Correct Sample Rate and Bit Depth for Stems
Choosing a technically consistent delivery format — how to pick the right rate and depth, when conversion is necessary, and what dither and 32-bit float actually do.
Should I export stems at 44.1 kHz or 48 kHz?
This is the most common sample-rate question in stem delivery. There is no fixed answer that suits every project — the safest rule below explains how to choose.
The safest rule: match the project and the recipient
There is no single sample rate that is correct for every stem delivery. The correct rate is normally the native rate of the approved production session, unless the mixing engineer, broadcaster, game engine, or label specifies another format. All files in the package should be technically consistent unless an exception is intentional and documented.
Unless requested otherwise, stereo stems should normally be delivered as interleaved WAV files rather than split left/right files.
Should stems be WAV or AIFF?
WAV is the safest default unless another format is requested.
Sample rate in practical terms
Sample rate describes how many samples represent each second of audio. The most common music-production rates are 44.1 kHz and 48 kHz, but also include 88.2 kHz and 96 kHz. Changing the rate requires resampling; changing only the metadata or importing at the wrong rate changes playback speed and pitch.
| Situation | Recommended approach |
|---|---|
| Session recorded and approved at 48 kHz | Deliver 48 kHz unless another rate is explicitly requested. |
| Recipient specifies 44.1 kHz | Use one high-quality conversion from the approved source. |
| Mixed-rate source files inside session | Let the DAW resolve them, then export a single consistent session rate. |
| No specification is supplied | Confirm before export; 24-bit WAV at the native session rate is a common professional default. |
| Lossy reference file | Keep it as a reference only, not as a production stem. |
Bit depth in practical terms
Bit depth affects the resolution and available dynamic range of fixed-point PCM audio. For stem delivery, 24-bit is widely used because it provides ample headroom and compatibility. Sixteen-bit may be required for a specific final format, while 32-bit floating-point can preserve values above 0 dBFS and avoid fixed-point clipping during technical interchange — but only when the receiving workflow supports it.
Don’t increase specifications for appearance
Dither: when it belongs
Dither is normally applied when reducing fixed-point bit depth, such as converting a high-resolution mix to 16-bit. It should usually be applied once, at the final reduction stage. Repeated dithering, or dithering files that will continue through substantial processing, is unnecessary. Most 24-bit and 32-bit float stem deliveries do not require additional dithering unless you’re reducing from a higher fixed-point bit depth and the recipient specifies it.
32-bit float: useful but not universal
- Can preserve sample values above 0 dBFS for later gain reduction in compatible software.
- Can reduce the risk of fixed-point clipping during intermediate interchange.
- Creates larger files and may not be accepted by every application, portal or institutional system.
- Does not repair clipping that already occurred inside a plugin, converter or analogue path.
- Should be agreed in advance rather than used as an unexplained exception.
Common technical mistakes
| Mistake | Consequence |
|---|---|
| One stem at a different sample rate | Import conversion, wrong speed/pitch if misinterpreted, or drift concerns |
| Mixed bit depths without explanation | Uncertainty about source quality and export settings |
| MP3 or AAC stems | Lossy artefacts and reduced editing flexibility |
| Normalization during export | Changed balances and unexpected peak behaviour |
| Bit-depth reduction without appropriate dither | Potential low-level quantization distortion |
| Repeated conversions | Additional processing, possible artefacts and harder troubleshooting |
Recommended default when no special format is required
Common professional default
Engineer checklist
QC your stems in seconds — no DAW needed
Stem Checker loads and plays your stems in sample-accurate sync and flags length issues in your exports — mismatches, missing audio, duplicates, clipping, dual mono and more. A faster, more reliable way to review and organise your stems and multitracks.
Save time. Deliver with confidence.