For cfitz and KCK.
In the PlexTools Professional Help file, it states:
C1: indicates BLER, which is the number of *E11+E21+E31.
C2: indicates the number of E22.
CU: indicates the number of E32.
These are the number of frames that have correctable or uncorrectable errors, counted per second.
The existence of C1 and C2 errors on a CD is perfectly normal, since they can be corrected by the drive's CIRC logic. However, these numbers should not be too high since they can create uncorrectable errors when the disc's playability decreases (for instance, by means of dust, fingerprints, scratches, etc.).
In case of CU errors, there is a chance that the data will become unreadable.
*Edit- C1 errors are labeled incorrectly in the PlexTools Professional Help file. E12 & E13 should be E21 & E31. The quote has been edited to correct this error.
It is interesting to note that Plextor is disclosing their adherence to an accurate definition and method of determining BLER. That is, it is not C1 errors per se that are being counted but the number of frames
that contain C1 errors of any type. This is the correct method for gathering information to determine BLER. Instead of the misleading acronym for BLock Error Rate, BLER is actually the frame error rate. Each block (sector) has 98 frames (24 bytes in each frame).
This would also seem to imply that all sectors are being counted during error measurement, as a frame is the smallest divisible data container on a CD. This may be one of the differences with KProbe, and may somewhat explain why KProbe may allow testing at the higher speeds. KProbe's high percentage sampling may allow it to run at the fastest speeds, while Plextor may be doing a complete error count. This is of course speculation on my part.
As to the question of why only E22 C2 errors are disclosed, at least they have disclosed what they are measuring. Personally, I have no problem with this method, as it differentiates between the less serious E12 C2 errors. One perspective is that E12 errors would be primarily represented by the E31 errors included in the C1 BLER (frame error rate). Under normal conditions, most E31 errors would be corrected at the E12 stage.
Finally, CU reports E32 errors, the type of error where data may become unreadable. This is Plextor's definition, following the standard practice. Disaffected gadflys, please complain to Plextor if you wish to correct someone.
The thing that I like about the Plextor error test is that is possible to calculate BLER, which is probably the most often quoted CD error measurement reference. I like KProbe a lot, but I would like to know more about what is actually being measured, or if it can even be used to calculate BLER.
As I understand it, BLER is not calculated in the Plextor test, but the raw data is there. My method to calculate BLER is to divide the total frame errors by the number of seconds on the test disc. Using the data from the Plextor Plex Tools Help, total C1 errors of 22,585 divided by 4,440 seconds (total seconds in a 74 minute disc) would result in a BLER of 5. The ISO standard for BLER requires a 10 second average, but the calculation above is the average for the entire disc.
All in all, GUI and other user preferences aside, I am happy to see Plextor involved in the error measurement business for the home user. As more information is made available as to what is actually happening in these error testing programs, the art will advance. It is always useful to have options and competition in order to advance the field.