Actually, I need to come a little bit cleaner about the "zero-degradation" concept. While a CD-R should, in theory, be a 'bit-perfect' copy of the CD, I in fact do find small listening differences between original CDs and copies I burn. I do not know how to explain this, as a dropped bit here or there shouldn't be audible as an overall change. I might attribute this to something having to do with the playback optics' ability to read a CD-R accurately compared to a real CD. When I auditioned the Marantz CDR-500 vs. the HHB CDR-830 (which I bought - see the archived article in my threads), I found the HHB to not only outperform the Marantz with analog sources, but also to my surprise on digital dubs, all using the same blank media and playback reference system, leading me to wonder whether the HHB had a superior burning laser (which might result in CD-Rs with better readability). Jitter isn't a factor until the bits are transduced by the DAC, and so shouldn't affect the digital dubs (and indeed, I haven't found that my otherwise effective jitter-reduction box has any influence on the finished product when inserted for digital dubbing). But in general, with revealing source material, I do hear very slight reductions in transparent clarity, HF extension, tonal color saturation, microdynamics, soundstage separation, image body, and background 'blackness', when A-B'ing a copied track against the original (or put another way, the copy sounds a little veiled, thin, flat, and lacking in life by comparision). Subtle differences, to be sure, but I believe real, though I haven't performed blind ABX tests or anything. But again, I do not know if these differences are to be found in the encoded data itself, or are an artifact resulting from transcription difficulties. Anyway, for now the upshot is that the original still sounds best by a small margin in the reference system, though it makes no difference in my car, and is not enough to dissuade me from playing my custom comp's at home either ('course, most of my source material tends to be analog, making this an oft-moot question).