Help understanding CD/DVD Speed's "Score"

[_H_itman]

I've been in the BenQ forum lately because I'm getting a 1620, and I have come across a few scans that puzzle me. I'm probably overlooking something, but I don't understand why the 2nd pic below (Burned by BenQ) scores better than the first scan (NEC 3500)? The only thing I can think of would be because the NEC scan doesn't show Jitter. Or is it because the BenQ scan was performed at 8X?
[b]
16X Burn on TY02 with NEC3500 Liggy's 218 firmware
Scanned at 4X with Liteon 832S (QUALITY SCORE=70)

Why does the below scan score better than the NEC scan above?
SONY08D1 burned at 12x on BenQ DW1620
Scanned at 8X with BenQ 1620 (QUALITY SCORE=96)

BTW, I found the 1620 scan HERE, posted by zevia.
[/b]

[Jucius_Maximus]

This is speculation, but perhaps it's dependent on the scan speed as well? The second scan definitely shows more errors, but then again these were found when the disc was spinning at a higher speed.

[Jamos]

Simple, your nec burn has high PIF spikes above the 4 limit (liteons scan pifs at 1ecc). Typical 16x NEC burn, it may not matter on some dvd players for playback (if its a movie), but on Sonys you may see some skipping and pausing on playback. The benq scans at 8ecc so the pif limit is 16 not 4. (note the different colors you want to stay in the green on the pif score).

so in my opinion the score is accurate..though you have higher pie (which really doesnt factor until you get in the upper 100s) your pifs are much lower on the Benq burn, hence the better score.

Also Max is right the second disk should of been scanned at 4x not 8x.

Though to be fair you should only compare scores with disks scanned (not burned) on the same drive at the same speed.

[_H_itman]

Jamos, I know they were scanned at different speeds... I put that in my original post. I always scan at 4X. I'm not really comparing scores here, I'm just trying to figure out why the extreme difference in "QUALITY SCORE"? That makes sense about the spikes above 4, but count the number of spikes above 4 in the SONY08D1 scan, and count the number of spikes above 4 on the 1st scan.

So does CD/DVD Speed give a handicap to 8X scans? Again..the extreme difference in Quality score is the only thing I'm talking about here. This isn't a BenQ vs. Nec thread. My 1620 should be arriving any minute.:iagree: Correction... just answered the door and my 1620 HAS arrived.:bigsmile:

[Sapa]

Hitman

CD Speed takes maximim PIF into account, on the basis that one single fatal error (the PIF) could make your disk unreadable.
The total count doesn't play a role here, nor does the PIE count.

Erk Deppe stated he would reconsider the maximum PIF as the sole factor in the calculation of quality score in a future version.

[_H_itman]

@Sapa, sorry, but I just now realized you posted in this thread. Taking maximum PIF into account also makes sense, but the scans had the same maximum pif.. 7.

[Sapa]

[quote="'[H"]
itman'][b]@Sapa[/b], sorry, but I just now realized you posted in this thread. Taking maximum PIF into account also makes sense, but the scans had the same maximum pif.. 7.
[/quote]

I noticed the relation between quality score and maximum PIF on scans done on the Benq and confirmed by an answer from Erik Deppe in the CD Speed thread.
There is no sort of handicap allowed when you scan at higher speed, that I'm aware of.

Your scan with the Liteon 832 defeats my statement.

I think the code is re-written between the Benq and the Liteon parts of the software. Cannot explain otherwise.

Edit : Also note that the Danger Zone for PIF's in the Liteon scan (Red) starts at 4 already whereas it is not visble in the Benq scan. ECC 1 for the Liteys and 8 for the Benq ?

[socrates007]

Hitman,

What Jamos said above is correct. Notice on the first scan (the LiteOn one) that the PI Failures above 4 have a red background. Below 4 the color is green. That's because on a LiteOn the max PI Failure is 4 for the "safe area". Once a LiteOn gets above 4, then it's going into "dangerous" territory regarding a potential playback problem.

