Very late to the party here, but just wanted to echo everyone's sentiments that having Rob participate to the breadth and depth that he does is hugely appreciated. I had the pleasure both of meeting him and hearing Dave at the Audio High event recently (and thanks to Hiyono for cross posting my brief review of that evening at #567 in this thread). From what I heard that evening, and even in the very challenging circumstances that prevail at those events, posters here who have "pre-bought" Dave won't be disappointed. The ability of Dave to bring back depth to the hi-fi experience was very noticeable, and something I heard even before Rob's talk went into more detail, in particular about where the technology was different relative to his earlier work on Hugo and hence highlighted what he learned during its development.
I will be very interested to see the results of Rob's further investigations here. Initially, I was a bit surprised that it was improving the SN ratio through adding further high-order noise shapers that seemed to be the thing that made the most difference in this area.
If you think about audio depth perception, numerous studies have shown that it's driven by the ear hearing time-delayed reflections from the original sound source interacting with the environment. Indeed, if you sat in an anechoic chamber with one fixed and one moveable sound source, moving one source further away would just make it sound quieter, not more distant, which is not what happens in real world acoustics. In practice, environmental reflections reach the ear both delayed and significantly attenuated, with both parameters providing clues about the acoustic in which the sounds occurred. The brain then does its own remarkable thing and creates the impression of a soundstage that we can bring to mind even with our eyes closed, one that can be intimate or expansive, studio or stadium. But this attenuation also means that the depth cues are, therefore, akin to signal-modulated noise, heard by the ear some time after the primary signal. Therefore, I went away from Rob's talk thinking that it was his work in removing the tendency of a DAC to itself modulate the noise floor of the signal chain with the source waveform that was improving depth perception, but that doesn't seem to be the case.
Of course, we live on the world of engineering here and not pure theory, so it may just be that it's in the interaction of the noise shapers with the - now - signal-independent noise floor that makes the difference; pull down the quantization noise far enough and it enables other engineering parameters to become evident. Regardless, my speculating won't answer the question but Rob's experiments can, so I hope he will keep us in the feedback-loop as he gets closer to un-picking the complex, interwoven threads in this area.
Looking forward too to hearing first impressions as people get to plug Dave into their own systems, as well as what comes back from professional reviewers once they get to do more of a comparative analysis than most of us can do from our own resources.
Exciting times indeed, and thanks again, Rob, for coming to visit and spending time chatting! As an expat Brit it's always fun to meet visitors from the UK, especially when they come with 20+ years experience in digital audio design
.