MTM2.com

How do you guys verify whether or not a cd / dvd is any good? "DVDinfo pro" comes bundled with Roxio but I'm wondering what else you might use.

I use Nero's CD/DVD test utility. I don't have it now but i'll install it later. I use version 6.6

I copy all the files to my drive and if it doesn't error I asume it works

Ditto, I was doing that as recently as a few days ago when I was trying to back up some work and my POS work Dell only turned out coasters

I do remember reading a magazine review of CD/DVD media where they tested different brands by burning data to them and then running them through a program that counted the errors on the disc (DVDs in particular will always have some errors so they have quite robust error correction - the rationale of the review being that the less you have to rely on the error correction, the better the quality of the media). I can't remember the name of the program that was used in the review (I'll try to dig it out later), but it looks like that DVDInfoPro does all of the same, so I'd say you're already going about it the right way if you use that.

i meant from CD to HDD, if u dont get read errors then it works. preferable use a normal CD drive, A burner has a stronger laser that will error less

I knew what you meant and that was what I meant too: I burnt the disc and then, because I was paranoid, copied the disc contents back to the hard disk. And then, because it errored, I got PO'd at my work Dell.



I found my mag review. They used the freeware tool KProbe, however it's apparently only for Lite-On burners (it might work with others - I think the review used drives other than Lite-On).
http://www.k-probe.com/ (pretty horrible site actually - loaded with ads and embedded YouTube)
http://www.cdrlabs.com/kprobe/ (download link)
http://club.cdfreaks.com/f96/big-kprobe2-thread-93944/ (discussion thread and download link)

Alternative programs I found listed at a couple sites:
- DVDInfoPro: as listed by Phin. You have to pay for a license but apparently old versions are freeware.
- Nero CD-DVD Speed/Nero DiscSpeed: appears to be free and comes with Nero suite software (so guess what, I already have it and should be using it rather than just copying disc contents ).
- PlexTools: software from Plextor for their drives, which are apparently highly regarded for disc quality checking.
- PxScan & PxView: freeware utility written for Plextor drives, a Linux port PxLinux also exists.

Take your pick...

Thanks for the info and links.



> KProbe, apparently only for Lite-On burners

I used a lite-on once; it was the worst burner I ever owned. Shortest life span too.

But unless the program checks the manufacturer's name on the drive, and refuses to work unless it's a specific type, I can't see what difference it would make since we're testing file data.

And a horrible site to boot. Makes me wonder how good a program can be if the site it's on is that bad. I mean, isn't the site indicative of the mindset that wrote the program?



> PlexTools from Plextor

Piece of garbage. This is the part about looking for new tools I despise. Not only do you have to install and uninstall garbage, not only do you have to sift through a configuration system that would challenge einstein, not only do you have to waste time on junk, but the stuff doesn't even work on the company's own drive. And the interface stinks too. This thing was complete, utter, useless junk.



> DVDInfoPro

It comes bundled with roxio so you only have to pay for a license if you want it standalone.



> Nero CD-DVD Speed and/or Nero DiscSpeed 4

The downloads appear to have restricted functionality, tho you can trick them into doing a scan. Which makes me wonder... isn't it possible to read the surface successfully but still fail a crc test? I mean, can't it happily copy corrupted files to a hdd? For instance, these tools distinguish between a read test and C1/C2 tests (whatever they are) so it seems to me there are different 'depths' of reading.

The emphasis on speed is interesting. If my drive is slower than it should be, there's nothing to troubleshoot the problem or recommendations on how to speed it up. And why does data integrity seem to be an after-thought.



> PxScan & PxView

Haven't tried yet. Had enough for one night.




> Take your pick...

can you copy a corrupted file? no not really, if teh checksum is wrong, it will try to read intot he next file, if the filesystem is damage it won't find peices of the file, basically windows will hang trying to copy it. I know Nero data burning lets you do a check after you burn it but, it seems it will indicate at least half the files are different, when in fact they are indentical to their HDD counterparts. So I stopped using it and just try copying down, the only glitches I've ever had were burning audio CD's, occasionally in a song you will hear strange sounds, first time freaked me out cause it was a scraping sound, and I had it in my car blasting, I though I drove over a curb or something.

My Burner is really old though, got it in 2003, still going though, I don't use it for CD like tasks, only burning, I have a CD/DVD player drive in it for such tasks.

Yeah I prefer "let explorer tell me when something is wrong"

nd to explain curator situation, every couple of minutes curator goes into evocation or whatever, best thing to do is to pop trinket and amplify curse and then CoD one minute before this happens as when he goes into evocation the damage he takes will be increased by a rediculous amount for a short while and you will find your CoD HITTING for an insane amount. just like in the netherspite fight while in the blue beam you stick CoD on right at the end of the portal phase when the damage buff will have stacked heaps and your CoD will hit extremely hard.

