DW1620 problem: amount of C1 errors depends on reading speed

Today I finally got to testing some CD-R’s on my second BenQ drive. The first one had some read back problems with DVD+R’s: one time it said a disc had 19000 PIE’s and another time the same disc had 6000 PIE’s.
Luckily my second BenQ drive doesn’t have this problem, although I’m experiencing something similar with CD-R’s now.

I began with a burn @ 40x and tested the disc afterwards. I got a VERY good result @ 40x on my first BenQ drive: 1918 C1 errors, so I hoped my second BenQ drive would give about the same result.
This wasn’t the case at all: 8100 C1 errors. I burned another disc @ 16x: 6066 C1 errors. I noticed the errors increase rapidly after 40min on the graph.

I began scanning other discs that didn’t show this behaviour on my first BenQ, some of them have the same thing, others not. Strange.

Then I noticed something: whenever my drive slows down, the amount of errors decreases. I marked this on the first picture below in yellow. So my first thought was to scan the disc at a lower speed setting, and you guys can see the result below. I tested this with two different discs, same results.

Now, how must I interpret this? Is my BenQ a crappy CD-R scanner at higher speeds?

At two points I actually got < 2000 PIE’s when scanning this disc at 40x, but most of the time the graphs are like below. I don’t have this with all my discs either, some of them don’t show this behaviour even after several scans. I thought it was a batch problem but both TG1175’s and TG1313’s can have this problem, so at the very least I know they’re not from the same package (still have figured out the true meaning of that serial).
It’s not a writing problem either, because both discs burned on my Plextor and my BenQ have it. And also all tested discs had very low PIE’s on scans with my first BenQ.

This is the original scan on my first BenQ drive, right after the burn:


C1/PI seems to increase at higher speeds. Maybe we need a tool like Sandra and 3DMark for testing optital media writing quality. :slight_smile:

I have also just completed a cd-r test myself (small so far).

It is only one Disk but I am getting some interesting results…

It’s a HP 48x CMC disks. Used Nero CD Speed to create the disk.

First scan is with BenQ maximum speed, looks pretty bad.

Second scan is Kprobe with Liteon 166s, looks really good.

EagleClaw (or anyone else), can you do some tests with the BenQ and CDSpeed at lower speeds? On your graph, the C1 errors don’t seem to go down at the slowdowns so I don’t think lowering the read speed will matter, though I could be wrong. But I’m still curious. :slight_smile:

So, is it safe to assume that my scans @ 16x are more reliable than the ones @ 40x? They match the scans @ 40x of my first BenQ drive, as you all can see above. Hmz, I guess I’ll contact BenQ about this.

Do a scan with your Liteon DVD-rom with cdspeed en you will notice higher value’s en big spikes (usually at the beginning).
I always use kprobe for testing my cd’s.

I scanned with benq at 16x speed with nero and it was about the same as set at max speed.

I scanned with the liteon in Nero at max and 16x and they both looked like crap.

I think the cmc is just crap on the Benq?

Look at these Ritek 48x burned at 40x

Nero scan done at max and Kprobe done at max

That’s what I thought, probably just the media.

I guess my second BenQ drive is a weirdo too. Anyone else experiencing something like this?

I don’t want to in any way take over your thread here but I did some more scanning.

I made a Data disk with nero “Maxell” 48x -Ritek Co. disk.

What one of these scans can we trust as being the most acurate?

I usually just scan with Kprobe using my 166s.

Generally a CD/DVD burner drive is considered to be the one that’s the most accurate when doing a read scan, so the BenQ 1620 is the one that I’d vote for.

It’s my belief that a C1 error maximum below 20 along with no C2 errors represents an excellent burn. I prefer to limit my CD writing speed to 24X, since the burn quality looks a bit better for me at this lower writing speed, and the extra time (about 1 minute) is minimal for my requirements.

Would you look at this…

The disc below has been burned on july 13th 2003, probably on my Plextor drive.

First picture below is the scan I made on my first BenQ drive. Second picture is a scan I made a few minutes ago on my second BenQ drive. Notice the yellow markings on both pics, which mean that errors decrease with lower reading speeds. Also notice the shifts. Third picture is a scan I made after the second pic, but I lowered the read speed to 16x.

Conclusion: seems both drives have the same issue, scanning discs at high reading speeds is unreliable. Can anyone confirm this, or am I just having bad luck with drives?

edit: I also noticed right now, that the jitter seems to go straight down at most of the places where the errors decrease when speed drops to zero…coincidence?




