Official CD Doctor (WSES alternative) Thread

[dodecahedron]

This has already been done not only for Nero, but for some other CD burning utilities as well; see, e.g.,

http://club.cdfreaks.com/showthread.php ... adid=60752

why go far afield ?
http://www.cdrlabs.com/phpBB/viewtopic. ... 5632#45632
http://www.cdrlabs.com/phpBB/viewtopic. ... 5639#45639

[Inertia]

I'd love to run CD Doctor on my XP Pro box, but I'd like to avoid messing up my clean OS with Adaptec's ASPI. (Yeah, I know 4.60 is apparently harmless, but I'd like this Adaptec dinosaur to die eventually.) Did anybody consider any fixes, like putting Nero's WNASPI32.DLL in CD Doctor's directory?

The RPCDE revision cited as a fix for the socalled "ASPI bug" was caused by a bug in the original program code.

There is nothing risky or scary about the v4.60 ASPI for WinNT/2K/XP as installed by ForceASPI. It consists of four files:

ASPI32.SYS
WINASPI.DLL
WNASPI32.DLL
WOWPOST.EXE

and a WINNT.REG registry file which installs the following registry entries:

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\ASPI32]
"ErrorControl"=dword:00000001
"Start"=dword:00000001
"Type"=dword:00000001

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\ASPI32\Parameters]
"ExcludeMiniports"=""

That's all there is to it. It doesn't touch anything else. It won't mess up the operating system. It can either be automatically removed with the KillASPI.bat utility in ForceASPI or it is very quick and simple to remove it manually if desired.

[cfitz]

I'm with Inertia on this one. Although I think that what KCK was asking was if CD Doctor could be made to run without ASPI just as Rpcde was made to run without ASPI - he wasn't talking about fixing any bug in ASPI.

By the way, I did try running CD Doctor with Nero's wnaspi32.dll in the CD Doctor directory. CD Doctor started and tried to run, but wasn't able to do the actual measurement. If anybody does have luck with alternative ASPI layers, please let us know.

cfitz

[KCK]

Sorry for not being clear enough in my previous message. As for Adaptec's ASPI, let's say I'm (over)cautious after seeing so many complaints on various forums, and the current trend seems to be to move away from having to install Adaptec's ASPI.

dodecahedron:

I just linked to a thread I had seen by chance several minutes before coming here. Since I can't follow each thread here, I should have searched (maybe it works now). I only wanted to quote an example; I'm not interested in RPCDE, but only in it being able to run without ASPI, as discussed by cfitz:

http://www.cdrlabs.com/phpBB/viewtopic. ... 5632#45632

Now I see that I should have omitted any reference to RPCDE.

Inertia:

Thanks a lot for the useful info. Now, just like a typical user, I could install ASPI, followed by tons of other software that really confuses XP and Nero, and come back to you complaining "ASPI kills InCD!"

Seriously though, your information is reassuring, so maybe I'll try CD Doctor after my holidays.

cfitz:

Did you try W2K, XP or both? I'm not sure it should matter, but who knows (MS? Don't think so! ).

[dodecahedron]

dodecahedron:

I just linked to a thread I had seen by chance several minutes before coming here. Since I can't follow each thread here, I should have searched (maybe it works now). I only wanted to quote an example; I'm not interested in RPCDE, but only in it being able to run without ASPI, as discussed by cfitz:

hey, that's cool, i was only joking!

[cfitz]

Did you try W2K, XP or both? I'm not sure it should matter, but who knows (MS? Don't think so! ).

W2K, but I suspect the results would be the same for both.

cfitz

[Kennyshin]

@cfitz

I can see you here, too.

[cfitz]

@cfitz

I can see you here, too.

But just as a guest! My true loyalties are with CDRLabs.

cfitz

[mdenny]

I am new to this forum, and just tried to run CD Dr on my laptop using XP Pro.