<center></center>

need I say more? roflol

I have bought six game cds in the past few weeks and every single one of them fails a disk scan.

Hence my reason for asking. Given D2S' survey and my own experience along with your notes, disk checking is anything but a proven process.


Well, that's both true and untrue. It's true in the sense that if the file system is damaged there will be serious problems. It's untrue in the sense that a checksum is not the file system but a part of it... one that can either be used or ignored.

To make a long story short, I was able to create disk images from a cd. I tried nero, roxio, imgburn and ashampoo. All of them successfully created disk images from a cd that dvd info pro reported crc errors on. This raises many concerns. First, what is dvd info pro doing that other scanners are not doing? Were the images created properly and how would I know? Would burning them back to disk (I haven't yet) result in the same errors? If I ask you guys to do a scan of your mtm2 cd, would that prove anything?

For what it's worth, I sent an email to the guy who made dvd info pro and asked about this stuff. I'll let you know if I hear anything back.

After a search to decode Mal's reply I came across the following forum threads that I found to contain interesting info:

http://club.cdfreaks.com/f61/pifs-pies- ... ly-136579/

http://club.cdfreaks.com/f33/pifs-how-d ... en-154485/


(Only just now see your second post Phin, arriving while I was composing this.)

How you got those replies from Mal's post is a bigger leap than I'm capable of.

Anyway... I don't know what a PIF is. I had PIE tonight at dinner time. It was apple.


then it has general issues with " error correction "



The purpose of scanning is to ensure data integrity.



These quotations harken back to the differences in test tools. Each one seems to be testing for different things, and a couple of them seem geared to making nero, for example, appear better than it might be. The only one that actually says it's testing crc is dvd info pro.


I have a disk here. It's scratched and marked. Needless to say, it fails the tests. Should it be trusted just because it might rip to an image? I mean, just because something contains errors doesn't necessarily mean those errors are meaningless. I've gotten bad disks in the past. I've made bad ones that wouldn't read. Sometimes it's the media, sometimes the burner, sometimes who knows. But if we can't use a testing tool as a guide, what do we have left? I am not a hit or miss sort of person.




[ edit ]
[pre] http://en.wikipedia.org/wiki/Parity_Inner_Failure (PIF) http://en.wikipedia.org/wiki/Programmed_input/output (PIO) http://en.wikipedia.org/wiki/Pie (PIE) [/pre]

I wish I coudl take a picture of my worms armageddon CD, in 1999, "someting" got stuck on the eye of my CD drive while the disk was in, this was back before CD drives turned off when not in use, so basically it spun this "somthing" on the disk for about a month, it litiritly cut 2 MM into the disk, about 1.2 CM's inwards fromt he outside of the disk, a perfect circle. Needless to say, the disk could not be read by the player anymore, HOWEVER, it COULD be read by a drive that was also a burner, I was told it was because burners have a stronger laser. I have since made several copies of the disk, straight to other disks, that read just fine in all players.



and sorry, I mean tthe file checksum, hopefully whatever program the file belongs too would check it. Exceptions should only be things like MP3's and other media files that don't require the entire file.

If I have an mp3, I want the complete thing.



Anyway, here's what the guy from dvd info pro had to say.



So now I guess the task is to test a few retail disks and see what percentage of them turn up good or bad. (I wonder if the roxio cd will show errors).

I typed a lot up last night, hoping that the thread wouldn't move too far from what I was trying to understand and address, but of course it has


I don't think that's true. They should all test the same things because there's only one standard for error correction for CDs and for DVDs, though the programs might have some different tools and present the results in different ways.

I can talk about the error correction in DVDs because I've read a bit about this stuff before, and last night I typed out an explanation but now I think getting into that may be going off base. CDs I've never thought about before, but I'm convinced now the method employed is similar. The short of it is, there are two levels of parity information stored on the disc, providing two levels of error correction (two lines of defense, if you will). Testing programs like Nero CD-DVD Speed (which I have just been playing with) and DVD Info Pro can (or should) count the errors that are corrected (or not) by these two levels (as reported by your drive) and show them to you.

Here's some links I found that look to contain some good detailed info, although I confess I haven't read them through:
DVDs: http://club.cdfreaks.com/f76/interpreti ... ans-80545/
CDs: http://club.cdfreaks.com/f77/interpreti ... ans-75573/
Cyclic Redundancy Check: http://en.wikipedia.org/wiki/Cyclic_redundancy_check
Reed-Solomon error correction: http://en.wikipedia.org/wiki/Reed-Solom ... correction

Disclaimer: I assume this isn't an audio CD we're talking about, because apparently the error correction is totally different for them (ie. there is hardly any).