With your second scan (the BenQ one) the PI Failures show all green in the background. That's because your PIFs never got past this "safe" area. On the BenQ, the PI Failures "danger" area doesn't start until 16 PIFs. If you had PIFs of 16 or higher, you would have seen a red colored background on your BenQ scan in the PIF area just like the red background on your LiteOn scan.

There's a 1:4 ratio between the LiteOn and the BenQ PI Failures. That is to say, a PIF value of 1 on the LiteOn equals a PIF value of 4 on the BenQ. A PIF value of 2 on the LiteOn corresponds to a PIF value of 8 on the BenQ, etc. Since your max PIF value reported in your LiteOn scan was a 7, that would be like getting a max PIF value of 28 on your BenQ. That's why your LiteOn scan is in fact of worse quality than your BenQ scan. The max PIF value of your BenQ scan was a 7 which would be like getting a max PIF value of 7/4 on your LiteOn (7/4=1.75, a little under 2). Do you now understand this?

Jamos explained it correctly when he said, "The benq scans at 8ecc so the pif limit is 16 not 4". Whatever max PIF value you get with your LiteOn, multiply it by 4 to get a corresponding BenQ equivalent value. Likewise, whatever max PIF value you get on your BenQ divide it by 4 to get a corresponding LiteOn equivalent value. PI Failures are much worse than PI Errors regarding causes of skips, jumps, or unreadable discs. That's why Nero CD-DVD Speed uses that factor so highly in determining the quality score of a given scan.

I was tempted to say the reason your scan got a higher quality score in your BenQ drive than in the LiteOn, was because the LiteOn scan was of a NEC 3500 burn and not a BenQ DW1620 burn.

All joking aside, you obviously know that if you want to compare a NEC vs. BenQ burn you have to do the following: Test the exact same movie burned on exact same media type (one in the BenQ and one in the NEC), then scan both burns on the same drive (either the LiteOn or the BenQ, but not both) and do both scans at the same scan speed. Anyhow, I hope this helps you out.

[_H_itman]

@socrates007, thanks for explaining that. That makes sense. It's funny that up until now I hadn't even noticed those background colors.(green & red) Of course I run Kprobe much more than I do CD/DVD Speed.

That also explains some questions I had about PIF's when scanning with my BenQ. So having a PIF of 4.. or even 5.. is perfectly fine when scanning with BenQ/CDSpeed? But having a PIF of 16 or higher (on benq scan) is bad.

BTW.. which do you prefer. Kprobe or CD/DVD Speed? Just curious. :bigsmile:

[socrates007]

[quote="'[H"]
itman'].. which do you prefer. Kprobe or CD/DVD Speed? Just curious. :bigsmile:
[/quote]

I sent you a PM answering your question.

[Inertia]

...There's a 1:4 ratio between the LiteOn and the BenQ PI Failures. That is to say, a PIF value of 1 on the LiteOn equals a PIF value of 4 on the BenQ. A PIF value of 2 on the LiteOn corresponds to a PIF value of 8 on the BenQ, etc....

This is technically incorrect, socrates007, but your statement is understandable since it relates to the graph color scheme. The Liteon scans PI errors @ 8 ECC blocks and PIF errors at about (slightly over sometimes) 1 ECC block. The BenQ 1620 scans both PI and PIF errors at 8 ECC blocks. Therefore PI errors are directly comparable on the drives, but the ratio of the PIF errors is 1:8, Liteon vs. BenQ. The limit of 4 PIF errors for the Liteon matches the limit set by the standard for acceptable PIF errors, namely 4 PIF errors per 1 ECC block. Since 8x as much territory is being sampled by the BenQ, theoretically the approximate limit allowed by the standard would be 32 errors per 8 ECC blocks (4 x 8).

