Continued from other discussion. *will add link
Here’s some reading, I will add more when I can.
3 Likes
From that PF “Mixing in Pure DSD” article quoting NativeDSD;
Over the past several years, Tom and NativeDSD have worked with Jussi Laako at Signalyst to use Jussi’s HQPlayer Pro (HQP) software to mix DSD tracking channels. The mixing is done in HQP through a process called “modulation.” This process keeps the signal entirely in the DSD domain.
Hmmm, I don’t think so 
I’m pretty sure HQP is converting the DSD signal into some multi-bit format internally then applying whatever mixing changes, before remodulating to DSD.
My guess is Tom doesn’t know because Jussi has never commented on exactly what’s going on inside HQP Pro when used for DSD mixing/mastering. Even Sony’s Sonoma workstation which was the original platform that touted “native” DSD mixing actually converted to a multi-bit format internally before remodulation to DSD. My guess is HQP Pro is doing exactly the same thing.
Some additional reference material;
1 Like
Me neither, I’ll try and find some more indepth math/signal processing oriented explanation. For the mastering engineer it’s honestly just a toolbox/black box, which is fine because they do what they do best and decide what sounds better to them, but not necessarily why.
https://www.jstage.jst.go.jp/article/transinf/E107.D/8/E107.D_2023LOP0006/_pdf/-char/en
So apparently according to this last more recent article it is possible to stay in the delta-sigma domain doing transformations while keeping noise-shaping. Thing is if that can do the operations that the mastering engineers need and if it then sounds better than the alternative which is conversion to dxd and editing in dxd pcm.
Interesting paper, and quite recent too, thanks for that. Even more interestingly, all the key references to the processing “[5]-[8]” are from the early ‘90s! Which suggests for whatever reason, no one’s really using these methods for native delta-sigma digital signal processing?