I get the message "missing manufacturer.txt file or ASPI layer".

I guess, like others, I would like to not clutter up XP with ASPI...who knows, maybe Nero will object...maybe now, maybe later. Is there an easy fix to this?

[cfitz]

I know you don't want to hear this, but the easy fix is to install Adaptec ASPI 4.6 with forceaspi. There is no alternative to installing an ASPI layer because CD Doctor requires it. Nero won't object to the Adaptec ASPI layer, since it uses its own ASPI layer that has nothing to do with any system level ASPI layer. Ahead uses its own ASPI layer specifically to avoid any problems with system level ASPI layers. I run both Nero Burning Rom and Adaptec ASPI 4.6 with no problems.

Frankly I think Adaptec's ASPI layer is the most tested and stable one out there. I really don't know how Adaptec's ASPI managed to get such a bad reputation with so many people. Probably because so many software producers were distributing and installing it willy-nilly with their products, without any sort of version control considerations. And since any developer could obtain and use the ASPI SDK, no doubt a number of developers did their jobs badly and put out bad code that uses ASPI. Then the bad experiences a few people had with those bad applications snowballed, Internet rumor style, into "ASPI is evil". Perhaps Adaptec should have kept a tighter leash on the third-party developers.

Installing ASPI 4.6 with forceaspi isn't going to do all sorts of awful, mysterious things to your computer. And it can easily be removed in the off chance that you do have some problems. Check Inertia's post above - he said it perfectly. And, finally, remember that ASPI isn't going to take over your system, inserting itself into all your programs and breaking them. It is simply another API, and it will only be used by programs that specifically call it.

I'm not telling you to install ASPI or not. That is your choice. But the fear of ASPI that so many people seem to exhibit is just not justified. Just my two cents...

cfitz

[spath]

Does anyone know what this tool is
really measuring : correctable or
uncorrectable errors, counted in
bytes or frames ?

[kgatwork]

cfitz:

Thanks for the info, finally installed the aspi 4.6 layer and the program now runs on my machine.

[cdrfreak]

If CD Doctor had a sliding scale graph that would adjust to the error rate, similar to CDSpeed's, it would great. As it is, it's still a terrific tool for C1 measuring in Windoz, and the price is right.

There will be a C1/C2 error test in CD Speed.
The guys from CDRInfo posted some info and a few screenshots of this new test in their writing quality article (great stuff btw.)
This article is mentioned in the EAC forum:
http://www.digital-inn.de/showthread.ph ... adid=17426

[cdrfreak]

CDSpeed sees lots of C2, perhaps it's CDSpeed that has the problem.

CD Doctor doesn't like it much either, looks like a couple C2 indicated at the start. The gaps interest me, also the doctor has several times given me a graph that is blank on one half. I presume that the scale in CD Doctor is such that 1000 would be the same distance up from 100 as 100 is from 10.

This is a problem with the way the C1/C2 errors are measured.
If the drive can't keep up then it will miss some data samples. This is indicated by the gaps.
This anomaly is also present with some drives (LiteOn LTR-32123S for example) when running the CD Doctor test at full speed instead of the default 8x.
If you want to compare the C2 graphs with CD Speed you should set the read speed to 8x before running the CD Quality Test.

[cfitz]

There will be a C1/C2 error test in CD Speed.

I didn't see anything about C1 testing in CD Speed in the links you provided. Unless I misunderstood the quoted material, they are just talking about the advanced DAE quality error test in CD Speed v1.02. This is an interesting new test (although it requires a specialized test disc), but doesn't test C1 errors. On the contrary, it is specifically designed to test the accuracy of C2 error reporting.

cfitz

[rdgrimes]

The folks at CDRInfo have apparently come up with a bete version of CDSpeed that measures C1. I've seen some screen shots, it's much the same only with an added color on the display for C1. I saw no mention of where this was obtained or if it will be released any time soon. I'm not sure if the article has been published online yet, but I have a .pdf of it somewhere if you want it.

