Thanks for your input, I will read the suggested subjects above. Quality discs is an adequate reference. I still prefer an actual, functional sequential read at max speed, and then compared using DOS. And then the option to repeat the DOS FC at lower read speeds from the drive, sequentially, if the FC fails. Max is 20x right now but 22x is available (I do not have this yet) and these speeds can be used to determine quality burns. Much like the concept of using "Picky drives" as referred to in this thread, using higher speeds may accomplish the same thing to determine quality of the burn, for any specific drive.


Probably, true, but I have seen a Scandisc turn out different every time I scanned the same disc. Not so for the TRT on the same disc. My interpretation: probably a bad drive. Also I am referring to a time referenced test. A week, a month, a year, whatever your preference. Speed and Time are both required in the equation to determine a true quality burn. Dq scans will not encompass the time factor of the equation.

I'll check this out. I have a Terabyte and still growing. Not too much of a problem.

Thanks for your input, I'll check the points you noted.

It's unfortunate because I think my disc is dying fortunately I do have a backup, somewhere.

And it IS a Data disc. And it is clear that bit failures have started pressing the redundancy margin into service.

Every drive I've scanned this Data CD-ROM with shows damage.

BTW, here is the Quality Scan you asked for.

Eeeeer.. no. While apparently logical at first, this is not true. If you put your mind to it for a moment, you'll realize why.
I'll give you a hint: picture youself testing a half-size burn (~2GB).

@SeanW: you had me re-testing 29 marginal burns and cross-check TRTs vs. Scandisc for each one of these discs. This took some time, and explains why I come back only now.

Well, I could not reproduce something similar to what you posted. The TRT has (as usual) been the pickiest of the two in these cross-checks.

I think we would need to know exactly what CDSpeed reports as "damaged sectors" in Scandisc's read test before going any further... and I don't know! Until now, I thought it was slowdowns that were marked as yellow....

I'll do some research, and if I can't find an answer, I'll try asking directly Erik Deppe. Of course, if someone around knows this, well, welcome!

So we all need to buy the worst reader on the market. If the disc is readable in that it will be readable everywhere.
The worst readers I presume are the NEC/Optiarc drives, right?

For the pickiest / most paranoid of us out there, who want only the absolute top of the crop, that's the idea. Not necessary for everyone, though: who needs the absolute best possible compatibility these days? Also a really bad reader would give too many false positives. So not really the worst, but very fussy reader, yeah. Nope. That's a reputation owned in 2006 and may well be, finally, as sound as a hoax. NEC drives from the 3550/4550 line are indeed good companions for TRTs in my book (not very good readers, but not terrible ones either).
3500, 3540 and 3520 are actually good readers. I know that the ones from the 7200 line are excellent readers. In-between NEC drives (7100 line for example) I don't know, we'd have to ask Dee-27. But the thing is, it's not a clear cut with NEC, much less than the usual rumours anyway.

My pickiest drive here is a Lite-On 16P1S DVD-ROM. I've heard some ASUS drives are terrible readers.

Maybe those Asus driver based on Pioneer or Pioneer drives in general.
those are fussy readers
I don't know about Optiarc 7200 but those 5200 are very bad readers.

Looks like you'know more than I do about some NEC lines.

The other thing to remember with testing (and this is valid for PIE/PIF testing as well), is that there are many different reasons why a drive can have problems reading a disc. Tracking, reflectivity, localised defects, unburnt spots, jitter... and unfortunately (unfortunately for testing, but fortunately when recovering data!), not all drives are sensitive to the same problems!

For example, I noticed (and I'm not the only one BTW) that while BENQ/Nexperia drives are very sensitive to high jitter, they're rather impervious to poor tilt (flatness). On the other hand, my NEC 3540 units are totally impervious to very high jitter and low burning quality, but have huge problems with cheaper discs with poor manufacturing quality. The list goes on...

Nothing is ever simple. I've searched a lot for a perfect and simple testing method, and found none until now. Combining low-speed PIE/PIF scanning, high-speed PIE/PIF scanning and TRT, all this in 3-4 drives, is not that bad, I'm happy with it, but sometimes I dream of seomthing simpler than that!

The neat thing with TRT is that as it's a real-world test, you know that if it's successful in any drive, you can count on this drive to retrieve the data later if the disc show poor compatibility in other readers. With PIE/PIF scanning, unless the scanning drive reports POF accurately (something that doesn't seem to be the general rule ), there's always room for doubt.

