The CAPS 3.0 runs on battery power.
OK, I didn't know that. The next step improvement (according to our tests) is to supply each rail (12V, 5V 3.3V) from individual battery. The ultimate config might be to bypass the VRM switcher (Voltage Regulator Module) & supply direct power to the relevant electrical points? I don't know, I'm just surmising - it may be a case of diminishing returns here?
It may be a case of the various PC supplies are cross-contaminating noise between each rail & separate supplies reduces some of this? This, to me, is the problem with computer audio & why there are fights about it - what's acceptable for digital functioning is not necessarily OK for digital AUDIO functioning! So most mobo are built to fulfill the brief of digital processing functioning & no consideration given to audio. We are left with working with tools (PCs) not designed for the purpose (audio) & is it any wonder that different software players sound different, that different PS approaches sound different?
But I2S Pro reclocks using the DAC clock. I'd be shocked if this were inferior to outboard reclocking, since folks using I2S Pro report getting the best sound out of their MSN DAC. In fact, the MSB should be switched to reclocking mode for synchronous sources (and asynch USB), which reclocks inside the DAC as well, so with MSb reclocking outside the DAC is never the best architetcture.
Yes, I saw your answer to Vincent & concur that timing issues seem to be dealt with using the config you outlined, if I understand it correctly.