[cfitz]

Thanks rdgrimes and cdrfreak, I guess I did miss that. There was one broken link buried in the link cdrfreak posted. Perhaps that had the information about C1 testing in an unreleased version of CD Speed. Anyway, no need to get me that pdf. I'll take the combined word of you two, and just wait for the official release.

cfitz

[cfitz]

Does anyone know what this tool is
really measuring : correctable or
uncorrectable errors, counted in
bytes or frames ?

At the very least it is measuring correctable errors. The C1 errors that it measures are certainly correctable. As for the C2 errors, as you personally know, spath, there is more uncertainty and disagreement in general as to what constitutes a "C2 error" and whether or not "C2 errors" include uncorrectable errors as well.

My educated guess would be that as far as CD Doctor and WSES are concerned, that which is labeled C1 consists of E11 and E21 errors, while that which is labeled C2 consists of E12 and E22 errors. Whether the actual number reported is E11 + E21 or E11 + 2*E21 or some other formulation, I don't know. I do believe that the number is reported as "events" per second, since the graph is labeled "cps", which I take to mean either "characters (bytes) per second" or "corrections per second".

If I get a chance, I will write Miyuri and see if he knows. There is a chance that even he does not know, since he might just be reporting information delivered by the chipsets of the supported drives, and the documentation for those chipsets may not be as detailed as we would like.

For those with less technical interest, think of the graphs that CD Doctor and WSES produce in the following manner: The C1 portion of the graph shows the number of smaller, less serious random errors that occur while testing the CD, while the C2 portion of the graph shows the number of larger, more serious errors. The C1 errors can be corrected with the first stage of CD error correction, while the C2 errors have to be corrected at the second stage. Don't worry about the exact units and definitions for the graphs. Instead keep in mind Miyuri's comments from the readme.txt file and my comments from another thread in this forum ( http://www.cdrlabs.com/phpBB/viewtopic. ... 6103#46103 ) :

Because CD Doctor is not dedicated measurement equipment, please use it as a gauge of performance rather than an absolute measurement.

<When testing CD-R media> the results we see are a function of the writer, reader, media and a host of other variables. So, we can't specify the absolute quality of media X, and, in fact, it doesn't even make sense to try. At best one can say media X performs at some level under some conditions.

However, this doesn't invalidate the quest to characterize media and media/drive combinations. Certainly relative assessments of media quality are fair game. We may find that media X does consistently better than media Y when tested under identical conditions on the same drive. We can then say that media X is superior for those conditions on that drive. And we may find that media X does consistently better than media Y under a number of different sets of identical testing conditions. Then, we can expand our conclusion to state that media X is generally superior to media Y (keeping mind that there may always be exceptions for specific test conditions).

Free software tools (such as CD Doctor, WSES and CD Speed) running on consumer-level CD-RW drives are not going to be able to characterize CD-R quality to the same levels of precision and exactitude that dedicated CD testing equipment costing tens or even hundreds of thousands of dollars can. Keep this in mind, and use these tools to gauge relative performance, and you will get good service from them.

cfitz

[spath]

> As for the C2 errors, as you personally know, spath, there is
> more uncertainty and disagreement in general as to what constitutes
> a "C2 error" and whether or not "C2 errors" include uncorrectable
> errors as well.

Well, I have no doubt about the official meaning of "C2 error" (any
error, correctable or not, detected by the CIRC's C2 stage), I only
have doubts about what CDDoctor means by C2 error

> If I get a chance, I will write Miyuri and see if he knows.
> There is a chance that even he does not know, since he might
> just be reporting information delivered by the chipsets of
> the supported drives, and the documentation for those chipsets
> may not be as detailed as we would like.