I don't know why Eric's coloring schemes for PIF errors are using a 1:4 ratio for the green section. It is interesting that the limit of the second color level for the BenQ is 32 PIF errors, which would be the approximate limit allowed by the standard (4 x 1 x 8). On the other hand, the comparable Liteon second level is 10 PIF errors per 1 ECC block, far above the standard limit of 4. In fact errors as high as 13 on the Liteon can still return a quality score of 47. At first glance, it appears that the Liteon results may be given extra latitude with error levels.

If I had to speculate as to why Eric has seemed to position the BenQ measurements in a more conservative way, it would be for the following reason. The standard specifically requires the limit of 4 PIF errors to be measured in 1 ECC block. Since this can't be done with the BenQ (and other non-MediaTek chipset drives), it isn't as simple as dividing the errors by the number of ECC blocks measured in the sample. If in a series of 8 ECC blocks, one block has 32 PIF errors and the other seven blocks have 0 PIF errors, this serious error in one block could go unreported by being combined with seven good blocks. I believe the standard requires PIF errors to be measured per 1 ECC block to prevent this type of problem. I think that maybe Eric tightened up the PIF allowance for the BenQ to allow for the possibility of high single ECC block errors being somewhat disguised when combined in a series of 8 ECC blocks. This is just my speculation, and I hope that Eric will let us know his thinking on this subject.

[SubversiveAgent]

Soooo.. does that mean that the LiteOn is more theoretically accurate at sniffing out and reporting PIF errors than the Benq?

[Inertia]

Theoretically, perhaps, but in my experience my BenQ has been more reliable in catching problems with some media that the Liteon passed with flying colors. My Liteon seems to have become somewhat erratic. It never was a very good burner, and now I can't rely on it for testing. Consequently, I have quit using the Liteon for testing and will eventually replace it with another BenQ 1620.

[_H_itman]

@Inertia, what Liteon burner do you have? Is it an 811? Just curious because I've seen quite a few people with 811's say that their burners don't scan very well at all. I have to say that my 832S(original) has been an overall great drive.

[socrates007]

Inertia,

The 1:4 ratio I was speaking about was in relation to reading Nero CD-DVD Speed. It was specifically about the green/red color background used for PI Failures. I know that it doesn't "technically" correlate to the actual difference in scanning between a BenQ drive and a LiteOn drive. I was trying to help Hitman correctly interpret the reason why his original scans gave the different quality scores even though they didn't make sense to him. Why EriK Deppe subjectively chose to write Nero CD-DVD Speed the way he did in relation to the BenQ drive is unknown to me as well. Maybe he will explain himself in this thread. Based on Erik's green/red color scheme used for PI Failures, the ratio is in fact 1:4 regarding "safe" vs. "unsafe" values.

What the BenQ, and a LiteOn drive for that matter, actually scan in reality is unknown to us. Even if we were given the specs from the manufacturer, there is considerable difference in tolerances between any given drive; even of the exact same model and the exact same firmware for a particluar model. Throw in differences due to various system environments (one system vs. another) and you really have subjectivity at its best. Therefore, I really can't blame Erik for making his color scheme the way he did. He had to make a decision somewhere and that "somewhere" is the criteria that only he knows. Like I said, maybe he will reveal it to us; maybe he won't though. Even if he does, any software developer has to make a "subjective" decision on how he/she is going to code things because there's no real objectivity/consistency that can be achieved for error scanning on our drives.

One BenQ DW1620 will scan a given disc and report one set of values. That same BenQ DW1620 can re-scan the exact same disc and will report a different set of values for the second scan and so on..............Then, take another BenQ DW1620 with the exact same firmware version as the first DW1620 drive and it can scan the same disc as drive one and will give its subjective value.........Even if both BenQ DW1620 drives are in the exact same system, these differences will occur each and every time a scan is performed. Throw in different drives - i.e. LiteOns etc. - and the situation is horrid. This is the "best" we can achieve unfortunately. When I hit the lottery (yeah sure), then I will get industrial grade testing equipment. Until then (probably never), I and all of us are stuck with less than reality.

