wuwulf10 posts01-15-2021 3:48am01-15-2021 9:37amDear audio2design,
but
one more thing I believe is rather curios. If digital is defenitely
superior, why than there are so many different solutions? Like using a
chip form a manufacturer versus programming your own chip. NOS versus
DAC with Filter. PCM versus DSD. Upsampling vs No Upsampling etc.
Sometimes I feel the dissonance beween different digital opponents is
bigger even than in analog where you have the dd versus rim versus belt
discussions.This uncertainty about how to process digital best does not
neccessarilymean that digital is inferior. What it shows to me is:
solutions in digital or analog are easier to build than to be explained
:-)
Sonic differences between what I will call technically equivalent DSD and PCM virtually all come down to implementation details.
Other than that, most of what you are mentioning is either implementation detail and/or an intentional euphonic manipulations and moving away from accuracy. I am not at all against that if you like the outcome.
NOS DACs at redbook rates are not technically accurate. The are fraught with both audible band artifacts and near audible band that can subharmonically isolate. You can put a brick wall analog filter, but then you have other issues.
Upsampling a NOS DAC with a standalone upsampler will fix most of the ills of a NOS DAC. However, if you prefer one sound for one genre and a different sound for another genre, it makes total sense.
Every chip DAC designed in the last 20 years, and most high end standalone DACs where the DAC is discreet upsample. Sometimes they upsample a lot. In most cases that is very good thing. It makes analog filters simple and easy, and improves SNR in the audible band. It's not going to overcome bad design. There are some DAC chips that measure better at lower sample rates, but not so much modern ones.
Modern DACs also have different filter setting. Most of these "new" filter settings are technically less accurate. Some prefer the sound.
Other implementation details come into play like how well they reject noise on data / electrical connections, and how well they reject jitter on optical/co-ax connection. A virtually jitter free isolated USB interface is not rocket science.