I have decided not to continue testing/experiments with the "new" CD player, because I have an issue where I need to turn the 'volume' way up to get any sort of synch lock (I'd reported this before -- when I turn the contrast down I lose synch lock). But more to the point, even with the volume turned way down, doing a manual synch on the disk, and turning the brightness up -- I still get the way odd corruption on the image (the top of the black circle in the test pattern missing, for example).
So I'm going back to my "old" CD, and abandoning MP3.
I may still keep the "negative video" circuit, I think -- it could be useful at some later date.
However, there appears to be another problem. Originally I'd wired in the wire coming from the CD line out directly to the contrast potentiometer. To allow the "negative video" circuit to be easily plugged in, I wired in an RCA male/female plug sort of setup, so that I could plug the CD's line-out into the "negative video" circuit, or directly to the contrast potentiometer (via the RCA on each).
BUT, now that I have had a very close look with the "negative video" circuit NOT connected -- that is, connecting the line-out directly to the RCA of the contrast potentiometer, and turn the brightness down significantly, I can see some artefacts in the image. Artefacts that I'm pretty sure were NOT there originally.
I suspect a couple of things. One, the RCA connections might be a bit lossy, and the signal getting to the NBTV circuits is no longer "good". Or possibly there's even noise being introduced. More likely, I think, is that I've increased the length of the wire that the signal has to traverse. Or possibly, I have a very minor "leak" or short circuit in my wiring somewhere and I'm getting a bit of interference.
It will be an interesting one to track down. I did not spot this problem on a couple of other tracks that I looked at -- specifically, track 36 was at issue. I saw the problem mostly around the L of "32 LINE" when the brightness was quite low. It was a sort of cyclic 'fluttering' of the background image/colour. Now that I think about it, though, why should it be cyclic... that tends to suggest a problem in the original video/source.
Mmmh... well,more experimentation needed.