Define "fail". Speaking at least about Nero which is what I've got here, it doesn't give a binary result like the OK/NOT OK on Mal's disc checker.
On a CD I can run a Disc Quality Test, which counts the C1 and C2 errors. C1 means a bad error occurred which was corrected simply. C2 means there was a bad error, but C2 are still correctable within a certain limit.
And I can run a Disc Scan, which lets me run two tests. A File Test checks all the files and reports any that are unreadable. A Surface Scan draws a chart representing all the data blocks on the disc, coloured to show which sectors are readable, which are damaged (readable with error correction) and which are unreadable.

None of these declare a disc passed or failed. It just presents the result. I might say a disk has failed when some parts on it are unreadable, and I won't be able to copy at least one file from it, but that's my interpretation.



I just copied an image from a known bad disc using Nero. (The disc was a poor quality burn, Explorer can't copy all the files off and CD-DVD Speed indicated unreadable parts.) The rip went right to the end, but it signalled unrecoverable read errors through the latter part of the disc, which I know is where the disc scans have identified unreadable sectors. I assume the rip probably inserts dummy sectors anywhere that the drive couldn't read data.

If I read you right (as of last night), the crux of the matter here is that you have ripped a disc and want to verify the integrity of the disc and the rip?
If something like Nero or DVD Info Pro indicates parts of the disc are unreadable, then I'd say the rip wouldn't have been able to get those parts - the ripping program should have indicated there were errors to corroborate this. I don't know how you'd test the integrity of a rip. Once it's on the hard drive, the dummy sectors must have as much integrity as the rest of the data that could be read. You'd only know by looking whether all the data is actually there.



I have a burned audio CD that was damaged on its printed side, it appeared to have a hole ripped right through the reflective layer, so you could see through. Somehow it still played...



I can virtually guarantee you that ANY disc, whether stamped or burned, WILL show errors. That won't mean every CD is bad. The distinction will be to look at the levels of how many are "friendly" errors (C1 or PIE, which are easily corrected), "nasty" errors (C2 or PIF, which may still be readable), or unreadable.

They "should" all test the same things.


> http://en.wikipedia.org/wiki/Cyclic_redundancy_check

Read through that the other day. There's a joke in those pages - don't know if it's deliberate or accidental - but I won't side track.


> Disclaimer: I assume this isn't an audio CD

We're talking all instances, but of particular interest today are mixed mode disks.


> Define "fail"

CRC error.



I understand but that sort of terminology and generalization drives me around the bend. Does it, or does it not, verify file integrity via crc? If one read fails, how is it correcting - an alternate file system? Which file system is it reading first? How prioritize the reads?



> A File Test checks all the files and reports any that are unreadable.

How would it know without a surface scan? The best it can do is see if the file is there.


> A Surface Scan draws a chart representing all the data blocks on the disc, coloured to show which sectors are readable, which are damaged (readable with error correction) and which are unreadable.

How? At what level is it reading? Does it rely on the operating system to read what it can and then report back whatever return values it provides or does it actually go through the bytes, or the bits? Is it running a crc comparison between what it reads and what is stored on the disk? This stuff is too vague and doesn't tell us much at all.


> None of these declare a disc passed or failed. It just presents the result.

Failed is my word. I've done well over a thousand, maybe a couple thousand disks. I check them all. If any errors are reported, ANY, the disk is thrown out. That is a failure. Once a disk has errors, there's no predicting how it will perform. So, it's no good to me, and I won't pass it along to anybody else.


> I just copied an image from a known bad disc using Nero.

I did an image from the main disk I'm concerned with. I used nero, roxio, imgburn and ashampoo. (I didn't realize but dvd info pro will rip an iso). All of them created the image and none reported errors. Yet, the crc scan reports plenty of them.


> If I read you right (as of last night), the crux of the matter here is

The crux of the matter is that I have bought several disks, most used. I want to know if they are good bad or otherwise.



I imagine it could be done a number of ways. Total size, block by clock, read the file system and compare file by file, etc, for both the disk and the image.



If that's a money back guarantee you're going to lose a lotta cash. As I noted above, I've done thousands, and none (that I keep) have errors.

For today, this was the test program.
<center>
<img src="http://mtm2.com/~forum/images/cdtextabout.gif" width="365" height="217">
</center>

I don't have a lot of my software disks available at the moment - one of the reasons I bought a few - so these are the only commercial disks I could test:

• <a href="http://mtm2.com/~forum/images/cdtestroxio7.gif">Roxio 7</a> - ok
  
• <a href="http://mtm2.com/~forum/images/cdteststudio6.gif">Visual Studio</a> - ok
  
• <a href="http://mtm2.com/~forum/images/cdtestxp.gif">Win XP</a> - ok
  
• <a href="http://mtm2.com/~forum/images/cdtextwin98.gif">Win 98</a> - readable errors
  
• <a href="http://mtm2.com/~forum/images/cdtestbasic6.gif">Visual Basic</a> - unreadable errors

3 of 5 were perfect. 1 with readable errors (something I've rarely seen). And one failed crc, which, by the way, won't install from the cd (I've ripped an image and install from that, or copy the whole thing to hdd and install from there - oddly, making an image or copyng to disk did not complain it was unreadable).