I THINK it checks the total data per byte, including both the data byte and the correctional margin, to see if there is correctable failure of individual bits.

Did your "marginal burns" include both CDs and DVDs?

Because I've only seen full-speed Yellow run-overs on damaged CDs - dodgy DVDs just tend to straight to Red.

If you have marginal CD burns, you might check them now.

As for the screenshots I posted, I have had other CDs like this but like I said, I binned them after the scan.

Agreed, big time. I also use regular (once a year) testing for all my important discs, but through TRT. Also PIE/PIF rescan of enrolled samples from each disc/burner/speed combination. I'm probably the most obsessive CDFreak when it comes to disc stability. It takes me something like 2 hours a week to conduct all the tests needed to survey my collection and get peace of mind. *LOL*18 DVDRs and 11 CDRs... maybe it's a thing with the drives used for testing, or the type of burning "defects" we have. Experiences may vary a lot depending on each combination of variables.

I tested these marginal burns in Samsung SH-S203B, NEC 4550 and Benq 1650. Who knows, maybe the new 7200/7203 (great burner BTW, from the scans I've seen in the NEC forum ) specifically allows the detection of specific issues through the read test..? Go figure...

*ËDIT* BTW, care to post the brand and ATIP of the disc used above?

The Disc above is a silver-pressed Data CD-ROM (not a CD-R) supplied by the game creator Maxis.

For reasons that should be obvious, I cannot supply either a brand or an ATIP.

The other CDs I have noticed similar issues with were mostly CMC MAG, most of which in turn were Memorex.

I still have 4 (I think) maybe marginal discs left over from a large scale stability test of all my media (where I checked every disc I own) about 3 months ago. I am now going back over the most likely suspects to see if I can find them again (when I started the recheck, I had an NEC ND-3550-A and a Samsung SH-W163A) where I had defined any case where one drive found yellow blocks and the other didn't as "marginal." I am now trying to find these again, if I do I will post full ScanDisc, TRT and Quality Scan of any that I find.

Why is that? - (just kidding )Interesting. I don't like Memorex, and I don't trust CMC CDRs much, in general (though they can be very decent depending on the sources). But that's off-topic.

Mmmh. Not sure where to go from here. I'll let you know if I find relevant info.

*EDIT*Just checked you second post. If you feel like discussing media stability and testing for stability in general, better do that in a new thread, this would be rather off-topic here. Drop me a PM if you open a thread with your test results and comments.

We need Dragemester, it looks like he knows several things about the Scandisc test. Why am I not surprised.
http://club.cdfreaks.com/f87/pioneer...ml#post1959155
Excerpt:
Some drives don't report "Damaged" sectors (yellow dots) in a meaningful way, however, when running a ScanDisc Read Test.

Hmmm, that link doesn't point to the right page for me. Perhaps because we're using different number of posts per page?

Here's a link to the post in the thread and a link to the single post.

I don't have any inside knowledge of ScanDisc - only what I have learned from observations. Erik Deppe would be the one to ask about inside knowledge.

It's my observation that all drives can report red (bad) sectors in ScanDisc for sectors that fail to read. A few drives report all other sectors yellow (damaged) when connected externally. Most drives can report sectors as green (good) but not all drives will report yellow (damaged) sectors.

So in other words, some drives will only report green (good) and red (bad) sectors but not yellow (damaged) sectors. In my opinion such drives are not suited for ScanDisc Read Testing.

The behaviour might be different depending on whether a DVD, a data CD or an Audio CD is scanned. (This is my speculation based on the difference in error correction between these media types).

It's not completely clear to me whether the damaged sectors are a result of the software having to re-read, the result of the drive reporting that it had to re-read, or something else such as e.g. C2 info. You'd have to have internal knowledge of the program to know that for sure.

That must be the explanation indeed, as I changed the standard number of posts per page. Thanks. Too bad, thanks anyway.Very interesting! Would you happen to have a couple of examples of drive models doing that?

I found this example in my scan collection, a CD-R scan of an externally connected (by FireWire) Optiarc AD-7170A in a Prolific PL-3507 based enclosure.

C2 Test ScanDisc (Optiarc 7170 FW) @48x.png

All sectors are incorrectly reported as damaged.

Same disc in same drive in same enclosure connected by USB does not have the same problem:

C2 Test ScanDisc (Optiarc 7170 USB) @48x.png

Thanks for sharing, Dragemester!