QUOTE(westgroveg @ Sep 18 2004, 12:14 AM)
My opinion is that software error recovery is more useful than slight differences in firmware or optics. I have had many drives (Plextor, Lite-on, Cyber Drive, Pioneer, LG) & when one fails on a track/position they all fail. The only times I was able to recover a CD which failed the first time was when I switched software, I used EAC & then tried Plextools which recovered all errors (which EAC couldn't) so this means that errors can be corrected outside of the drive & anytime you rip a CD with burst mode, it fails & then secure mode can rip it without errors, error correction has been performed outside of the drive.
If I am wrong please explain your vague statement JeanLuc
Error correction in terms of audio CD ripping means using CIRC circuitry ... nothing more, nothing less. Any software method you use will not perform "real" additional error correction - EAC's and PTP's success on ripping scratched CD's IMO mostly results from drive slowdown (so the drive itself will do a better job at the error correction stage).
Plextools Pro come close to "software error correction" with the extensive use of CU (or C2-uncorrectable) flags but in the end, PTP do rely on similar procedures as EAC which means:
slow down the drive and read the CU-flagged parts again according to the user settings..
I also discovered that there are CD's that might be recovered with PTP and a Plextor drive when EAC's secure mode fails to retrieve 8 matching results out of 80 re-reads - but always keep in mind that you should validate PTP's results by ripping twice (I use two different Plextor drives for that which renders the ripping process even more secure), too (you might be surprised that PTP can fail, too).
I cannot agree with your statement about different CD drives either ... on a real badly scratched CD like the ones used by CDR-Info.com (ABEX TCD-721R IIRC) you will see
significant differences between the different drive's error correction abilities ... the tested drive's results with the ABEX TCD-726R as a test disc (which does not include that much errors as the 721R and thus may be a more realistic test) range from 1) unable to rip to 2) rip without errors at all.
QUOTE(westgroveg @ Sep 18 2004, 12:14 AM)
Define a good drive?
A good drive shows the following features (order like my personal importance)
- The drive is tolerant towards out-of-spec jitter and low disc reflexivity
- The drive is tolerant towards scratches (within the limits of CIRC specs)
- The drive can report CU flags to the ripping program
- The drive will overread its read offset into lead-in/out
- The drive is fast (ripping should start at 8x min.)
- The drive is still fast during EAC's re-reading
- The drive works with different DAE programs
- The drive can "defeat" copy protections
- The drive doesn't produce too much noise during operation
- The drive should work well with unbalanced discs
Just a few points I consider when searching for a good drive ... Accurate streaming (it's the same as "non-jitter-reading" or "constant read offsets") is present with any of today's drives and caching isn't the issue since nearly every burner uses its internal buffer for DAE (and there exists the FUA command or EAC's flushing trick which should take care of that).
QUOTE(westgroveg @ Sep 18 2004, 12:14 AM)
don't think there is a very noticeable difference for extracting audio CD's with EAC's secure mode.
So as long as you are speaking of error detecting & correction using EAC's secure mode I think "a good drive" is pretty much anything on the market these days.
From my personal experience with some 30+ optical drives during the last 5 years (with all of them being capable of delivering an "accurate stream") is that there are significant differences between drives using EAC's secure mode if you have to rip "problematic discs".
Your staement is perfectly valid for pristine discs, though ...