C2 pointers

[sredmyer]

I originaly posted this in another thread (here http://forum.dbpoweramp.com/showpost...7&postcount=24) but thought it might be more likely to be seen if I gave it its own thread.

2 passes for C2 pointer which are supported well should be enough.

I see alot of comments like the one above regarding how well a drive does (or does not) support C2 pointers. Comments that seem to indicate that for a drive that supports C2 pointers well, Batch Ripper/CD Ripper should be configured to use them. On the other hand those same comments seem to indicate that if the drive does not support C2 well (or at all) they should not be used. I also know that the configuration allows us to test if our drive is reporting C2 pointers.

Question is, how do we know just how well C2 pointers are implemented by our drive. To further confuse this issue (for me anyway) is the fact that although the drive may support C2 pointers well (according to the both the manufactures documentation and Spoon), dbp may say it does not. I understand that this is usually due, not to the drive, but the more likely to the IDE/firewire or IDE/USB bridge being used. So a drive that supports C2 pointers very well (ie. an older "real" plextor) attached to a USB/IDE or Firewire/IDE bridge which either does not pass them or does so in such a way that Spoons code can not read them, may result in sub-standard support of the C2 pointers.

So the question is, how do we know how well our particular setup (drive and any interface hardware) handles C2 pointers. Then once we know that, what level of support for C2 is sufficent to make using them worthwhile and below what level of support does using them cause more harm than good?

[Spoon]

Re: C2 pointers

Unless C2 pointers interfere with Ripping (sometime the whole track will mark as c2 errors for every frame), they should be used.

Testing the quality of c2 pointers is very difficult, you could only do this if you had many drives and the same damaged disc to compare.

[bhoar]

Re: C2 pointers

Testing the quality of c2 pointers is very difficult, you could only do this if you had many drives and the same damaged disc to compare.

Indeed.

It's really only feasible to test C2 pointers with many days of work and following something like the following setup:

1. Three identical CDs purchased at the same time (to ensure they are the same pressing).
2. Verify the CD(s) are in accuraterip with a good matching score.
3. Damage one CD to the point where most drives cannot reliably rip an accurate track off of it (but some can).
4. Damage one CD to the point where most drives require rereads, but in generally they can still get an accurate rip.
5. Leave the remaining CD pristine.
6. Handle all discs with gloves in a touch free and dust free environment during the tests.
7. Run tests with and without C2 pointers as well as with low *blooper*s of rereads vs. large *blooper*s of rereads (increased *blooper* of matches required) across manufacturers' drives, including multiples of each make/model/firmware release.

This kind of test should cover most C2-impacting failure modes (lost sync, passing of unflagged but erroneous data, incorrect error flagging of correct data, drives that are able to internally reconstruct data better "around" an error (whether C2 reported or not), etc.).

Note also you're always testing an entire system: ripping software, filter drivers, device drivers, OS, interfaces/bridges and the drive itself.

-brendan