tigre

Hi there!

Trying to rip "Grönemeyer - Mensch" (CDS 200 protected) there are several problems - most of them are solved, but one remains.

------------------------
First the solved problems FYI:

- corrupted TOC/data Track in 2nd session:
I've tried 4 drives so far. With 1 drive (no-name) the CD is not readable at all, no matter what I try, with 1 Drive (liteon 8x burner) all tracks are shown correctly by EAC, 2 drives (LG 12x DVD + Plextor 121032A) need "special treatment". This means either "detect TOC manually" or the marker / post-it trick.
I decided to use the post-it trick - so far so good.

- C2 errors and interpolation:
The CD is in near-to-perfect condition, so most tracks can be ripped in secure mode with all drives without error-correction kicking in. Test+Copy CRCs for each drive are always identical (1 exeption: later).
Comparing the results between the different working drives showed different CRCs, wave substraction showed small bursts of different samples. The differences were that big that the liteon drive produced audible clicks. On most parts of the tracks there were no audible problems with the lg & plextor, but I found at last two positions where I could hear a slight click produced by the lg drive while the plextor sounded OK. So I decided to use the plextor for ripping as this drive seems to do the best job on interpolation.

- gap of silence problem:
At "normal" secure mode settings (speed: actual, cache on, C2 off) the plextor drive produced a short annoying gap of digital silence (~0.xxx seconds) somewhere in each track.
Lowering extraction speed solved this problem.
---------------------
The remaining problem: The last track, > 20 minutes long, containing a 3:xx minute song, silence and a 3:xx minutes "hidden bonus song" is ripped "error-free"/no suspicious positions, but everytime I rip this track the CRC is different (using the plextor 121032A drive). I've tried lowering extraction speed, burst mode ... it's always the same. If I compare the .wav files from different extractions, there are 1-5 different samples at various locations (always in the music parts of the track not in silence). There are a few positions where EAC's error correction kicks in (red lights ...) but it seems to happen at different positions each time. It's not really something I can't live with, because there are no audible problems. I'm just curious about the reason for this. Why can all other tracks be ripped in secure mode, 4x speed with matching Test&Copy CRCs and without audible problems, but the last track can't, even at 1x speed?

Cheers tigre

Halcyon

1. You have Plextor

2. Download Plextools 1.21

3. Select First session only in Plextools for your Plextor drive

4. Select DAE extraction speed at 1x (not any faster)

5. Insert the CDS 200 protected disc and proceed to rip to WAVs

6. Listen to the tracks yourself to hear whether they are OK

I have used this method successfully on my 121032 SCSI Plextor and so have others. Later models of Plextor can rip up to 4x speed without errors, but at least I couldn't do any faster than 1x.

regards,
Halcyon

PS Plextools will automatically re-read erronous parts, select the best possible of multiple reads and adjust read offset automatically.

tigre

Hi Halycon!

Thanks for your input. I already have Plextools and use it once in a while. The problem with this CD and Plextools are the mastered CD errors that make extraction painfully slow (several hours for 1 track) - the same with EAC - C2 activated BTW.

I'm satisfied with the result already (BTW - I've done ~15 extractions of the last track now and got the same CRC 3 times). There are no audible problems.

I didn't ask for a solution (like "How to get a better/more secure rip of this CD?") but for a (rather technical/theoretical) explanation why I get different CRCs only for 1 track.

I know of the phenomenon of not matching T & C CRCs in properly configured secure mode if a CD has slight scratches, it happened to me once too, but in this track it happens on different positions ... Is it possible that they tried a newer version of CDS copy protection for this last (bonus) track that causes this? How could this work? ... Or am I paranoid and don't notice some slight scratches?

Halcyon

tigre,

I'm sorry, I didn't read your post properly. My fault

Your question is indeed interesting and I have no answer.

I would suspect a small scratch myself, but can't know for sure.

You could try (if you have time) and go out to buy another copy of the same disc and see if it behaves similarly for that erroneous track. It's not available in Finland, so I can't test it myself.

You can probably return the second copy (or the first one if you so desire) to the store as an unplayable cd not conforming to CD standard. You should be able to get your full money back.

At least we can do that in Finland and the store absolutely MUST take them back, no questions asked.

Funny thing that, copy protections actually make it cheaper to copy CDs, now that you can copy them and take them back for a refund.

Somehow I just don't understand the record industry...

regards,
Halcyon

Flanagan

Maybe you could find some answers in this link.
Its german, though.

http://www.pcwelt.de/ratgeber/software/15859/4.html

In concluding words:
The CRC is calculated on basis of the digital information stored on each disc, which stay the same, unless physical harm is done to the disc.
But it depends on your drive which results you get, because it calculates the pre- and post-gaps which are insatlled by the drive and may differ from read to read into the CRC.
The more equal the CRCs, the better quality is your drive.

@halcyon
In Germany usually the RecordLabels have CDs without CopyProtection. Although they never go out to stores they usually replace the copy protected one with a cd w/o copyprotection, but only if you ask.

Greets, Tick Raw

Halcyon

Flanagan,

thank you for the enlightening information.

