But in this case, how do you explain the Ricoh scan ? the C2 rate goes up to 6000 per second, which was not the case when it was burned, while the C1 rate is still zero in very large aeras (no C1 error in more than 10 minutes). The Fnac gold scan from 2000 shows that it does not come from a mechanical defect of the burner that would give high error rates past 62 minutes. We would have seen a C1 increase near the end. I think that there is definitely something attacking the CD from the outer edge, and, in a lesser extend, from the inner edge. I sometimes experienced the same phenomenon, too ... and if these C2 errors haven't been there before (how could they ...), you surely are right about the process of media degradation (although I still believe that K-Probe does not work that accurate with a DVD-ROM). Some scans of mine show higher error rates at the start (no visible damage to the CD itself) so I can also confirm that some CD's are beginning to degrade from the inner edge ... A well-manufactured piece of media should show a somewhat consistant (and low) error rate throughout the whole disc after writing (like my Verbatim and Taiyo Yuden scans show) and a non-matching or non-properly aligned write strategy should have a constant impact on error rates. But what about tracking/focusing errors rising to the end? Nearly every piece of media (even highest quality E-Grade TY) shows higher deviations of FE/TE with Plextools Pro towards the end - a writer not being capable of following the track properly (or a writer not using a running OPC like modern drives do to adjust write speed and fine-tune laser power output) will surely induce a higher BLER towards the end - not even regarding the heat-up of the media during the burn process which (in my case) was some "problem" with the CRW6416S. This leads to an increase of errors towards the outer edge of the CD and, in conjunction with media degradation (which is measureable and definitely takes place) will cause media failure. The real problem is that we cannot know how the CD error rate relations (start/end - BLER) were right from the beginning (after the burn) since no comparable tests have been conducted (and no tools for the masses - except for early versions of Nero CD Speed - have been available) back then. What makes me somewhat doubt that this is media-related only is the fact that I regularly bought CD's with a friend - we bought the same media (I grabbed two ten-packs, he grabbed two). My CD's all failed (using a highly acclaimed quality SCSI write device) whereas my friend can still read every single CD (he did use some questionable Mitsumi ATAPI 4x writer). IMO, the Yamaha CRW6416S was far from being crap, but support was not that strong from Yamaha's side - I would have e.g. wished some guidance in what media to buy. In my opinion, the problems with older media that we face today are surely not entirely related to media degradation - high error rates after the burn on lower-quality media due to somewhat unlucky combinations of burner/media might introduce or accelerate the media degradation we encounter, disregarding the fact that we have been surely sold some media with badly-sealed edges (so any corrosive agent can diffuse into the CD between the reflective layer and the dye). But MPO as a european quality manufacturer with lots of ISO 900x certificates shouldn't have produced such bad kind of media IMO ... I'm glad that it is much harder to sell crap these days ... be it media or hardware ... thanks to communities like HA or CDFreaks, people can get well-informed before they actually buy.