Interpreting C1/C2 error scans

vbimport

#1

This thread is for discussion of C1/C2 error scans and how to interpret them.
Please begin by reading the media FAQ to get the basics on errors and general media info.

There are several testing utilities around that measure either C2 or both C1 and C2. for this discussion we are using Kprobe, which is designed to work with LiteOn burners. If you have a different burner or reading drive that reports errors in another program, by all means use it. The principles are the same. Please note however than scanning discs in a different drive than they were burned on will introduce additional nuances into the scan results. When using that method, you should give less weight to the actual error counts and focus instead only on the differences between scans.

The point of scanning is to find the combination of media and burn speed that produces the lowest error rates on your burner Each scan that’s done for the purposes of comparison should include only one change in the variables. In other words, don’t change both the media and the burn speed, don’t change both the burn speed and the read speed, change only one thing and scan again to compare results. The point is to get burns with the lowest possible error counts.

The actual error counts are not very important in and of themselves, and comparisons of scans done in different drives is NOT a good idea.
The one and only hard rule with error scans is: less is better. If you lower your burn speed and you get a scan with lower error rates, you are moving in the right direction. If one media has lower counts than another, use that media.

For our purposes, it doesn’t make a lot of difference what is being reported as “C1” and “C2” by the drive and the software. All we really need to know is that C1 is the “friendly” error and C2 is the nasty error. All discs show C1 on them in varying amounts, but a good disc should not show any C2. You will see these errors showing in varying colors on the scans that are posted, colors are set by the user and don’t mean anything. The C1 and C2 graphs will often be combined into one chart, or may displayed separately in 2 graphs. Please consult the Kprobe thread to learn about how the program works and how errors are displayed. If the charts are combined, the C2 chart will be layed on top of the C1 chart.

You will notice that the scan images you see posted are different than the screen shots in the Kprobe thread. This is because Kprobe saves the basic disc and error info with the image of the scan charts. We are looking here at the saved report from Kprobe. Now let’s look at one:

I chose this scan cause it’s so pretty, :wink: , and because it’s a good example of what you do not want to see on a scan.
The first thing to understand about Kprobe is that the graph is auto-scaled, so the errors always fill the graph. You have to look at the scale on the side to know where the max error levels fall. A scan with only max=10 level errors can look the same as one with max=1000 errors, so you have to check.

Along the bottom of the chart, you see relative time measured, indicating the location on the disc. On the left is the start of the disc, (inside of the disc), and on the right is the outside of the disc. (the end). The chart will end (on the right) at the end of the burned portion of the disc, not at the max capacity of the disc.

The error values are also displayed in the top area of the image. You will see the “max”, “total” and “average” counts showing there. Max and total are the most important values, the average count is just that, an average of all the error values on the scan. It can be misleading sometimes, so don’t fall into the trap of using the average counts as a measure of “quality”. A disc can have fairly high average C1 value, and still be a very good burn.
In VERY general terms any C1 average count under 10 might be acceptable, but higher values are not always the end of the world either. But, on CDR’s we prefer to see avarage C1 counts under 2, with max counts under 20.

The total error count is more of a measuring stick for comparing one disc to another, but again the actual value is not so important.
Here’s another scan that’s a bit easier to look at:

As you can see here, there are a few C2 and high C1 values. this disc would be considered “marginal”. It’s playable and the data is intact, but the scan indicated a problem with the media, the burner, the reader or burn/read speeds. In other words, something needs to change. In this case, it’s a RW disc so very little can be done to reduce the error rates other than lowering the read speed. In Windows, most drives will slow the read speed to correct these errors. It’s also a disc that should be considered less reliable over time because the errors can degrade to become much worse.

Now let’s look at a really good disc:

Note that the max error values are under 10, and total and average counts are also low. This is a disc you can trust. Discs with even 2-3 times these rates are considered “good enough” by most people. A pressed CD will usually have far more errors than this.

The other thing you need to pay attention to when looking at these scans is the scan speed. Speed = more errors. Most people will scan CDR/RW discs at max speed. but if you encounter a disc that has problems, lowering the read speed will make it easier to scan accurately. The read speed is also included on the image near the middle of the top portion.
Be wary of scans that are done at low speeds, as they will almost always show low error rates, and may not tell you much about how the disc will perform outside the scanning program.

