Correct, the Realiser will be the master clock in such a setup, and both DAC8 will just follow it. The only unknown would be if there's a guarantee that the USB stack will have the exact same latency for both DACs, i.e. the Mac may be a few whole samples off when starting playback.That's correct but it's not really a big deal. On macOS it's set up as a 16ch aggregate device and one of the dac8 is the master clock. The other one has drift correction enabled. I don't have any problems with this. However, since the dac8 in AES/USB mode uses the clock signal it recovers from AES to also clock the USB data transfer, technically speaking it should be the Realiser that sets the effective clock for both units and drift correction probably isn't even needed. However that goes beyond my knowledge.
Edit: this has actually been discussed over on audiosciencereview.com, but I can't recall at the moment what the result was.
Last edited: