems, and no encoding setting touches them.

The returns fall off sharply. Going from 64 to 128 kbps is an obvious improvement on almost any material. Going from 256 to 320 is inaudible to most listeners on most content, and costs bandwidth you may need for video.

For a stream that is mostly speech, the default 160 kbps is comfortably sufficient and 128 is genuinely fine. Speech occupies a narrow frequency range compared with music, so it survives compression far better.

Going higher for a podcast or a talking-head stream mostly wastes bandwidth. If you are hitting bandwidth limits, the audio bitrate is one of the least painful places to save, because dropping from 160 to 128 frees capacity your video will use better.

Only drop to 96 or 64 kbps if you are severely constrained. Speech remains intelligible down there, but it starts sounding thin and papery in a way listeners notice even if they cannot name it.

Music is where the difference becomes audible. Cymbals, reverb tails and sustained high frequencies are exactly what lossy encoding struggles with, and they are what tell you a stream is under-encoded.

For a music performance, a DJ set or anything where the audio is the main attraction, 192 kbps upwards is worth it, and 256 to 320 is reasonable if your platform accepts it and your connection can carry it.

Game streams sit in between. Game audio has music and effects in it, but viewers are watching gameplay rather than listening critically, so 160 to 192 is a sensible middle.

When you stream, your bitrate is not what viewers receive. Streaming platforms re-encode incoming audio to their own specification, and several cap what they will accept. Sending 320 kbps to a service that caps at 160 gains you nothing at all.

Check your platform's published limit before raising the setting. The number is easy to find in their streaming guidelines and it saves you spending upload bandwidth on data that is discarded on arrival.

Local recording is different, and this is the case for keeping two settings. Nothing re-encodes a local file, so if you are recording to disk for later editing, a higher bitrate is genuinely preserved. In Advanced mode you can give your recording track a higher bitrate than your streaming track.

Under Settings and then Audio you will find sample rate, offered at 44.1 or 48 kHz. Pick 48 kHz and leave it there. It matches what video workflows expect, and mismatched sample rates between OBS and your interface cause resampling artefacts that sound like faint crackling.

The channels setting in the same place is worth a moment's thought. Stereo is the default, but a single microphone produces the same signal on both channels, so you are spending twice the data on one voice.

None of this matters if your levels are wrong. Set your gain so peaks land well below the ceiling first, because clipping at any bitrate sounds worse than clean audio at a low one.