What is audio bitrate?

The short answer

A bit is the smallest unit of digital information, a single one or zero. Digital audio is just a very long stream of those bits, and bitrate is how many of them go by each second. It is usually written in kbps, kilobits per second, where one kbps is a thousand bits every second. A larger stream, such as high quality video audio, may be written in Mbps, millions of bits per second.

The rule is simple: more bits per second means more of the original detail is kept, at the cost of a bigger file or more bandwidth to move it. Fewer bits per second means a smaller, lighter file that has had to leave some detail behind. Bitrate is the number that sits in the middle of that trade, and knowing what it controls is what lets you pick a sensible value instead of guessing.

How uncompressed bitrate is calculated

When audio is stored raw, with nothing thrown away, its bitrate is not a setting you choose. It falls straight out of three numbers that describe the recording. The sample rate is how many times per second the sound is measured, the bit depth is how many bits each of those measurements uses, and the channel count is how many separate streams there are, such as two for stereo. Multiply them together and you have the bits per second.

So a studio recording at 48,000 samples per second, 24 bits per sample, in stereo works out to 48,000 times 24 times 2, which is 2,304,000 bits per second, or about 2,304 kbps. A standard CD, at 44,100 samples per second and 16 bits in stereo, comes to roughly 1,411 kbps. For the two ingredients that drive this, see what is sample rate and what is bit depth.

That is why raw audio is so large. Every second of that stereo studio take is a little over a quarter of a megabyte, and a few minutes runs into tens of megabytes. It sounds pristine because nothing has been discarded, but it is a lot to store and far too much to send smoothly over most internet connections. That weight is exactly the problem compression exists to solve.

Compression and the kbps you choose

Most audio you download or stream has been through lossy compression, which shrinks the file by throwing away the parts of the sound a listener is least likely to notice. Common formats like MP3, AAC, and Opus all work this way. Here bitrate becomes a genuine choice: the encoder aims for a target kbps, and that target decides how much of the original it is allowed to keep. A higher kbps keeps more and sounds closer to the source, a lower kbps keeps less and can start to sound thin or smeared.

• Speech and podcasts. A voice does not need many bits, so 96 to 128 kbps usually sounds clean and keeps files small.
• Music streaming. 128 to 320 kbps is the normal range, and 256 or 320 kbps is close enough to the original that most listeners cannot reliably tell the difference on everyday gear.
• Archiving or heavy editing. Skip lossy compression and keep the audio uncompressed, or use a lossless format that shrinks the file with no loss at all, so you never bake in damage you cannot undo.

The catch with lossy formats is that the loss is permanent. Once detail is gone it cannot be rebuilt by raising the bitrate later, and compressing an already compressed file only stacks the damage. That is the reason the honest workflow is to work from the highest quality source you can and only compress at the very end, when you know the final destination.

Which bitrate should I use?

There is no single best number, because the right bitrate depends on whether you are capturing audio to work on or delivering it to a listener. The difference between those two jobs is what should guide the value, not a habit of always reaching for the biggest figure.

• When you record. Capture uncompressed, or at the highest quality your recorder offers, so every edit, effect, and mix works from the full detail. You can always compress a clean master down later.
• When you stream or upload. Your streaming or recording app chooses the output bitrate, so pick a target that suits the content: a middle value for talk, a higher one for music. Many platforms re-encode your audio on their end anyway, so a wildly high number is often wasted.

A good instinct is to keep quality high while you can still change your mind, and to compress only once, at the last step. Past the point where a listener cannot hear the difference, extra bitrate just costs storage and bandwidth without adding anything you can actually perceive.

Bitrate, sample rate, and bit depth

These three terms get mixed up constantly, so it helps to see how they fit together. Sample rate is how often the sound is measured, bit depth is how finely each of those measurements is stored, and for uncompressed audio the bitrate is simply the two of them multiplied by the number of channels. In other words, bitrate is the result, and the other two are ingredients.

The important twist is that this only holds for raw audio. Once a lossy encoder is involved, bitrate is set on its own and no longer equals that multiplication, because the format is deciding how much data to spend rather than storing every sample in full. If you want the difference between the two ingredients laid out side by side, see sample rate vs bit depth.

Bitrate in patchd

In patchd, a Windows virtual audio mixer, there is no bitrate to set for what you hear. Monitoring is live, uncompressed audio flowing through the engine in real time, not a compressed stream, so the concept of kbps simply does not apply to the sound coming back to your own ears. Bitrate enters the picture only at the moment you record or stream, and that choice belongs to the app doing the recording or streaming.

What patchd does is hand that app a clean, fully mixed feed to compress. It carries your sources through a bus, a virtual output that other apps pick up as if it were a microphone, and your recorder or streaming software then applies whatever output bitrate you set there. The engine runs over ASIO and WASAPI Exclusive, the low overhead Windows audio paths, so the audio it delivers stays clean before any compression happens. For how sources reach those buses, see what is audio routing, and for capturing system sound, see how to record desktop audio on Windows.

The takeaway

Bitrate is how much audio data moves each second, measured in kbps. For uncompressed audio it is fixed by sample rate, bit depth, and channels, and it is heavy. Lossy compression trades some of that data for a much smaller file, and the kbps you pick sets how much detail survives. Keep quality high while you record and edit, compress once at the end to suit where the audio is going, and there is no reason to pay for more bits than a listener can hear.

patchd is pre-launch, a real-time Windows mixer that keeps your monitoring uncompressed and clean while your recording or streaming app handles the bitrate on the way out. If that is the setup you want, join the waitlist and get notified when it is ready. The paid Studio tier is $39.99 per year, and the engine is identical on both tiers.