Are the Jtest measurements loopback? With a common clock some of the jitter will be suppressed. This will help interpret the jitter: https://www.bramjacobse.nl/wordpres...21/04/TN-118_Measuring_Jitter_with_J-Test.pdf The spectra don't show the 250 Hz stuff. Maybe a digital loopback will show better what a jitter free display would look like. Usually on the 16 bit stuff there is a pattern at 250 Hz intervals:
