Volume / Gain could be synchronised - but you would get random phase shift between "channels" as the clock circuits / ASRC are not locked to each other.
I started writing an even longer post (!) in reply to this, but I've run out of time to complete it now - I might get a chance to follow up with the detailed reasoning, but in summary:
- I don't think that's particularly important because...
- the DACs all individually have very stable clocks and the same ASRC performance characteristics
- the ASRC aspect is effectively happening at frequencies much higher than the data rate
- and it is ASRC and not synchronous reclocking
- the individual ASRC circuits will have very similar (the same?) behaviour to the incoming signal as each other
So I don't think there's any problem with relative drifting of the ASRC between different DACS - they're all to the same high standard. I don't think any phase shifts will be random as in 'jitter' random.
- the DACs are all receiving the same source of SPDIF signal (possibly they could be receiving the
same multi-channel SPDIF signal - ignore that for the moment)
- the (spdif) source - the DSP crossover - will have have a single high performance stable clock driving high performance SPDIF outputs.
So we have the limitations of SPDIF but nothing worse than any other SPDIF source. And all the SPDIF outputs are driven by/synchronised to the same source clock so to some extent variations will/could be common across all SPDIF outputs at the same time.
But overall, whilst it
may be better to have one universal clock, running independent DACs and leaving them each to sort out their own ASRC is the lesser of many other potential evils:
- Active overcomes so many limitations in the speaker/crossover/equalisation chain that the missing 0.X% of performance that
could be further achieved by having a single master clock is outweighed by the greater benefits of an active system with independent ASRC recovery in each DAC
- The alternative is to choose/be limited by whatever DAC implementation has been chosen as part of the DSP design and I think that has another set of compromises
If we think there's a problem with DACs wandering relative to each other then an interesting experiment would be to feed the same high quality stereo SPDIF signal simultaneously into 2 MDACs and use one for each channel. Is there any degradation in sound quality because their ASRCs aren't locked? My feeling so far is that there wouldn't be.
Alternatively, there may be arguments about whether to use 2 stereo DACs horizontally or vertically. Eg if locking the performance/clocks between the L & R HF channels is determined to be important, then horizontal 'bi-DACing' would provide the best results
What is the delay on the spdif output from the MDAC? The experiment could be run by taking the (high quality de-jittered) SPDIF output from one MDAC and feeding into a second MDAC. That may overcome any concerns about variation between the different SPDIF input signals
PS. the point about synchronising gain between multiple MDACs rather than doing it in the DSP is because I believe that the right point to control gain digitally is at the output stage using all of the Sabre's 32 bits
(into ML Ascents)
