Well, I don't know the architecture of the processor, but in essence I don't know how the analog signals could mix with the digital ones once the analog signal hits the A/D; all of the processing within that section is digital. The most likely scenario is the delay and add approach. That is, the only way that I can see crosstalk occurring is via delay and add; I can't think of a way for the analog signal to get into the digital signal (and indeed, they are two completely different domains).

It's funny because I did in fact try to remove the optical feed from the source when I first noticed it (leaving the analog connected - see thread's origin)...and soon I discovered that it wasn't he source (verified by using yet anotehr analog source). Again, per the post, I have routed the analog "CD" input to zone 2, and this is the input that I used with the stand alone analog source, bot to listen and confirm, as well as make the recordings.

If you want to know what to listen for, you could email me, and I could send you the mp3's (they are 320 kbps so they are very, very close to the original wav files). If you listen to these, you will know immediately what to listen for. I could show you the plots as well if you are interested. One thing to note - you may need to slightly adjust the voluem of the mp3's when you play them, because I didn't botehr to normalize their levels. Still, if you compare the mp3 from zone 2 to any of the ones taken from the front (stereo, bypass, upsample) the difference is immediate and unmistakable.

Thanks for being willing to check into this - I really do appreciate it.