The best result for any scan is this:
Full speed scan with low C1 and no C2. Some scanning programs allow the drive to slow, and also display the drive speed during the scan. Look for scans that complete at full speed with no slow-downs.

Error scanning 101:

Error scans are one of the most misunderstood things around in burning circles. People often fail to remember that the media and the burner are only 2 of several variables that effect the outcome of the scan.
You cannot always judge a burner by the error scans, or the media either.

The only truly legitimate use for error scans is to compare different burns in the same drive. In other words, finding the right burn speed, the right media, and the right firmware for your media. Beyond these simple, limited purposes, you are exceeding the limits of the data by drawing any conclusions about the drive quality or media quality.

It’s easy to spot a really good disc, and a really crappy one. But most scans will fall somewhere in between where interpretation is more of an art than science. When you get to know your drive and media very well, you know what to expect from them and small differences can be noted easily. However, every scan varies to some extent, so drawing conclusions based on slight differences in scans is not helpful and will ultimately drive you crazy.

There are at least 5 variables that affect the outcome of the error scan.

  1. the burning drive

  2. the media burned

  3. the burn speed

  4. the reading drive

  5. the reading speed

  6. Every drive has a “preference” for the media it likes to burn, which can include different batches of the same media.

  7. Every media varies from batch to batch, disc to disc, and even from the start of the disc to the finish.

  8. Every burner has a “preference” for burning speed, regardless of what media is in the drive.

  9. Every drive has a preferance for what kind of media it likes to read.

  10. Every drive has a preference for what speed it likes to read a given media at.

You cannot draw any general conclusions about any particular drive or media type based on a few scans. If you wish to use a particular media, you should try different firmware revisions and burn speeds to “fine tune” your performance if you want the lowest error rates.

Drive firmware revisions, wear and tear, variations from one drive to another, overclocking a drive, and a dozen other factors will affect how it behaves with a given media. Even multiple scans of the exact same CDR can produce varrying results in the same drive.

Another important note: Discs that are “marginal” in burn quality can sometimes exhibit remarkably different scans when scanned multiple times. On one scan, they look OK, the next time the results are very different. On such a disc, lowering the scan speed a notch may resolve the differences between scans.

There is no real evidence to support the conclusion that higher C1 rates (within reason) will adversly impact the performance of a CDR. We do know that lesser quality CDR’s with higher C1 rates may tend to degrade over time, but given a good quality CDR (that is burned reasonably well initially), there is rarely any degradation seen. The Cyanine and azo dye types have earned a reputation for greater stability over time, even with higher error rates, but this may be due to the fact that they are better made in general.

The error scans should be viewed as more of a benchmark of the performance of a given disc in a given drive, not as a test of the relative quality of either the drive or the CDR. It’s just another piece of information in the bigger picture.


GH24NSC0 - trouble reading some data discs - Switched to BDR-209DB
#2

Do C2 errors imply bad sectors?
Do bad sectros imply C2 errors?


#3

Short answer: no.

Every drive has different error correction ability, but all drives will fail when C2 reaches a certain level.


#4

What is a bad sector then?


#5

“Sector” refers to hard drives, not really a correct CDR term. If the error correction data is corrupt, then that block of data is unreadable. It may, however, be readable in a different drive.


#6

what are cu errors, reported by plextools?


#7

In VERY general terms any C1 average count under 10 might be acceptable, but higher values are not always the end of the world either. But, on CDR’s we prefer to see avarage C1 counts under 2, with max counts under 20.

can you provide reference values for c1 and c2? like it was done with PI and PO erros (DVD), i want to know till when audio CDs(data Cds) will be redable by most of devices

tank you


#8

There are no specs for C1/C2. No C2 are acceptable, but very low C2 are generally not problematic in most players. C1 values do not appear to affect playback, but are considered an indicator of a disc’s reliability. In other words, high C1 should be considered an indicator that the disc will eventually degrade and fail. Iv’e seen discs with C1 max values in the 1000 range that played just fine.


#9

As far as I know C1 values should not exceed 220 per second, and C2 values should not exceed 1 per hour(!).

I do have CDs however with a few thousand C1 errors per second and thousands of C2 errors per hour that are still readable without any major problem.


