@Frnch - It is very good to see someone arguably crazier than me trying different things with a 1000Pro! One thing I have learnt over the last year is just how much very subtle changes to the source can influence the subjective impression of the sound. Switching between two bit-perfect sources, the microRendu to the sMS-200Ultra, subjectively made a significant difference tonally to my system. Then we had another tonal shift with the Neo upgrade of the sMS-200, and apparently this is a capacitor change in the power supply. The point is, these very small changes appear to make significant differences to how the system sounds, and ultimately how enjoyable the system is to listen to. Of course, @flohmann is correct, you are using the Devialet to perform analogue to digital conversion, then another digital to analogue conversion in the ADH DAC, so you are still listening to a Devialet DAC, but the ultimate source has been changed (improved?) by the M Scaler upsampling. I am aware of the magic that the "1 million taps" M Scaler is reported to produce. This is a form of hardware upsampling, and a very fine form of upsampling at that, which I am sure can do something that is less subtle than a capacitor change in a network endpoint. So despite the fact that this goes against the purist "shortest signal path" approach, I can see that this would sound subtly different, and potentially in a way that sounds subjectively preferable. It reminds me a little of listening to the dCS Upsampler with a D800, this sounded better when upsampling to DSD 64, which makes no sense when the Devilaet's DSD "MAT" algorithm is converting back to PCM for the ADH. It is also interesting that the dCS hardware approach sounded superior to me versus doing similar with HQPlayer.
So in other words, you can have an approach which appears technically superior, but sometimes doing something slightly different can yield superb results, generally for reasons nobody seems to fully understand, probably in the realm of psychoacoustics.
With all that said, can I ask if you have tried your M Scaler feeding the Devialet via the optical connection? The reason I ask this is that with my REF10 / Mutec MC3+USB combination, I have found the best results when feeding the Devialet 24/192 via HQPlayer upsampling. One curiosity of the M Scaler is that it can be set to upsample to 176.4 via the S/PDIF output, but you can only fix to the Devialet preferred 24/192 via optical. OK, I know that some people think Toslink is not preferred for jitter reasons, but the Chord optical out is reportedly excellent, you get full electrical isolation, and you can set your M Scaler to deliver the Devilet preferred 24/192, upsampled from 16/44.1 or whatever else your source material might be. Have you tried this?
So in other words, you can have an approach which appears technically superior, but sometimes doing something slightly different can yield superb results, generally for reasons nobody seems to fully understand, probably in the realm of psychoacoustics.
With all that said, can I ask if you have tried your M Scaler feeding the Devialet via the optical connection? The reason I ask this is that with my REF10 / Mutec MC3+USB combination, I have found the best results when feeding the Devialet 24/192 via HQPlayer upsampling. One curiosity of the M Scaler is that it can be set to upsample to 176.4 via the S/PDIF output, but you can only fix to the Devialet preferred 24/192 via optical. OK, I know that some people think Toslink is not preferred for jitter reasons, but the Chord optical out is reportedly excellent, you get full electrical isolation, and you can set your M Scaler to deliver the Devilet preferred 24/192, upsampled from 16/44.1 or whatever else your source material might be. Have you tried this?
1000 Pro - KEF Blade - iFi Zen Stream - Mutec REF10 - MC3+USB - Pro-Ject Signature 12
