Sample rate is how many times per second the audio is measured, and bit depth is how precisely each of those measurements is stored. Sample rate sets the range of pitches you can capture, and bit depth sets how much quiet detail sits above the noise. For most music, working at 44.1 kHz and 24 bit and delivering at 44.1 kHz and 16 bit covers you completely.
The round numbers here are the ones almost everyone uses, and they are not magic. This is a plain explanation of what each one means and which to pick, without pretending that the highest setting is always the best one.
Sample rate is how many times per second the audio is measured. A rate of 44.1 kHz means the sound is measured 44,100 times every second, and each of those measurements is called a sample. A computer cannot store a continuous wave, so it takes a rapid series of snapshots instead, and the sample rate is how many snapshots it takes each second.
The higher the rate, the wider the range of frequencies it can capture. The rule behind this has a name worth remembering: a sample rate can capture frequencies up to half of itself. This is the Nyquist idea. At 44.1 kHz you can capture sound up to about 22 kHz, which sits just above the roughly 20 kHz top of human hearing, so 44.1 kHz already covers everything a person can actually hear.
That is why 44.1 kHz became the standard for music and CDs, and why 48 kHz is the standard for video and film. Higher rates like 88.2 kHz and 96 kHz exist and some people record at them, but the extra range they capture sits above hearing, so the gain is mostly on paper. Here is where the common rates land.
| Sample rate | Where it is used |
|---|---|
| 44.1 kHz | Music and CD, the standard for released audio. Captures up to about 22 kHz, above the top of hearing. |
| 48 kHz | Video and film, the standard for anything paired with picture. |
| 88.2 kHz | Double 44.1 kHz. A higher rate some use while recording and processing. |
| 96 kHz | Double 48 kHz. A higher rate some use while recording and processing. |
| 192 kHz | Very high, rarely needed, and heavy on CPU and disk. |
A sample rate captures frequencies up to half of itself. So 44.1 kHz reaches about 22 kHz, which already covers the roughly 20 kHz limit of human hearing. Going higher captures sound above what you can hear, which is why bigger sample rates buy far less than the numbers suggest.
Bit depth is how precisely each measurement is stored. More bits mean more possible values for each sample, which means a finer scale for the level and more room between the loudest sound and the quiet hiss at the bottom. That distance, between the loudest undistorted sound and the noise floor, is called dynamic range, and bit depth is what sets it.
Each bit adds about 6 dB of dynamic range, so the numbers fall out simply. A 16 bit file gives about 96 dB of dynamic range and is the standard for CDs and for final delivery. A 24 bit file gives about 144 dB and is what you record and mix at, because that extra room is headroom: you can set levels safely below the ceiling without your quiet detail sinking into noise, and you have space before clipping if something peaks louder than you expected.
| Bit depth | Dynamic range | Where it is used |
|---|---|---|
| 16 bit | about 96 dB | CD and final delivery, the standard for released files |
| 24 bit | about 144 dB | Recording and mixing, for headroom against clipping and a lower noise floor |
| 32 bit float | far more than needed | A working format many DAWs use internally, so hard to clip that peaks stay recoverable |
You record at 24 bit not because you can hear all 144 dB, but for the headroom. It lets you track at a comfortable, lower level with plenty of room before the ceiling, so a loud surprise does not clip and your quiet parts stay well above the noise. You then deliver at 16 bit, which already gives about 96 dB, more than almost any listening room reveals.
For most music, record and mix at 44.1 kHz and 24 bit, then deliver at 44.1 kHz and 16 bit. If your work is going into video, use 48 kHz instead of 44.1 kHz. The bit depth choice matters more day to day than a high sample rate does, because 24 bit is what gives you headroom while you work.
Match the sample rate to where the audio ends up. Music and streaming live at 44.1 kHz, and video and film live at 48 kHz. Pick the rate of the final destination and work at it from the start, so there is no conversion step later. If you record real sound, that rate is set in your audio interface and your DAW, and it is worth checking both agree before you start.
Do not chase very high sample rates. Rates like 96 kHz or 192 kHz mostly cost you CPU and disk space for very little you can actually hear, since the extra range sits above human hearing. They can help a little with heavy pitch shifting or certain processing, but for most people 44.1 kHz or 48 kHz is the sensible choice, and the difference on playback is inaudible to almost everyone. Bigger project files and a busier processor are the reliable part of the trade.
Record at 24 bit, deliver at 16 bit. The 24 bit working file gives you room to set levels safely and mix without noise creeping in. The 16 bit delivery file is the standard for released audio and already carries about 96 dB of dynamic range. Once the music is made and mixed, a separate question is how loud your master should be, which is about level, not about sample rate or bit depth.
Work at 44.1 kHz and 24 bit for music, or 48 kHz and 24 bit if it is going into video. Deliver at 16 bit. That covers the overwhelming majority of projects, sounds no different from far heavier settings on playback, and keeps your files and your processor sane.
Dithering is a tiny amount of low level noise added on purpose when you reduce bit depth, for example when you export a 24 bit mix down to a 16 bit file. Cutting from 24 bit to 16 bit throws away the finest detail, and the leftover rounding creates a faint, harsh distortion that is most audible on quiet moments like fades and reverb tails. Dither replaces that harsh distortion with a gentler, more natural noise the ear ignores.
In practice you only meet it at one moment: the final export to 16 bit. Your DAW has a dither option in its export or bounce dialog, and you switch it on when you render that 16 bit delivery file. You do not dither if you are staying at 24 bit, because nothing is being reduced, and you do not dither more than once, since each pass adds a little noise. Add it once, at the last step, and only when the bit depth is actually going down.
Dither is a small, deliberate noise added when you lower bit depth, so the quiet parts stay smooth instead of turning grainy. Use it once, at the final export to 16 bit. If you are not reducing bit depth, you do not need it at all.
Sample rate and bit depth are settings, and once they are sensible they stay out of your way. The music is the part that takes the time, and it is worth spending your attention there rather than on decimal places.
Our instruments make sound inside the computer, so they run cleanly at whatever sample rate and bit depth your project is set to, whether that is 44.1 kHz for music or 48 kHz for video, 16 bit or 24 bit. You do not set a rate on them or match anything by hand. They follow your session, in the browser or in your DAW, so you can focus on the music and not the numbers.
When you want them in a project, The Collection ships as AU, VST3, AAX and a standalone app on macOS, signed and notarized, and a VST3 on Windows, so it loads into whatever session and settings you already have.
Free to hear in the browser, no signup. It follows your project's sample rate and bit depth either way.
We write these when there is something worth writing down. One email when a new one lands or a new Tunary instrument ships. No newsletter, no schedule.
Tunary