C2 Spikes on NEC 3540A

I have started seeing isolated spikes in the Nero Disc Quality plots. The C2 spikes are usually about 2000 high. Sometimes they are repeatable, but sometimes a second pass will be clean, or will show a spike in a completely different location, as in the first two figures at:

http://home.nyc.rr.com/dmcguire/_nero.htm

I see this on discs I burn, and on discs given to me by friends. I see it on cheap blanks, and also on good ones - the examples are from a Taiyo Yuden burned by a friend.

It almost looks like the spikes are in playback, or perhaps it’s just a problem with the NEC 3540A I use for these tests. My Lite-On drive doesn’t show the spikes (see third plot, on same Taiyo Yuden). But the Lite-On is generally much better at error recovery than the NEC, and if I want to simulate what a typical player (car, discman, etc) will see, I try to use the NEC to run these tests.

Are these spikes real, or are they phantoms? I have listened to a few of these, on a boombox or on the NEC, and I don’t hear any problem.

Thanks in advance.

Your LiteOn SHM-165P6S is virtually useless as a CD quality scanner, so lets ignore it when it’s telling you that everything is fine. If you want to know why, read this thread.

The spikes you’re seeing remind me of spikes I occasionally see when using my NEC ND-4551A as a scanner.

After comparing scans between my NEC 4551 and other drives, I have come to the conclusion that some spikes are due to a “real” problem but many spikes are just “phantom”.

If a Data CD really had a spike of 2000 C2 errors, the disc would have a spot that’s difficult or impossible to read, so you can check that by performing a Read Transfer test or a ScanDisc. If there’s no reading problem, then your NEC 3540 is either imagining things or at least exaggerating wildly.

An Audio CD could have a real spike of 2000 C2 errors, and your drive might still read it because it’s designed to read the audio samples even when they’re damaged. If you use a “secure” ripping mode however, your rippng program probably wouldn’t be able to read that spot correctly.