Can someone explain to me in a few sentences what dither & dithering is and if I need to use when exporting my mix?
A member on a internetforum where mixing engineers discuss with each other asked this clear question. Obviously he had googled around to find an answer, but all the information he could find wasn’t clear to him. After reading long and often technical stories, he had gain little or no knowledge on the subject of dither. The software he uses for creating his mixes has an option to dither a track, but what should be done with that option?
I replied to his question and according to his reaction, this explanation is clear. Hoping to help others, here is my reply.
What is dither?
In digital audio each sample is stored in a certain amount of bits. On a cd par example 16 bits per sample is being used. The higher the amount of bits per sample, the more accurate the audio can be. We call this bitdepth.
When we are working on a project with 24 or even 32 bits per sample, which we need to master for an audio cd, we will need to lower the bitdepth to 16 bits per sample, as required by the CD standard. The reducing of the bitdepth will introduce rounding errors, which are predictable and will give a distortion in the audio,
Dithering is intended to prevent this, and is the addition of a soft random signal (noise; dither), making the rounding errors are not to be predictable anymore. Samples will still need to be rounded (you can’t avoid that when lowering the bitdepth), bu by adding dither the rounding errors at least will no longer be predictable.
Why should I need dither?
The human ear is much more sensitive to the distortion that is introduced by the rounding errors than the noise that is added by the dither signal.
Dithering is only needed when you reduce the amount of bits per sample.