#10

I am somewhat perplexed in some of the testing I have done. I oversee a small production operation in which I sell about 500 CDs a month of CDs produced on an Imagemaker 150 by Microtech (4 Plextor burners and Rimage Prism thermal printer). I recently noticed high error rates on a lot of the CDs that I tested on a Clover Systems CD Analyzer (bought about 4 years ago). After exhaustive testing I narrowed down the cause of the high error rates to the printer. I would get excellent burns and then print on the CD and test it again and the error rates skyrocketed. I replaced the printer’s printhead and the error rates dropped dramatically for a time. I have assumed that the printhead got out of spec and was too hot. I am starting to see the pattern return, although not quite as bad. However, I don’t know what to make of the different readings I get with different scanning speeds when testing the CDs. A little while ago I tested a CD at 8x speed and got a grade A (Avg. BLER not greater than 5, Peak BLER less than 100, no E22 or E32).
I then tested the same CD at 4x speed and got a C grade with a total of 88 E22 errors. I have taken some of the CDs that got the poor grades home to test on my Plextor PX-716A with Plextools and the errors tended to be lower and more in the acceptable range. The fact that I am selling these raises my anxiety levels and I need some peace of mind. Which readings do I trust?


#11

@stevenhal
Unfortunately, both scans are “correct”, for the drive they are done in. I take it that after the CD’s “rest” a while, they don’t improve? Suggest trying different media, you don’t say what you are using.


#12

I did consider the possibility that the CDs needed to rest after the printing but it didn’t make any difference. My main media is white thermal printable Taiyo Yudens. I also tested other thermal printable media like Mitsui, Prodisc, and the silver laquer Taiyo Yudens. They all exibited the same pattern: Good results on scan tests before printing and lousy results after printing. Changing the printhead on the printer did make a big difference. The degradation is not nearly as bad, yet there still is some. Since my original post I tested the CD that had inconsistent test results at the 4x and 8x in the Clover Systems Analyzer on my Plextor drive at home. The following are the results:

Scan at 4x: C1 - 2.4 avg/sec., 32 max/sec., 10959 total, no C2 errors
Scan at 8x: C1 - 2.4 avg/sec., 32 max/sec., 11257 total, no C2 errors
Scan at 10-24x CAV: C1 - 2.5 avg/sec., 34 max/sec., 11749 total
C2 - 0 avg/sec., 10 max/sec., 12 total

These scan results seem to show a bit more consistency than the Clover Systems drive, which I believe is a modified Plextor drive. I guess my bottom line is do I really have a problem here? Even though “both scans are ‘correct’ for the drive they are in” is it really representative of all other drives being used out in the real world? I need to be sure that our customers are getting decent quality. I could invest in a new printer at a cost of about $2400 but we are presently planning on switching our operation to a new system already purchased that will also support DVD burning which has an Everest printer. Unfortunately it will be a few months before all the programming is completed. (It’s an automated system where customers create their CDs online and they are automatically burned from files on a server on the network)

One other interesting note regarding the Clover System’s drive. I had absolutely horrid test results on the Prodisc media after printing with the old printhead. Got graded with an “F” with a lot of E22 and E32 errors. However these same CDs tested with a grade eqivalent of “C” on my Plextor at home with some C2 errors but no CU (E32) errors. It’s as if the drive in the Clover Systems drive absolutely despises the Prodisc media. The million dollar question is are these CDs really bad? I guess I should just try and play the files (about 10 hours of MP3 files) and see what happens.

Thanks for the info in your posts. Very helpful.
Steven


#13

I always scan CD’s at 40x, low speed scanning doesn’t always separate the good from the not-so-good. As I said before, both drives are giving you accurate scans - for the drive. Clearly it seems your thermal printer is cooking the discs, and this is not good. To do further testing, suggest picking up a LiteOn CDRW drive for about $35 at NewEgg and using it to double-check scans with KProbe2.

Your Plextor scans are perfectly good, though too slow for my taste. Since you are using good quality media, I think you can hope that they won’t degrade much over time. One thing, if it’s an option, is to reduce the amount of label being applied to the disc. (smaller area)


#14

Well I tested a few discs today coming off the system and they all tested good with a grade “A” at 24x speed scan on the Clover Systems Analyzer after having been printed on with the thermal printer. This printer can be like a person, hot one day and cool another.
It does appear that the new printhead has solved the bulk of my problems. I did test one of the discs that got the “A” grade at 24x speed again at 4x speed and got a “C” grade because of 16 E22 errors which were picked up at the slower scan speed. I have read that faster scans will result in higher error rates yet the opposite seems to be happening here. I think I understand what is stated in the Clover Systems manual when they say that “the disc itself does not have an error rate; playing the disc produces errors” which in essence is what you have been saying. I think I will take your advice and purchase one of the LiteOn drives for comparisons.