Do you know if anybody has done tests of various drive's CRC quality using CDS 200 discs?

regards,
Halcyon

Pio2001

The point of CRCs is to completely change as soon as one bit is modified. There is no CRC "more equal than another". It must mean "the more CRC are equal". But if it is an offset problem, the CRC are equal if and only if the offsets are corrected, exept the first/last ones, one of which being possibly different if the overread is not working properly. Again, it matches or not, and it is not related to the drive quality, "unless physical harm is done to the disc".

In CDS200 CDs, CRC, even offset corrected, are different from a drive to another, because there are intentional uncorrectable C2 errors, and different chips handle them differently : uncorrectable for one drive is not necessarily uncorrectable for another, thus possibly different consistent readings on different drives, thus possibly different consistent CRCs for each drive, as Tigre gets for all other tracks.

The errors in the last track can be an average pressing quality, pushed into the uncorrectable domain by the error correction codes being destroyed by CDS200.

__________________
Pio2001

tigre

So this can be considered as kind of measured evidence that CDS200's C2 errors cause quality problems because "real" errors start to become uncorrectable earlier. - Right?

Halcyon

Pio,

let me get this straight, so I don't lose you behind the language barrier:

1. CDS200 C2 error code manipulation results in deterministics results (with one particular drive) from one read to another

2. If CRCs (from, say EAC) for two different rips are NOT equal on a CDS200 protected disc, then there is something else wrong with the disc than just the CDS200 faulty C2 codes? Of course, I'm now referring to two rips done with the same drive?

Just want to make sure I understand you correctly.

regards,
Halcyon

Pio2001

Yes, as Tigre rips show.

Said like this, yes.
But it doesn't necessary mean that the pressing is defective, because there are things in CDS200 that we don't understand well. For examlple why a given Plextor drive gives silences at high speed and not at low speed. If for a given reason the test and copy are not at the same speed, the CRC will be different because of CDS200. It shouldn't be related with C2 uncorrectable errors introduced, but since I don't know what's going on here, I can't be sure of it.

__________________
Pio2001

Corsair

Maybe that Plextor drive just can't correct errors fast enough (at least some of the errors), if the speed is too great - and in those cases the drive might insert silence.

Andre Wiethoff

Yes... All C1/C2 correction circuits usually work deterministic, from which follows that the result on the same (wrong) C1/C2 information will always be the same.
But these CDs are like already scratched CDs (with many read errors). So there is no buffer for some wrong symbols on a sector anymore. Any new scratch/hair /whatever will cause a real read error with a much higher probability!
So, do you like to buy a new CD that is already scratched?? That is what you have done!

cu, Andre

tigre

If I would have bought it.

Anyway thanks for your answer.

Halcyon

Call me thick Pio, but I don't understand your logic

Tigre wrote:

"but everytime I rip this track the CRC is different (using the plextor 121032A drive). I've tried lowering extraction speed, burst mode ... it's always the same. If I compare the .wav files from different extractions, there are 1-5 different samples at various locations (always in the music parts of the track not in silence). There are a few positions where EAC's error correction kicks in (red lights ...) but it seems to happen at different positions each time. "

Doesn't this imply (assuming now that the pressing is OK and there are no scratches) that the way CDS200 destroys CD Audio information may result in non-deterministic reading results?

I mean, how is it possible to get different results with every read, using the same drive and settings?

Either

A) Disc is defective (pressing bad, scratches, smudges)

B) The drive is bad (it's not deterministic even on totally faultless discs)

C) The copy protection causes some things to be undeterministic in the read and/or C2 correction process

D) Or some combination of the above together

I mean, does somebody know the itty bitty details of CDS200 so well that one can for sure say that it destroyes the readability of an Audio track in a deterministic manner in all drives?

Sorry to drag this along, but I don't understand how Tigre's results are any proof one way or another. Maybe there is something I don't know / understand?

regards,
Halcyon

Pio2001

No problem.
In fact, this discussion started much earlier and you can review the begining here :
http://www.hydrogenaudio.org/index.p...370#entry54479

Tigre said, about all the tracks that return a constant CRC This gives in average
(893-845)/(5-1)=12 samples between each stereo error.
This burst is 460384-459820=564 samples long, thus it has
564/12=47 errors in each channel in average, that is about 100 errors in a burst.

264600 samples are 264600/44100=6 seconds, thus there are about 100 errors every 6 seconds in all tracks but the last.

If the CD is 80 minutes, the last track being 20 minutes, there are 60 minutes of repeatable errors, that is
60*60/6*100=60000 errors, everyone of them being exactly the same, reading after reading

This is the proof that CDS200 errors are repeatable. There is in addition something else going on in the last track.

This can be a CD with a high error rate : each part can be read with confidence after some trials (errors happening "at different positions each time") , but since the errors are many, there is very little chance of getting all of them right. As soon as you get one part right, other parts get wrong again.

CDS200 errors are very localized, so they destroy the error correction ability for very small parts around them. When an uncorrectable C1 error falls in such a place, it should not be able to be corrected by C2, weakened by the CDS200 error.

It's probably what's happening in Tigre's CD.

__________________
Pio2001