I've been looking into the performances of individual channels on the Cymac by computing their individual PSDs and corresponding CSDs that show their noise relation to each other. It appears some channels do have lower noise floors than others, and some combinations of these actually do perform similar to the Red Pitaya (showing below the CSD between CHs 3 and 5), although it doesn't look like it's much of an improvement. The best method as of now still appears to be phase-locking two separate ADCs to reduce the correlated noise floor further.

This can be further discussed at the July 22, 2025 group meeting.