If you can ask Miyuri please do so, that would be great. By the way
you could also suggest him to add the uncorrectable C2 errors to his
plots, because it's a very important figure which is currently missing.
Another practical method to get some infos about the plots could be
to use an artificially degraded disc (say some tape on a cd) and test
it with CDDoctor to find out the maximum C1 and C2 values it can report.
Since correctable errors are measured in bytes and uncorrectable errors
are measured in frames, the maximum values will be different between
the two.

[cfitz]

Note to the general reader: The following, in my opinion, is discussing minutiae that don't have any relevance to the average user of CD Doctor who is using it as the author intended - for relative comparisons of burned CD-R quality. So, either don't bother reading it or read it with that in mind, and don't get worked up about it.

Well, I have no doubt about the official meaning of "C2 error" (any error, correctable or not, detected by the CIRC's C2 stage), I only have doubts about what CDDoctor means by C2 error

That is just my point. Although you have a firm and reasonable definition in your head, apparently not everyone shares that definition, or at least there is reasonable doubt as to whether or not everyone shares that definition. Unless there is a generally accepted common usage for the term or an official definition spelled out in a standards document, I think "uncertainty and disagreement" accurately characterizes the current state of affairs.

Evidently there isn't a generally accepted common usage, or we wouldn't even be having this discussion. As for a definition spelled out in a standard, I can't find a definition of the term in ECMA-130, and I'm not about to shell out $120 for the Sony/Phillips document. If you have a reference to an official definition in a standards document then for heaven's sake stop being coy and share it with us!

In the meantime, your definition spells out two fundamentally different types of C2 errors (in your nomenclature) that are measured in two different sets of units, so I would contend that even your usage of the term ends up being ambiguous and, hence, uncertain. Your request to have both “uncorrectable C2 errors” and “correctable C2 errors” separately plotted implicitly acknowledges as much. By the way, if it can be done I do agree that it would be a nice addition.

On a related note, what do you call an E31 error that can be detected, but not corrected, at the C1 level, and can be corrected at the C2 level? Is that a C1 error, a C2 error, or both a C1 and C2 error? I could see reasonable arguments for any of the three characterizations.

I will try to write up an email and ask Miyuri if he can shed any light as well as pass along your request, but I won't make any promises as to when I will get to it. Technical discussions are harder for me to translate.

cfitz

[spath]

> Evidently there isn't a generally accepted common usage, or we
> wouldn't even be having this discussion. As for a definition
> spelled out in a standard, I can't find a definition of the
> term in ECMA-130, and I'm not about to shell out $120 for the
> Sony/Phillips document. If you have a reference to an official
> definition in a standards document then for heaven's sake stop
> being coy and share it with us!

Well, the standard definition used in the industry is that a Cx
error is an error detected at CIRC's Cx stage, nothing more and
nothing less. Then these errors can be split into two categories,
correctable and uncorrectable. No idea if this is really written
in any official document, as the "C1 error" expression looks quite
self-explanatory to me.

> In the meantime, your definition spells out two fundamentally
> different types of C2 errors (in your nomenclature) that are
> measured in two different sets of units, so I would contend
> that even your usage of the term ends up being ambiguous
> and, hence, uncertain.

The reason is that when you get uncorrectable errors you cannot
know how many bytes are wrong, so you obviously cannot count
uncorrectable errors in bytes and must use another unit.

> By the way, if it can be done I do agree that it would be a
> nice addition.

It's very easy, so if Miyuri agrees to add this feature I can
explain him how to do it.

> On a related note, what do you call an E31 error that can be detected,
> but not corrected, at the C1 level, and can be corrected at the C2
> level? Is that a C1 error, a C2 error, or both a C1 and C2 error?

It's both an uncorrectable C1 error and a correctable C2 error.

[cfitz]

Well, the standard definition used in the industry is that a Cx
error is an error detected at CIRC's Cx stage, nothing more and
nothing less. Then these errors can be split into two categories,
correctable and uncorrectable. No idea if this is really written
in any official document, as the "C1 error" expression looks quite
self-explanatory to me.