Hi Hitokiri,

It looks like the drive is doing regular recalibration during reading. I’m not sure why this is, as I would have thought that recalibration would only be needed for writing - this might be some sort of a firmware bug. The points you’ve circled are NOT error points - just temporary slowdowns based on this regular recalibration. Furthermore, it seems that this recalibration does NOT occur when you scan at 16X.

Your CDSpeed quality score is 98% (excellent) for each of the above scans - based on the C1 error maximums of 14, 16, and 14 respectively for the three graphs.

Note that the laser is burning pits onto the disc - if you were to magnify a burned disc with a microscope, it would look a bit like craters on the surface of the moon. So it stands to reason that the faster the scan speed the more difficult the reading process becomes.

The error totals count doesn’t really matter all that much, since all CD readers have hardware error correction built in - it’s the maximum value of any individual error that’s important, since it might exceed this error correction threshold.

All that said, there appears to exist a MAJOR weakness in the 1620’s error reporting in that it appears that it does NOT report C2 errors. Not only that, but the 1620 does not stop scanning when it hits a hard error spot on the disc, so you DO NOT KNOW when you have a bad disc.

To determine whether or not a disc is fully readable you must use the ScanDisc function and check for any red errors. The BenQ engineers should probably do some work in here to improve things.

Hi Spartane,

A few remarks.

The points you’ve circled are NOT error points - just temporary slowdowns based on this regular recalibration.

I think you’ve misread what I wrote. The yellow markings indicate places where the amount of errors decrease due to a slowdown in reading speed (gaps in the graph). If the drive didn’t slow down or recalibrate itself, there wouldn’t be such gaps in the C1 graph. Seeing that gave me the idea to test the discs at a lower reading speed.

Note that the laser is burning pits onto the disc - if you were to magnify a burned disc with a microscope, it would look a bit like craters on the surface of the moon.

I always thought this was a common misconception and that in fact the laser only darkens the areas that would represent pits?

So it stands to reason that the faster the scan speed the more difficult the reading process becomes.

True, but how would I have gotten that 4th scan at all then? That one was made @ 40x. I have many other scans like that, made with my first BenQ and some with my second drive.
It seems that at one time the drive can scan a disc perfectly @ 40x and another time it just exaggerates with C1’s. There’s no “inbetweens”.

All that said, there appears to exist a MAJOR weakness in the 1620’s error reporting in that it appears that it does NOT report C2 errors. Not only that, but the 1620 does not stop scanning when it hits a hard error spot on the disc, so you DO NOT KNOW when you have a bad disc.

What do you mean? The 1620 does have C2 error reporting capabilities, my discs probably just don’t show any C2 errors because there aren’t any. Wouldn’t that be VERY bad? With data discs that would only leave the ECC to correct errors. And non-data discs would be damaged. :confused:
I have a scan of a disc that degraded over time WITH C2 errors, so the drive is certainly capable.

To determine whether or not a disc is fully readable you must use the ScanDisc function and check for any red errors. The BenQ engineers should probably do some work in here to improve things.

If I’m not mistaken, the scandisc function only reports uncorrectables (the final step in C2 decoding), green means no uncorrectables, yellow or red means uncorrectable. The difference between yellow and red is the density of the uncorrectables. Thus the disc quality test would be far more accurate.

Though I could be entirely wrong, I’m not much into the matter.

Hi Hitokiri,

You’re right, I did misunderstand. I can see that the errors did indeed drop at these recalibration points.

I saw some pictures a long time ago that showed various kinds of CD burning issues. Some discs had pits that had ragged edges, some had pits that look stretched, and others looked like bolders on the face of the moon. I don’t know if this is still true with modern burners or not, but in general they all will show increased errors at increased scanning speeds, although this effect appears to be perhaps minimal with some of the better quality discs.

The fourth scan might be better for a couple of reasons:

  1. Maybe the burner/system temperature was cooler for the fourth scan. There have been stories of people getting better scans a few hours after their burns complete.
  2. Maybe the disc was perfectly centered the fourth time. Just a slight offset could cause a minor error increase.

In either case your burn quality is excellent despite this variation.

I’m surprised that you’ve seen C2 errors - if you still have the disc, would you mind rescanning it with the current firmware and posting the scan, since I have yet to see any with my drive, and seeing is believing. I think that you’re right about the ScanDisc not reporting C2 errors - I’ve just assumed that it didn’t report them because they weren’t being made available by the firmware.

I’ve attached a couple of scans made from a bad CD that fails at the 85% point. One is the “disc quality test” scanned at 16 times so as to eliminate those annoying recalibration points, and the other is a “surface scan”. Note the single spike of 187 and note the absense of any C2 error reports on the “disc quality test”. Also note the “surface scan” shows a hard error at the 85% point. It’s really easy to either miss or to just ignore these single-spike errors. However, this disc will not read in most readers (but will in a few).



