Dithering is a technique applied when reducing the bit depth of digital audio (e.g., from 32-bit float to 16-bit for CD). Without dithering, the truncation of samples introduces correlated quantization distortion — unpleasant artifacts especially noticeable on quiet passages.
By adding carefully shaped random noise before truncation, dithering converts this correlated distortion into uncorrelated noise — which sounds far more natural and is less perceptible.
When to dither:
- Exporting a final master to 16-bit (CD, some streaming)
- Any time you reduce bit depth as a final step
- Do NOT dither between processing stages at the same or higher bit depth
Types:
- TPDF (Triangular Probability Density Function) — the standard; flat noise floor
- Noise-shaped dithering — shifts dither noise to less audible frequencies