It sounds more logical that you have to check that box
even though you only have 1 of 2...

If your drive has a negative read offset or a positive read offset correction it also needs overreading into the Lead-Out in order to have no missing samples on extraction. If the drive has a positive read offset or a negative read offset correction it needs overreading into the Lead-In.

Am I right?

Yes, the drive will never attempt to overread into lead-in.

I was curious to whether I have understood correctly that a positive read offset correction value (like +1182) means, that to be able to get all samples during extraction, the drive would only have to be able to read lead-out? In other words, the ability to read lead-in is of no use when the drive has a positive read offset correction value?

I also got curious to whether checking the box (ability to read lead-in & lead-out) for a drive that only supports one of them, would mean
a) nothing, neither lead-in nor lead-out is read
B) the one supported is read
c) risk of erroneuos behaviour

Quote

I was curious to whether I have understood correctly that a positive read offset correction value (like +1182) means, that to be able to get all samples during extraction, the drive would only have to be able to read lead-out? In other words, the ability to read lead-in is of no use when the drive has a positive read offset correction value?

AFAIK the answer is yes, reading lead-in would do nothing for you if you have a positive read offset correction (and hence a negative read offset).

We are talking about REALLY small amounts of sound here, though, if you ever lose any.

Quote

We are talking about REALLY small amounts of sound here, though, if you ever lose any.

I'd like to make clear that if you loose 500 samples of audio, the effect is perfectly audible and annoying.
The point is that about 95% of drives have a positive read offset correction. Thus what you loose is at the end of last track, which is silence 99.9 % of the time. For the 5 % that have negative read offset correction, the beginning of track 1 is silence 95 % of the time, I guess (should be checked, if someone is interested).

I tested little more, I checked overread and I got read error on the very end of last track (or sync error, I don't remember anymore), then I unchecked and everything was fine.

If you attempt to overread with a drive that doesn't support it, it can return null samples instead of data, it can return C2 errors and null samples, it can also return a read error.

I also got curious to whether checking the box (ability to read lead-in & lead-out) for a drive that only supports one of them, would mean
a) nothing, neither lead-in nor lead-out is read
b) the one supported is read
c) risk of erroneuos behaviour

If the offset correction is bigger than the pregap of track 1 (exemple : if the pregap is 0), overread into lead-in is attempted if you rip track 1.

Magic75, I'm now sure, b) is right. I have a drive that only reads into Lead-Out and has a positive read offset correction value (negative read offset). I always have the option 'overread into Lead-In/Lead-Out' enabled and the drive does overread at the end of the disc, because there never were any samples missing. I know this, because the extracted files were identical to the ones extracted by another drive that reads into both Lead-In/Lead-Out.

If the box is checked, there are three possible cases :

If your read offset correction is null. Overread is never performed.
If your offset correction is positive, overread into lead out is attempted when the drive finishes the last track.
If your offset correction is negative, if you copy image, overread into lead-in is attempted at the beginning. But if you copy tracks, there are two possibilities :

I always have the option 'overread into Lead-In/Lead-Out' enabled and the drive does overread at the end of the disc, because there never were any samples missing. I know this, because the extracted files were identical to the ones extracted by another drive that reads into both Lead-In/Lead-Out.

Quote

If the offset correction is bigger than the pregap of track 1 (exemple : if the pregap is 0), overread into lead-in is attempted if you rip track 1.

Pio2001, I hope you know that the pregap of the first track never can be smaller than 2 seconds (only longer). A first pregap shorter than 2 seconds is illegal for Audio-CDs. It is the only gap that can't be 0. (But a first pregap of only 2 seconds isn't visible to Audio-CD players.)

Quote

Magic75, I'm now sure, b) is right. I have a drive that only reads into Lead-Out and has a positive read offset correction value (negative read offset). I always have the option 'overread into Lead-In/Lead-Out' enabled and the drive does overread at the end of the disc, because there never were any samples missing. I know this, because the extracted files were identical to the ones extracted by another drive that reads into both Lead-In/Lead-Out.

No, I don't think b) is right. The reason you don't have any problems is because you have a positive offset correction, and EAC hence only needs to read into lead-out, which is supported. If you had a negative correction value it would attempt to read lead-in instead and the drive could return garbage.

So, neither a), b) or c) is correct. In fact the whole question is incorrect...

It doesn't prove that your drive can overread. I've got some drives that accepted to overread, but returned null samples after the end of the last track instead of the audio that was there. For example if I set the read offset correction to +1000, I got a wav file with exactly 1000 null samples at the end, while another drive returned some different audio data reading the same range.

So you must find a CD with data in the lead out in order to really test for overread. And you must have one drive that overreads in order to test this, or at least a drive with an inferior read offset correction. Try old AAD CDs. Or if you can't find one, burn yourself a sine on a CDR with a burner that has a write offset superior to the absolute read offset of the tested drive (this way you don't have to have another drive, since you know what you're burning).
Rip the last track with offset correction, open it in a wav editor with vertical zoom ability, and see if the signal is present until the end.

All of these three methods show me identical sample values at the end (or at other sections) of Audio-CDs. The probability that all of them are wrong in the same way (cause they are identical) isn't worth to be mentioned.

Your method is correct, but I'm not sure if you performed the first step : finding a CD with non null data until the very last sample of last track (or creating one yourself).

I largely agree. But I still think it is a good question, it's just not specific enough. You mean that not only overreading capability is important, but the combination with a positive/negative read offset correction value, am I right ?

An audio CD begins like this :

----------Track 0

Lead In

----------Track 1 Index 0 = Absolut time 0:00:00:00

I suggest "pregap of the audio session" for this zone

----------Sector 0 = Absolut time 0:00:02:00

I suggest "pregap" or "gap of the first track" for this zone

----------Track 1 Index 1

In its writing window and cue sheet, EAC adjusts the beginning of the gap of the first track at absolute time 0:00.00.[a href="index.php?act=findpost&pid=230024"][{POST_SNAPBACK}][/a]

I also noticed that on self-written audio CDs my own and one other CD player are able to search back into the pregap of the audio session.
[a href="index.php?act=findpost&pid=230024"][{POST_SNAPBACK}][/a]