Page 1 of 1

Comparison of PlexTool Pro results with Audiodev

PostPosted: Sun Jun 29, 2003 8:01 am
by MediumRare
A couple of months ago there was some discussion on test quality in this thread: http://www.cdrlabs.com/phpBB/viewtopic.php?t=10416. I mailed the (German) magazine that triggered the somewhat heated debate and they said they were planning an article on consumer testing tools.

It materialized in the latest issue of c't, No. 14, available on Mon. in Germany and neighbouring countries. It is primarily concerned with the PlexTools Pro but mentions CDSpeed, WSES and K's Probe as well. Unforunately, the article is not available online.

The only really quantitative comparisons are carried out for the PlexTools Pro and the PlexWriter Premium, where they actually compare results with the Audiodev CATS-SA3 for one 80 min. disk. Here a summary (in my words):
  • C1/C2/CU: Plex C1 values scanned at 4x are flat at ca. 2 and very similar to the Audidev BLER values up to ca. 60 min. After that, Audiodev results (scanned @1x !!) increase somewhat to ca. 40. The Plex scan at 10-24x also increases from 60 min. but goes >100 and also has some C2's there and in a lesser peak (30) at the start. Plextor "seems to be generally more tolerant of reading disturbances". Results apparently didn't vary much when they tried another Plextor drive.
  • Jitter: Results vary strongly from drive to drive. They deduce that the lines on the unitless Plex graph are ca. 5 ns apart. 3T pits are generally too short with firmware 1.01 (-> increased jitter).
  • Beta: The Plex values bear little relation to what Audiodev measures. Ignore these results (no Audiodev results shown).
  • FE/TE: test is not very useful because it does not tell what the proper burning speed is- you have to resort to trial and error anyway. 52x is allowed (initially) for all media and the drive then relies on Powerec. This often reacts too slowly and (e.g.) 48x Verbatim are spoiled @52x when the linear speed exceeds 48x. 40x is OK.
  • Varirec: fingers off! It'll only get worse.

They did not mention which media were tested or how they were burned.

Regarding CD-Speed and KProbe, the main statement is that the results depend on the drive used ("the fundamental problem in measuring error rates of CDRs"). The results shown apear to be for different disks than the Plex (other values for end-LBA, qualitative differences.) They also used an antedeluvian drive (LTR-24102B) with KProbe.

On the whole, PlexWriter Premium users have learned something, the rest of us are not much better off. They missed the chance to compare the diagnostic properties of the various tools quantitatively by testing the same disc with each tool.

G

PostPosted: Sun Jun 29, 2003 10:08 am
by cfitz
Thanks for the report, MediumRare. I agree it is too bad they didn't run the same disc through more drives and test programs.

cfitz

PostPosted: Tue Jul 01, 2003 8:13 am
by MediumRare
Apparently the scans were for the same disk (feedback from author). But because KProbe ran at max. speed, the C1 rate was much higher- so a quantitative comparison of the same disc at similar read speeds still isn't possible.

I've been thinking about this a bit more. We now know that the PlexWriter read 1 disk with results similar to the Audiodev (and "better" at t>60 min.) One disk is certainly not a good basis for empirical results (they apparently did test more), but it does show that a consumer product can give reasonable results.

So now a very interesting question is:
how do the 8x or 4x scan results from PlexWriter Premium and (for example) LiteOn LTR52246S compare for one and the same disk?
In the absence of a direct comparison of Audiodev results with a modern LiteOn drive, this is the best we can do at present to evaluate a LiteOn's scanning capability.

To the best of my knowledge, this direct comparison has not been made. Can someone (Ian?) do this or point me at such results? It'd be best to scan several disks: at least a good one and a problematic one.

G
(Post no. 300 :D )

PostPosted: Thu Jul 10, 2003 5:38 pm
by MediumRare
OK- I've scanned the graphs of the scans from the magazine. The originals were rather small and the paper thin (some ghost images). Scanning from printed media is always a bit touchy, but thanks to cfitz's processing suggestions, the results are at least recognizable.

Scan with Audiodev CATS-SA3 @1x
Image

Scan with Plextools Pro @4x CLV
Image

Scan with Plextools Pro @10-24x CAV
Image

Nero CD-Speed quality test
Image

Scan with KProbe 1.1.12 @max, LTR 24102B
Image

There are more plots in the magazine, esp. regarding jitter- but my primary interest is BLER/C1/C2. I think there is a complete translation of the article in the works at CDRInfo.

Please read my prior posts as to why I was so excited about these results, particularly the Plex scans. All scans are for the same disk, but with varying scan speeds. KProbe seems to have slowed down twice during the scan.

Thanks to Aviationwiz, we also have a comparison between Plextools and KProbe (1.1.14) for a LTR-48125W, both @8x. Halcyon (at CDFreaks) had greater differences in the absolute C1-values between his PlexWriter and LTR52246S, but the distributions were very similar. He used scan speeds >=24x and not 4x or 8x, unfortunately. rdgrimes has shown in his tests of the new LTR-52327S at CDFreaks that the scans are very similar to a 52246S at low speeds, but differences become more pronounced at higher scan rates.

What can we learn from this?
We know that KProbe, CD-Speed, PlexTools etc. are useful for comparing the quality of burns on one burner. This is the most important application for these tools and really all that most of us need to know.
What these comparisons tell us, though, is that what these consumer drives measure (with all uncertainties as to sampling size and what the drives report as C1, C2, CU) does seem to approximate the results of professional tools at low scan rates. Of course these values aren't absolute values and will vary from drive to drive. Especially additional comparison of Plexwriter Pro with LiteOn / KProbe scans would be useful (at 4x or 8x !!! - Halc(yon)- can you do this?)

G

PostPosted: Fri Jul 11, 2003 3:24 am
by Halc
Thanks for the scans MediumRare!

I have already provided scans at 4x for both LiteOn 48246s (with 52246S firmware) and Premium in the following thread:

http://www.cdrinfo.com/forum/topic.asp?TOPIC_ID=10628

Look for scan at 4x. I can provide more later on, but I'm now in the midst of doing 1.1.16 testing and trying to find out a few issues with PlexToolsPro.

When I've understood these two issues a little bit better (and perhaps when Kprobe 1.1.17 is out) I'll do more slow speed testing.

cheers,
Halcyon

PostPosted: Fri Jul 11, 2003 2:23 pm
by MediumRare
Halc wrote:Thanks for the scans MediumRare!

I have already provided scans at 4x for both LiteOn 48246s (with 52246S firmware) and Premium in the following thread:

http://www.cdrinfo.com/forum/topic.asp?TOPIC_ID=10628

Look for scan at 4x. I can provide more later on, but I'm now in the midst of doing 1.1.16 testing and trying to find out a few issues with PlexToolsPro.

When I've understood these two issues a little bit better (and perhaps when Kprobe 1.1.17 is out) I'll do more slow speed testing.

cheers,
Halcyon

I saw the post but had forgotten that there was a 4x scan. My initial thoughts at seeing the first graphs (scanned at 24x) were that your Plextor is possibly damaged. On the other hand, your later comparisons at CDFreaks don't show such extreme qualitative differences.

I hope you resolve your issues with the testing programs.

A CD-Speed quality scan would be useful additional information for those crippled discs.

G