Continuing from RBJ’s point, the MQA music codec uses subtractive dither throughout the codec. MQA is the work of Peter Craven and Bob Stuart, Peter being the originator, along with Michael Gerzon, of the buried data method cited in RBJ's reference below. MQA makes extensive use of hidden data buried in pseudo-random noise along with signalling bits.
The HDCD codec ( developed by Keith Johnson and Michael Pflaumer) also used subtractive dither but in a more limited way. It was used to linearize the converters as I recall. But these are both codecs, in other words highly controlled processing from encode to decode. Vicki Melchior > On Nov 19, 2021, at 9:34 PM, robert bristow-johnson > <r...@audioimagination.com> wrote: > > Sampo, you should look into this buried data method of creating dither that > is subtractable. > > The late Michael Gerzon https://www.aes.org/e-lib/browse.cfm?elib=7964 . > > I can get a copy for you, just email me. > > -- > > r b-j . _ . _ . _ . _ r...@audioimagination.com > > "Imagination is more important than knowledge." > > . > . > .