Hello, I've been searching for various information on various CD-ripping obstacles and I just want to make sure that I didn't miss anything:
1. the offsets
2. drive caching
3. built-in c1/c2 error correction

Suppose I know my drive's offset, and it doesn't do cache and error correction. Will I be able to perform an exact rip from a clean CD?

P.S. I'm sorry, this has probably been asked a million times, but I failed to find a _complete_ list of possible problems, so I ask...

If your drive caches audio data and the respective option is not set, it will keep returning the data it first read when EAC attempts to reread the sectors (no physical disc access will be performed).
If you enable C2 and your drive does not support C2, EAC will miss some errors.

You must have a CD drive that overreads into lead-out when your read offset correction is positive or you must have s CD drive that overreads into lead-in when your read offset correction is negative.
Most of the CD's begins and ends with silence so this is not necessarily required but when you want to rip a CD which begins or ends with some signal your drive must have the above features...
The problem is, that what I remember the only drives thet feature the above... I remember are only Plextor's...

PS: Please correct me if I'm wrong...

In fact, overread I/O is carried out by "true" Plextor's ... 5224 CDRW, PX-504A (rebadged NEC ND1100A) and PX-166A/A2/A3 (rebadged Pioneer DVD-ROM drives) do not overread at all IIRC.

Additionally, I remember my former Yamaha 3200E being able to overread I/O, too

For questions on EAC drive settings, you can visit Uli's OffsetBase at

http://www.offsetbase.eac-audio.de/offset-en.php

+ the cd-rom must support overreading lead-in/lead-out as mentioned by k.eight.a and we've got no issues remaining, right?

So the PX-712A is not capabile of overreading into LeadOut? Also, do you have any idea if the LITE-ON LTR-52246S or the Plextor PX-712A can overwrite into LeadIn or LeadOut?

Well, the 712A is a true Plextor ... thus it can Overread I/O, as well as the 708A and the Premium.

From my experience, the Premium can overwrite I/O ... from my findings (I re-rip any CD I have burned), I still have to experience a CD that differs from the source after writing (and there are a lot that do not start with pure silence.

My LTR52246S cannot overwrite I/O, though ... but I do not care about 6 samples anyway ...

Offtopic:

I don't understand why are the manufacturers creating drives which can only overread into lead-in when they have positive offset correction and vice versa... What's the use of this "overread"?

À propos offsets... Why do offsets actually exist?

Around 1980, when CD hit the market, no player had a constant offset. Thus it was impossible to setup a given offset as reference for everyone.

Is there any possible reason that CD readers need to support overreading both Lead-In and Lead-Out in order to properly enable EAC's Overread option?

Can any drive with positive offset correction overread leadout (or vice versa)? Not that I know of. This is not some kind of rule, this is a wish from us users (we prefer a drive that meets our wish).

If your drive can read both leadin/out, fine, but it is not needed because one of them is already enough (for example the leadout for a drive with positive offset correction).

The reason behind this is that by setting an offset correction you force the drive to read out of its normal range, so it will end up in the leadin/out (depending on the sign of the correction) and there it needs to be able to read ("overread") the audio. If it can't, the (small amount of) audio that is situated in the overread area on the CD will not be ripped.

EDIT: overreading is not so important, since we are talking about a very small amount of audio.


No, I don't think so.

Then offset correction also must not be too important?

That's true in some point of view, ask Pio2001...

Offsets are only important if you use AccurateRip or for some reason must have bit-for-bit accurate audio CD copies.
edit: typo

They are indeed comparable, with the difference that offsets affect all tracks and overreading only one.
A typical offset would be 500 samples. The largest offset I've seen so far is roughly 2000 samples, which is equivalent to 50 milliseconds (=2000/44100). This is so small that in practice you won't hear the effect.


I know what you mean, it's a bit of a problem. That's why I've written a new guide for EAC (see my signature). It tackles, for example, the issues of this thread, including offsets.

Actually, the effect can be perfectly audible. When Nero cut between 0 and 587 samples of audio between gapless tracks, many users heard it and complained. With a well chosen sample, an offset of 1 sample might even be audible. Think about a track beginning with one sample at full scale, producing a click. Apply an offset of one sample, and the click is gone !

The point is that commercial CDs are already offsetted between them, and they are nonetheless mastered so that they sound properly, whatever offset the factory introduces in the final pressing. That's why offsets don't cause audible problems : they occur (on purpose) in silence, or during continuous parts. Rarely does a track begin sample-wise on a guitar note. In this case, the lack of a negative read offset correction would be very audible, because the "ping" noise of the guitar string being hit would be missing.

Good remark.

To put this clear:
500 samples (10 ms) is not neglectible. Why do we say then that they are inaudible? Well, because in practice they occur in places where they don't harm (silence in the leadin/out, silence or monotonous music between the tracks).

BTW, that is actually why my last post said "in practice" (in practice you won't hear it.

That's also why it is necessary to pay attention to home-made compilations from cassette or records. The one who made them may very well have put the track markers exactly at the place where the sound starts, which can cause problems if the write offset or the read offset correction are negative.

Hi

I've got an Plextor Premium CD-RW writer. I bought it to catch even the few samples usually lost when the drive is not able to overwrite into lead I/O. But some day a guy on digital-inn.de told me that even the premium is not capeable of gettin' all data from the lead I/O. He said my drive was not able to read far enough! Can anyone confirm this statement?

wbg. Fabian Stöckl

PS: Is there anywhere a good summary of information for users that want 1:1 copies (including offset correction, gap setting, cue, lead I/O and all the other important stuff I don't know about yet?)