Well, that certainly sounds like a reasonable definition. But to me the definition "a Cx error is an error correctable at CIRC's Cx stage" also sounds reasonable. (Okay, I will admit that is the definition I have been using in my own mind in absence of an official definition.) This alternate definition is not only reasonable but also less ambiguous than the one you describe. There is no need to further subdivide the categories into correctable and uncorrectable and quote results in two parts with different measurement units, and it doesn't end up counting the same error twice as in your explanation of E31 error categorization. Of course, to be complete one would also have to tally the number of uncorrectable errors.

That is just my two cents worth, and it doesn't count for much if the generally accepted definition is as you present it here. But I still don't know how generally accepted that definition really is. I wonder: if the definition you presented truly is generally accepted, then why did you even think to question the units reported by CD Doctor in the first place? And you yourself have said elsewhere that there are different definitions:

> Also, in CD Doctor, in the generated graph, what
> does the values in the legend of C1 and C2 represent?

It can indeed be several things, so I asked the same question
in the CDDoctor thread few days ago, but nobody seems to know
the answer.

(from http://www.cdrlabs.com/phpBB/viewtopic.php?p=46639 )

The reason is that when you get uncorrectable errors you cannot
know how many bytes are wrong, so you obviously cannot count
uncorrectable errors in bytes and must use another unit.

I do understand this. I was just pointing out that the definition you have presented requires the splitting of errors into categories that must be reported with different units and, thus, that under your definition usage of the term “Cx error” by itself, without splitting it into sub-categories, is ambiguous.

Anyway, this is all secondary to your actual question: what exactly is CD Doctor plotting on its C1/C2 graphs? I will write to Miyuri and see if he can clarify. I am still guessing that he is using my definition. In other words, the plots do not include uncorrectable errors. But, that is just my guess.

By the way, do you have information to this level of specificity regarding WSES? As far as I can tell, CD Doctor and WSES are measuring and reporting the exact same thing.

For those following along and wondering about the implications of spath's question and whether its answer has much significance in interpreting the CD Doctor (or WSES) test results you have obtained in your own testing, I would say the significance is minor and counsel you not to worry about it. The major point of spath’s question is whether or not uncorrectable errors are included in the C2 graph. Ideally you don't want to see any C2 errors on your freshly burned CD-R if it is of good quality, so when testing to decide what media work best on your system, you ought to be ruling out media that show C2 errors of any type. And by the time errors that are uncorrectable at the C2 level show up, your test results chart will already look so ugly that it will be clear that the disc is no good:



( see http://www.cdrlabs.com/phpBB/viewtopic. ... 5698#45698 )

cfitz

[spath]

> I wonder: if the definition you presented truly is
> generally accepted, then why did you even think to question
> the units reported by CD Doctor in the first place?

The only limitation is that uncorrectable errors cannot
be counted in bytes, so for instance correctable errors
can be counted in bytes, frames or whatever unit the
chipset's designers chose to use.

> By the way, do you have information to this level of specificity
> regarding WSES?

No idea, but since they support the same drives they must use
the same chipset's feature, so it's no surprise that the plots
are similar. A datasheet of the chipset used in these drives
would explain everything, but unfortunately Mediatek is very
secret about their work (if Miyuri has such docs, I'm also
interested )

[cfitz]

No idea, but since they support the same drives they must use
the same chipset's feature, so it's no surprise that the plots
are similar. A datasheet of the chipset used in these drives
would explain everything, but unfortunately Mediatek is very
secret about their work (if Miyuri has such docs, I'm also
interested )

My thoughts exactly. But I guess that error handling (particularly at the C2 level) is one of the places chipset manufacturers can differentiate their products, so they aren't too willing to disclose all the gory details to the likes of us.

I'll let you and everyone know what, if anything, I find out.

cfitz

[Kennyshin]

Media FAQ by rdgrimes at CDFreak