----------

My understanding of crc or any integrity check can be illustrated with pod files. In evo and later games, TRI included a crc as part of the pod file format. When writing a pod, a crc value is calculated from each file, then saved. The game, I assume, checks the files against this value when running it in the game. This serves two purposes. One, that the pod has not been corrupted anywhere along the line. Two, to detect cheating when playing online. It's the before and after comparison that gives the process its worth.

Yes, apparently audio disks don't have the crc included, so only read and size tests can be done.

In the case of cds/dvds, being unfamiliar with the format and file system, I can only assume the crc value is stored as part of the file information. When a check is done, the file is read, the crc is calculated, then compared to that stored on the disk. The two values should match. If not, the disk is junk.

Note. It is theoretically possible that a crc value was calculated incorrectly at the time the cd was made, or that it's not done properly at the time of testing, in which case a failed check would not be indicative of file integrity but of the creation and test process. I mean, what if the crc value didn't store right rather than the disk or file?

Still, after all this time and thought, I have to conclude that any disk that contains errors needs to be returned as junk.



PS. <a href="http://mtm2.com/~forum/images/cdtextebay698.gif">A disk that gave rise to this topic</a>. I stopped the scan after five thousand errors.

See this is what I love about you Phinny - you start off with what seems like a fairly innocuous question but then keep digging and digging with the result that it pushes me into learning new things that I didn't know about before


Mate, you and I are operating under different definitions of "error". I've adopted the ideas of C1, C2, PIE, PIF etc errors expounded upon in those cdfreaks threads, which it seems are generally used as a measure of disc quality (they're generally talked about for comparing media and burn quality). Which is maybe completely the wrong metric for me to be arguing with you. Your focus seems to be about checking whether the data on any part of the disc is irretrievable.

When I say any discs give will give you errors, I mean the "friendly" C1 or PIE errors that indicate the drive recovered from a glitch in reading the media. For the sake of example, here's my (reportedly 100% ok) WinXP CD:



It shows C1 errors, but using the disc scan nevertheless marks the whole disc as readable:



Compare with my failed burn CD: quality scan

Reality check: commercial movie DVD

That's all well and good but...


I have to confess I simply do not know. I've learned a little but not everything and not enough to be able to draw the link. I'll have to spend some time reading up, or find an expert to query, to get the correct picture of how all this is meant to go together...

> Your focus seems to be about checking whether the data on any part of the disc is irretrievable.

I want to know if the file coming off the disk is the same as that created by the maker of the file.

In a simple example, if an ACT file is put on disk BGR rather than RGB then there is going to be a markedly different result when it comes time to use that file - even tho the disk is readable and file properties appear correct. Unfortunately, there are a limited number of ways to tell. "If" the difference is extreme enough, we can tell by looking at it. The problem with that is the difference might not be substantial enough to differentiate with your eye. And you'd have to check all the files manually. A better, more reliable way, is to implement some sort of checksum system.

A more complex example would be an executable or dll. There is no way to visually inspect the file (manually or otherwise), and any errors might reside in lesser used parts of the program so that anomolies might never show up for some users. On the other hand, the difference between a 1 and a 0 could effect program execution, or you might get unpredictable errors that you spends months trying to troubleshoot and never find a satisfactory solution... because you're looking for bug, software, driver, hardware etc solutions, never even considering the file might never have copied to and subsequently from the disk correctly to begin with.

In the case of my visual basic disk above, the only disk I have with a known, predictable problem, I don't know why it won't install from the disk but will install from a mounted or burned iso. You'd think something is different but I don't know what or why. But it's enough of a problem to me that I flag it in my mind as a problem with the source.

And the question, then, is should I accept a disk with hundreds or even thousands of crc errors just because it 'might' work from an iso?

It seems to me that the standard for burners and for media is insanely low, and the tolerance for errors is almost idiotic.



In related news, you might find this of interest

http://www.isobuster.com/help.php?help=160 File Systems

It was partly due to isobuster that this whole thing began. Apparently a couple of the iso files I created have issues. They'll mount but not burn, or burn but not unzip. Also interesting is that the test rips I did above, some but not all the images are the same size. The program specific formats you'd expect to be different but the iso/bin files should be spot on, but are not. And the cue sheets generated by each program have differences. Makes me wonder what each one is reading.




> sundry

I don't know cricket, sorry.




PS. I can probably test a couple commercial cds or dvds if it would prove anything.