Can you define enormous and how it impacts audibility?
My reply was to what differences I
see in optical read mechanisms as posed by
@mkawa . The visible differences are enormous. Some RF (eye patterns) are a wasteland. By the time the signal is demodulated, buffered, error corrected/interpolated those differences can be non-existant or still present in the form of heavily interpolated data or even (ramp-down-ramp-up) mS semi-mutings. That is what testing uncovers.
High frequency pure sines are the easiest to see interpolation effects- the THD skyrockets. Don't forget, the digital filter and D/A converter just process the incoming bitstream. To those chips, the signal is golden, they don't know if its accurate or not- they just do what they're told and convert it.
But for testing CD players, you are testing the system as a whole. It's abilities with actual discs. For that, you need absolutely perfect, unmarked discs pressed to unwavering tolerances. They were very expensive at the time and impossible to find now. You also need discs with deliberate errors, eccentricities, pit-land length errors, data interruptions etc to accuractly determine the performance of the optical/servo/tracking/error correction and other systems.
Just because a disc plays, doesn't mean it is reading 100% perfect data.