Thanks again


#15

I’m using old Teac W512EB cd-burner and LiteOn 1653 dvd-burner. I have a lot of no-name CD-R discs. The disc surface looks like metal-AZO. Nero CD-speed and kprobe quality test results are strange – 0 C1, and average (!) 80 C2 errors, but there is no problem with reading of data. What that’s mean???


#16

This is a great interpretation, but can u point me to one as good that uses CD Speed as the measuring tool? The help does not help, and there are so many posts here, it’s hard to find a good one (many are based on very old versions of CD Speed). thanks!!!


#17

I’ve had some “gold” noname CD medias since the mid 90s. I never thought any good about them, as nothing is printed on, and therefor I don’t know what brand it is. But the bottom side, which always gets scratched is of a golden color. I took some of them, and run them though Q-check with my PX-755A drive, and guess what. Those medias have the same readings as my newly burned Verbatim quality medias, and they are 10 years old. Used out for DJ’ing, and is a little scratched. But the result surely surprised me a lot, never expected them to be that good after all these years.

Haven’t seen any “gold” CD’s even since that time, only the blue and green color CD’s.

A very good artical. Though I’m a little unsure about the C2 errors, as most of my medias contain them. The Cd’s I’ve checked are all CD’s that has been saved in those thin CD sleeves. Some of the sleeves has some protection filth build in, and some not. But for sure the CD’s gets a little dirty and also gets some fine scratches from pulling them out of the sleeves. There is also some fingerprints on them, which I have removed with a little Isopropanol on a very fine micro fiber cloth, that don’t scratch my CDs. But I can’t get rid of the C2’s, must be the small scratches and many of the pressed Cd’s is 10-15 years old.

But I don’t plan to start making copies of all of them (>5000 cd’s). So I started to take the worst CD’s, where the C2 error peaks way out of the scale in Plextools, and ofcause where there are CU errors.

Moef


#18

I have a very general question about interpreting CD-R error scans. I am archiving my favorite music to CD-R as WAV files. I burn the files in Mode 1 for maximum error detection and correction. After trying various media, I settled upon MAM-A gold as my archiving medium, as it consistently gave the lowest error rates. Taiyo Yuden was a close second. My interest here is in obtaining the lowest possible error rate, since this presumably gives the greatest longevity of my archives.

Having read hundreds of articles in this forum and elsewhere, there is still doubt in my mind as to how to select the optimum burn speed. Some experts say: test burn some discs at various speeds, and see which gives the lowest error rate. Then stick with that.

Others say: If another speed has a somewhat higher error rate, but its beta is closer to zero, then choose that. The disc will be readable on a wider variety of machines.

Still others say: Watch out for jitter. You want it to be low and flat. Make sure the jitter looks good, even if the error rate isn’t minimized.

To complicate things further, my burner is a Plextor Premium. Plextools does not give me an absolute number for jitter, but only a nebulous “high-to-low” graph. Even the Plextools manual is of little use to me in understanding how to interpret this.

Can someone give me the big picture here? How can I best interpret the results of C1/C2 and beta/jitter scans in Plextools to decide on a best burn speed (for greatest data longevity)? Despite the countless articles I have read, no one seems to adequately pull it all together.

I have a host of scan data that I can post, if someone would like to help me interpret it.


#19

@Jaslane

Your questions will be answered in the thread entitled “Secrets of C1/C2/PI/PO error and beta/jitter tests”, for the issues are highly related to what are being discussed over there.


#20

You have “analysis paralysis”. Burn at 32x and be happy. 32x gives the best results with the widest range of media and CD burners. 24x might be a better choice for DVD burners in general. Longevity of your discs is a complex issue, but with reasonable error levels (of any type), there’s no evidence that longevity is affected by small differences. Scan a few pressed discs and use that as a benchmark. If you’re below that level, don’t worry. Odds are that in a few years, those CDs will be outdated anyway, and you’ll be converting to some new format.