I’ve seen such pictures too, but I think they were badly pressed discs.

The fourth scan might be better for a couple of reasons:

  1. Maybe the burner/system temperature was cooler for the fourth scan. There have been stories of people getting better scans a few hours after their burns complete.
  2. Maybe the disc was perfectly centered the fourth time. Just a slight offset could cause a minor error increase.

Heat definitely isn’t the problem because my scans altnernate randomly.

I’m surprised that you’ve seen C2 errors - if you still have the disc, would you mind rescanning it with the current firmware and posting the scan, since I have yet to see any with my drive, and seeing is believing.

I’ve included the scan below. This disc has rotten away over time and I can’t get it to work in any of my drives.

I think that you’re right about the ScanDisc not reporting C2 errors - I’ve just assumed that it didn’t report them because they weren’t being made available by the firmware.

You probably mean the disc quality test. :slight_smile:

I’ve attached a couple of scans made from a bad CD that fails at the 85% point. One is the “disc quality test” scanned at 16 times so as to eliminate those annoying recalibration points, and the other is a “surface scan”. Note the single spike of 187 and note the absense of any C2 error reports on the “disc quality test”. Also note the “surface scan” shows a hard error at the 85% point. It’s really easy to either miss or to just ignore these single-spike errors. However, this disc will not read in most readers (but will in a few).

That’s interesting. If it’s a bad disc, then the disc quality test should normally report this. I see a huge C1 spike at 65 minutes, that should be the spot that scandisc indicates as unreadable. But I really wonder why it doesn’t show any C2 errors. Maybe because you didn’t use the BenQ drive for the scandisc? I tried to do one but I got an error that the drive cannot report damaged sectors.
If I’m right about this, then it just means that your other drive can’t correct that spike and reports C2 errors while the BenQ CAN correct it. That’s why you haven’t seen any C2 errors: the BenQ drive is an excellent reader (except from the problems I’ve been having).

If you look at my graph, you’ll see a small spike of 1 C2 error at 49 minutes. At that point the C1 graph reached 148. But it seems that even this amount of C1 errors (and higher) can be corrected by C2, for example at 60 min. Interesting. :iagree:


I’ll bet your disc deteriorated because it was exposed to the sun for a long period. A friend had some older unburned 8X Smart And Friendly CD’s in a spindle that was exposed to the sun. You could see a dark area all along the outside edge on one side of the discs. He still wrote to them - he just avoided filling them!

I’ve made an assumption that damaged sectors = C2 errors, so I did indeed mean ScanDisc. Perhaps that assumption is not valid…

The BenQ 1620 drive was definitely used for both images - I scanned them both just before I uploaded them here. And I’ve seen this same effect with DVD scans - zero PI Failures, but a red sector failure on the surface scan. The BenQ 822 also scans this disc similarly to the 1620 with the same perfect C2 error count of zero, yet a red sector in the surface scan.

Neither the 822 or the 1620 BenQ drive can read the file on this disc - it is an uncorrectable error and the read fails at that point. However, I believe that I was able to read it on my previous LiteOn drive (not certain about this, though).

Thanks for posting the scan - seeing is believing. From your scan it appears that the drive was correcting many C1 errors in excess of 160 without corresponding C2 errors.

The sun hasn’t touched the disc for months, maybe years. I stored it in a carton box at room temperature. Apparently some discs just die. :slight_smile:

The BenQ 1620 drive was definitely used for both images - I scanned them both just before I uploaded them here. And I’ve seen this same effect with DVD scans - zero PI Failures, but a red sector failure on the surface scan. The BenQ 822 also scans this disc similarly to the 1620 with the same perfect C2 error count of zero, yet a red sector in the surface scan.

Strange that I can’t do a scandisc on my BenQ. Maybe it’s not supported with firmware B7M9?

You should start a new thread about this issue in the Nero forum and/or contact ErikDeppe, the author of CDSpeed. Maybe he or others have some answers. I’m also very curious about this.

Dear god, took me a while but I finally noticed I wrote some errors in my first post.
My excuse: it was late at the time, I even posted this thread in the Plextor forum by accident first. :bigsmile:

Anyway, ofcourse I meant C1 wherever I typed PIE!!! :eek: But you guys probably already knew that. :wink:

So…is there ANYONE experiencing something similar here? Has anyone actually tested several CD-R’s on this drive at different reading speeds? :sad:

I contacted BenQ, they suggested I test it with the newest version of CDSpeed…same result. I hope they have other suggestions.

I have same results. On different discs… Dont know why =( Have latest fw and cd-dvd speed