Hi,

I have decided to post this "problem" on several forums on the web because I think it deserves to be discuss. And I really like to be informed. I have not yet found a thread dedicated. (sorry if such a thread already exists)

I would like to ask the important question of the conditions of explicit Read Transfert Rate tests. I mean the specific features that a drive must have to perform such a test with the best revelance for DVD writing quality. As it is a "stress test", I do know that reading speeds must be as fast as possible but the real question here concerns the reading capability of the drive used .

For those who want to test drives writing quality and burned media compatibility. It seems obvious, IMO, that the drive used for TRT tests must not be a real good reader. On one hand, if the drive has a good error correction and succeed in reading everything , It can't be enough sensitive to show, with a RTR test, the real way a disc has been burned, it will show only the big problems, I mean real failure as POF or very high PI values. But is it enough ? On the other hand, if the drive is too picky and too sensitive, will the RTR test be revelant in term of burned media compatibility ?

So here's the questions :

1) How "bad" must a drive be so that RTR test can be considered as revelant ?
2) What about BenQ drives (BenQ DW1640/1650/1655) for this kind of test ?

Thanks.

Hi

You'll have many conflicting opinions about this. TRTs are a sensitive subject on this board as the majority of members, here, put the emphasis on PIE/PIF scans and tend to disregard TRTs. Ony a handful of members have enough experience and empirical knowledge of transfer rate tests.

Personally, I think you yourself give the good answers to all of your questions, and that more definitive answers to your open questions are impossible to give as everyone relies on conjectures. It would need a specific study.

I do prefer so-so readers to perform TRTs, for the very reasons you mention. But:

1. Even with a so-so reader, a perfect reading curve doesn't guarantee that the disc will show and in-specs PIE/PIF scan in the same drive. I guess you're well aware of this but I wanted to stress the point anyway.

2. In Benq drives, potentially among the best choices to perform TRTs as they're rather picky readers, you can sadly in many instances have false negatives. Depending on the drive/firmware/config, odd irrelevant speed dips can occur. Some users don't experience this, some others do (including I - that's why I now use a NEC 4550 for TRTs).

As a sidenote, I personally don't use TRT to judge of the "burning quality" but rather to check the full readability of the disc at full speed. This tells me that:
1. There is nothing very wrong with the burning quality (but it can still be marginal)
2. The mechanical characteristics of the disc are satisfactory
3. The data is fully retrievable at full speed in at least one drive.

Yes ! Thanks for the reply Francksoy

In fact, there's a real good reason for me to ask that question, I am testing a drive and before releasing any review, I want the "protocol" to be as reliable as possible.
In this way, as I see the different drives (for RTR) used in all the reviews we can read over the net, I wonder in a first time if there is any consensus in drives choice for transfert rates. Sometimes yes, sometimes no, so the question about the readers capability. As we know that it is the comparison that makes you able to formulate conclusions about writing quality, a consensus seems important to me. That's one point.
Yes it depends on the use you have of the quality tests, if it's for a review, you need to provide, with error scans, some actual read tests, just to (maybe) give an idea of actual readability of the disc in other disc players. If you just want to check burning quality for you, you don't necessarily need to perform so many tests, even if it's better to run several types of test. ("I had a dream, that one day..." one drive will tell us the truth And if it's with RTR, it would be great because quicker !)
Fortunately for the moment, I didn't experience such dips in RTR that I could not explain with the PIE/PIF scan. (Except in PIE/PIF scan where my 1650 doesn't show smooth reading curve with some MID, I don't talk about PI values, so weird sometimes that I forget it for PI scans.) So my 1650, because of reading speeds and senisivity is my drive for RTR. (I recently compared it to the LiteOn 165P6S with differents disc if you're interested)
Yes, 100% agree, RTR just comes to confirm a PIE/PIF scan or tells you to perform one and most of all tells you if your datas are readable, finally it is all we want to know as end user. That's another point.

If someone could show me an example (in transfert rate) it would be nice !

I assume you can't especially use it to check DVD±R DL because of the reading speeds (8x for DVD DL I guess)? Am I wrong ?

Personally I think that the terms "false negative" and "false positive" are confusing when we talk about CD/DVD media scanning, because there isn't (yet) a general consensus about what it means.

When you test for something, a positive means that you have found it, and a negative means that you haven't found it. A false positive means that the test says it's there even though it isn't, and a false negative means the test says it isn't there even though it is.

That's why a negative is always what you want when being tested for the presence of a disease when you're being examined by a physician.

But when we test CD/DVD media, are we testing to find a problematic disc or testing to find a good disc?

The meaning of negatives/positives and false negatives/positives depends entirely on our testing philosophy (looking for something bad or looking for something good).

Do you see my point?

Hi DrageMester, pleased to see you participate to this thread
If by "it" you mean "what you're looking for" so yes I see your point !
Yes that's a real interresting point... just like "Do you see the half-empty or half-full glass?" If you're thirsty, you may see the first one. If you're not maybe one or the other.

I said that if I initially posted this problem, it was because I am testing a drive, so in this way I am not "thirsty", I just want to be as objective as possible. So I don't expect anything from the drive tested. I just want to relate how the drive behaves with such disc. What can I consider for criterias in such purpose ?

First I do know that I am not able to reproduce the real conditions to compare my quality scans to ECMA standards but I can't ignore them and have to refer to them. It gives the limits of acceptability (I don't know if it's the good word).

An other criteria is the actual readability, showed by RTR which comes to complete other tests. Considering that most drives have an error correction that supports higher values than those given by ECMA standards, it's problematic because actual readability doesn't mean neither good nor bad quality. It depends on the drive used for the RTR. (I just want to show here my way of thinkng).

I don't understand why such a test tend to be disregarded. Yes it can't be sure at 100% but which test can ?

I think that since we want to use RTR, because of the actuality of the test, among many for a review or as one revelant and single test, the features (sensitivity, speeds) of the drive used for the test have to be the same. This drive have to represente something as so-so (as Franksoy said) It's obvious that such choices have already been made by some users and especially by the «reviewers».

Except BenQ drives, NEC, and liteon LiteON SOHD-167T (with patched firmware, used in CDR-Info reviews), are there other drives that could fulfil RTR needs ? It would be nice to make out a list.

I don't disregard it, in fact I think Read Transfer tests are *more* important than PIE/PIF tests - they're just not as detailed.

Any drive can be used to perform a TRT (TRansfer Test), but I prefer to use at least two different drives that are somewhat picky readers without being really bad readers.

My BenQ DW1655 and NEC ND-4551 tend to react negatively to different disc problems, so I usually use those two, but that still doesn't guarantee that the disc can be read in all other drives.

In fact I don't think it's possible to find a drive that can guarantee readability in all other drives, even if it's a really bad reader, and if you choose a really bad reader as your testing drive, you will be throwing out a lot of good (but not great) burns, so I don't like that testing philosophy myself.

OK Drage, you force me (again) to be more precise...

What I refer as "negative" in the case of transfer rate tests (that's just the way it is in my book): anything that is not a perfect reading curve. Which obviously includes reading failures, but also any speed dips. What I'll call a "positive" is a perfect reading curve, not less. Nowadays there is no reason to accept anything else than a perfect reading curve in any drive. Once again, it's just in my book, but I'm of the opinion that it's legitimate.

So when I meet speed dips that aren't related to actual problems during the reading of the disc, but are just odd reading pauses due to the drive itself, I'll call it "false negative". I've met lots of these with my successive Benq drives (1620, 1640, 1650).

Hope this clears my point.

I won't add further comments on the subject of TRTs as it would only be (more or less) paraphrasing your own input.