Combined offset in EAC

[Psytopia]

Hi, do I really need to follow the steps for the combined offset in the tutorial? I have set the correct offset with Accurate Rip, shouldn't ripping the data track with IsoBuster and ripping the audio tracks with EAC/Accurate Rip be enough? Or is there something I'm missing?

Thanks in advance for your help!

[amarok]

Sorry, but yes, it is mandatory. Using the combined offset corrects not only your drive's offset, but also minor inaccuracies the mastering plant made when pressing the disc, so to say. So why do we correct these inaccuracies? Well, it's possible that different pressings of the same game have different factory write offsets, but are otherwise 100% identical. It'd be stupid to have duplicate entries for these, so we remove (i.e. correct) said factory write offset.

Don't worry, even though we "modify" it, the dump is still accurate, but more in a sense of how the disc was intended to be mastered if it wasn't for these inavoidable accuracies, and not how the disc actually came out of the mastering plant. I hope this makes sense to you

Besides, your dumps would be incompatible to all the other dumps in the DB. You wouldn't be able to match anyone's dumps, and neither could anyone match your dumps.

[Psytopia]

If I understand you correctly, this is ment to make sure the data track ends at the right sector and the audio track starts with the right one. Makes sense!

Is there no write offset for data tracks, only for audio tracks? If I get a negative factory write offset, doesn't that mean, the data track and the first audio track are overlapping?

[r09]

Data tracks have the same offset, but it gets corrected automatically by the drive, using the sync marks (those 00 FF FF FF... at the start of each sector) as reference.

[Psytopia]

Data tracks have the same offset, but it gets corrected automatically by the drive, using the sync marks (those 00 FF FF FF... at the start of each sector) as reference.

I see! And since there is no sync mark on the audio tracks, IsoBuster doesn't automatically know where the data track ends? Am I understanding that right?

[amarok]

There is an offset for data tracks, too. Drives recognize and correct it automatically. For an audio disc it doesn't really matter if your music is shifted by 1 or 2 hundredth of a second, at least from a pure listening point of view ^^ But non-audio data usually needs to be bit-accurate, so the drive has to adjust to both the factory and the drive's write offset.

And no, there's no overlap for discs with a negative offset. If the disc has a factory write offset of -647, for example, all data and audio tracks are shifted by those -647 sectors. But, as I mentioned before, the data tracks get adjusted properly, while the audio tracks don't, that's why we have to do that manually ^^

That said, it's possible that EAC loses samples at the beginning of the 1st audio track if the combined offset is negative. It's an annoying EAC quirk. You'll have to make sure that your drive's offset "outweighs" the negative factory offset. I have an LG with +667 which I use to dump -647 (a common PSX factory offset) or similar. If you're lucky, you have a Plextor. In this case you can use PlexTools, which doesn't seem to be affected by that problem

As for IsoBuster, it does know where the data track ends, but it appends the pregap of the 2nd track to it. The reason we need to cut that off is because redump handles the gaps differently. IsoBuster appends a track's gap to the end of the previous track (so the Track 2 pregap is put at the end of Track 1), while redump prepends the gap to the track it actually belongs to (pregap, as per cue specification ^^).

Appending the gap to the previous track is kind of a hackish solution to allow people to listen to their audio tracks without having the (mostly silent) pregap at the beginning.

I hope that answers all your questions

[Psytopia]

Thanks! Now it all makes sense! It's always better to understand what you're doing and why. I ripped many audio discs with EAC but those issues with mixed mode CDs never occured to me.

That said, it's possible that EAC loses samples at the beginning of the 1st audio track if the combined offset is negative. It's an annoying EAC quirk. You'll have to make sure that your drive's offset "outweighs" the negative factory offset. I have an LG with +667 which I use to dump -647 (a common PSX factory offset) or similar. If you're lucky, you have a Plextor. In this case you can use PlexTools, which doesn't seem to be affected by that problem

I have Pioneer DVD drive with a combined offset of +48, but I don't intend to dump any PSX titles. Also, shouldn't reading into the Lead-In solve the problem?

As for IsoBuster, it does know where the data track ends, but it appends the pregap of the 2nd track to it. The reason we need to cut that off is because redump handles the gaps differently. IsoBuster appends a track's gap to the end of the previous track (so the Track 2 pregap is put at the end of Track 1), while redump prepends the gap to the track it actually belongs to (pregap, as per cue specification ^^).

Appending the gap to the previous track is kind of a hackish solution to allow people to listen to their audio tracks without having the (mostly silent) pregap at the beginning.

Appending the pregaps to the next track is also needed to get ISO compliant Cue Sheets and according binaries, makes sense!

Thanks for all your answers! They helped a lot!


Would be cool if the developer of EAC decided to include automatic correction of factory write offsets using the data track to allign the sectors. Then mixed mode rips could be even "more lossless" than audio only rips.

[amarok]

I have Pioneer DVD drive with a combined offset of +48, but I don't intend to dump any PSX titles. Also, shouldn't reading into the Lead-In solve the problem?

Sorry, no idea ^^ I've never tried anything else than my +667 drive for these discs. I just remember that years ago, many dumpers had drives with large positive offsets so they could dump discs with large negative factory offsets.

The bigger problem I think is that you won't be able to detect a negative offset larger than your drive's positive offset with IsoBuster ^^ If you've read the guide you maybe know that you have to go back 150 sectors (in case of a 2 sec pregap) and count how many rows of junk data you see (because the end of the junk data marks the beginning of track 2). However, if the disc's negative factory offset is too large, you won't see said junk but only 00s, so there's no way to determine the start of track 2, respectively the combined read offset.

In any case your +48 should be enough for most PC discs. Common PC offsets are -22, -12 and 0. Tendentially, the older the disc, the more obscure the factory offset

[Psytopia]

The bigger problem I think is that you won't be able to detect a negative offset larger than your drive's positive offset with IsoBuster ^^ If you've read the guide you maybe know that you have to go back 150 sectors (in case of a 2 sec pregap) and count how many rows of junk data you see (because the end of the junk data marks the beginning of track 2). However, if the disc's negative factory offset is too large, you won't see said junk but only 00s, so there's no way to determine the start of track 2, respectively the combined read offset.

As far as I understand, you might have to go further back to another sector if the offset is too large. And of course I've read the guide.

Thanks again! I'll post a proper dump of Seven Kingdoms II (german) by tomorrow!

[amarok]

Well afaik that's only true for discs with a large positive combined offset (larger than 1 sector = 588 samples). But again, I've never tried to dump discs with a negative combined offset