DATBRAD, I make that kind of test by programmatically generating a sequence of digital samples whose values follow a predictable mathematical sequence. I store them as a wave audio file and burn that file to a CD.
I then play that CD on a player that has a digital output, which I connect to the digital input of the device under test. Then I record the result and look at the sample values in the second file. If they follow the pattern that I programmed into the first file, then the signal wasn't resampled (or was resampled perfectly, such that it doesn't matter).
I used a similar technique about ten years ago when I set out to evaluate computer CD-ROM drives. I wanted to know what really happened when people capture ("rip") digital samples from audio CDs at high speed--which, back then, was only 8X instead of the 50X and higher of today. I saw considerable skips, interpolations, samples repeated, arbitrary channel reversals, etc. on most of the drives that I tested, including "name brands." It's a fact that a lot of people apparently either don't know about or refuse to think about.
--best regards