Bad CIRC or some sort of DSP ?

[percy]

How common is it for CD players to perform incorrect data retreival from the disc ? like a bad EFM/CIRC decoding.

Recently I found out that the ripped copy(ucompressed WAV) of a particular track was not the same as the one that was recorded by playing the same track in a CD player and recording it via the spdif input of the soundcard.

Either the CD player is not pulling the values from the disc correctly, or it is doing some sort of dsp on the data before sending it out the spdif terminal.

The disc is not scratched or anything. The track on the cd and the ripped wav is 100% correct.

I tried ripping the track on a couple of different cdr drives and the files matched so its not the cdr drive used for ripping.

I did a loopback record of the file and the ripped as well as recorded files matched too indicating a perfect copy, so soundcard is not an issue either.

So what do you guys think is going on ?

oh btw, this is the Pioneer DV-563 "universal" player.

[oshifis]

I was curious about the same, and installed a LED in my Marantz CD-74. The LED is connected via a transistor buffer to pin 3 of the SAA7020 Error Corrector, that is the C1 uncorrectable error flag, and pin 28, that is the C2 error flag output. The C1 decoder can correct 1 error on a 32-symbol frame, and the C2 decoder can correct 2 errors in a group of 28 symbols. If there are more than 2 C2 errors, the Error Flag output is set. The interpolation and muting chip (SAA7000) then interpolates if a single errror between two adjacent samples are flagged, or mutes if two or more adjacent samples are flagged.
The LED is blinking very frequently, but there are also some CDs where it is dark. I think it is not the mechanism reading error, but the pressing error that causes the C1/C2 errors.

I just guess that the SPDIF output went through the above errror correcting process, but the ripped content is the raw data stream converted to WAV format. But I may be wrong.

[percy]

well c1/c2 error checking and correction happens in cdrom drives as well.
C1 errors - in more or less degree - always occur irrespective of the quality of the media or the transport. Uncorrectable C1 errors get turned over to the C2 error correction process. If C2 error correction fails the values are then interpolated based on the neighbouring "correct" samples.

The "quality" of the track that I am trying to rip is pretty good. It does not produce any C2 errors in both the cdrom drives and has pretty good C1 rate(less) as well.

so my question is -
Why is the C1/C2 error correction different between the CD Player and the CDROM drives ? Shouldn't it be producing the same result ?
Unfortunately I dont have another cd player with optical output that I can compare.

At this point my "guess" is that this could have something to do with the optics (laser and lens size) having to be able to handle multiple formats (CD, DVD, SACD, DVD-A).

btw, the "frontend" IC is STM6316 and the "backend" IC is STI5588 if that may shed any light. but I like your idea about the led. will check tonight if these chips have any pins like that.