My point was - who the heck cares about what is "technically correct" because we are not capable of reporting what is and isn't "technically correct" with our given hardware and software. We are seeing shades of reality, or dim reflections in a mirror; but the images are distorted. Even worse, we don't know where the distortions end and where reality begins.......ad infinitum.....
:sad:

[Inertia]

socrates007 ,

I agree with your well stated overview on this topic. Concerning technical correctness, we obviously don't have all the facts or know all the answers. It is a good start though, until further clarification is forthcoming, to clearly delineate what we do know (or think we know). :iagree:

[H]itman,

Yes, unfortunately I have the Liteon 811S (formerly 411S). As a cheap early model, it was an attractive buy influenced by prior good experience with Liteon CD-RW drives. This 811S drive has been a dud, and has been unimpressive in all areas. Obviously, converting to an 811S did not improve performance, nor did I expect that it would. Converting it was done out of experimental curiosity, since it didn't work well anyway.

I'm sure that if I had an Liteon 832S as you have, I would have had a much more satisfactory overall experience. At this point, though I wouldn't buy a Liteon just for error scanning. The BenQ 1620 is a superior burner, and with its associated testing capabilities and utilities it would be my all around choice.

[_H_itman]

One thing I wish BenQ would fix on the 1620 is having to eject a disc and re-insert it in order to run CD/DVD Speed without getting that erro message. Not having to eject and switch drives to run a scan was something I was looking forward to with the BenQ. It's not that big of a deal, but I do wish they'd fix it. And I also wish they'd fix the long hesitation when ejecting a disc after a burn. It doesn't matter if you burn a disc and let it sit in the tray for 8 hours, it's still going to hesitate and the red light will blink when you do finally eject the disc.

[Inertia]

[quote="'[H"]
itman']One thing I wish BenQ would fix on the 1620 is having to eject a disc and re-insert it in order to run CD/DVD Speed without getting that erro message. Not having to eject and switch drives to run a scan was something I was looking forward to with the BenQ. It's not that big of a deal, but I do wish they'd fix it.
[/quote]

Ejecting a disc after completing a burn has always been the best policy from the earliest days of CD-R burning. For instance, if the disc is not ejected the drive buffer may not be cleared, retaining information which is incompatible with another program or operation. A disc may not be readable unless it is ejected and reinserted. Some burning programs eject the disc by default after burning, and most others offer the option to do so. Of course, this varies with hardware and software combinations. Sometimes you can get away without ejecting, other times not. Still, the more trouble free and better practice is to automatically eject the disc after completing a burn.

[quote="'[H"]
itman']And I also wish they'd fix the long hesitation when ejecting a disc after a burn. It doesn't matter if you burn a disc and let it sit in the tray for 8 hours, it's still going to hesitate and the red light will blink when you do finally eject the disc.
[/quote]

I haven't noticed any exceptional delay when ejecting a disc after a burn. I wonder if your burning software might be involved if the delay is unusual. What burning program are you using when this delay occurs? I usually use CloneDVD and it pops up an eject option at burn completion. Selecting the eject option opens the tray without hesitation.

[_H_itman]

I spoke of the delayed ejection with blinking red light(without going into much detail) because I've seen quite a few people speak of it. I even saw one thread where someone questioned whether or not the drive was actually "completing the burn" upon pushing the eject button.:bigsmile:

So i don't know why you don't have the delay. The two main programs I use are DVD Decrypter and CloneDVD.

[Inertia]

Socrates007, et al

This post by Erik Deppe goes a long way toward explaining why Liteon scans might produce variable, non-repeatable, or seemingly unreliable results. The culprit is CPU usage. In tests that Erik ran, the Liteon 812S average ECC block scan for PIF errors varied from (average values) 1.14 ECC for low CPU usage to 8.71 ECC for high CPU usage. This relation to CPU usage obviously would have a major effect on reported results, and might make many comparisons meaningless.

On the other hand, most other drives that report PI errors such as the BenQ 1620 had ECC block scanning intervals which were constant and unaffected by CPU usage.

See Average ECC Scanning Interval