It's not needed. Your 'premium' drive also gives shit. So, I'm afraid your theory about PCE CD layouts is beaten 'Premium' subs (both normal and 'audio' ones) have either missing or doubled sectors in certain places, that affects pregap length. 'LH-18A1H' subs look reliable, I'll send you all the cues/logs/comments later.
Edit: or maybe the theory is not beaten completely yet...
yes, i'm aware of that. those .SUBs i've sent you were taken on Alcohol or CCD -
both those programs would return slightly 'improved' subchannel info.
AFAIR this could be witnessed most dramatically,
when an ordinary Audio Trap CD with multiple tracks is used to dump other CD
and there would be marks in returned subcode, matching to TOC entries from Trap CD.
i tried and could not reproduce this on current versions of CCD or Alcohol.
here is fragment from my older post:
Alcohol and CloneCD both masked some problems with subchannel, Alcohol seems to do it somewhat less but it has problems (track data actually gets damaged) with at least saturn data tracks when gap is large - 6 seconds (maybe 5) or more (mode changes on saturn, maybe it does not affect psx, i think i did not test, don't remember really). also when swapping toc with audio cd i think both Alcohol and CCD will output gaps in subcode from audio cd's toc. so it's strange, they like insert gaps in subcode instead of reading, i don't trust them at all after that. CDTool seems to be most accurate.
i used to report bugs like this, so maybe it was fixed with version updates, but i'm nut sure.
but it would be very difficult to convince me, that genereal
Audio->Data 3:00; Data->Audio 2:00; Audio->Audio 0:00
layout is incorrect
or for example, that each CD should have some random gaps instead of all audio - 0:00
on the other hand, i like very much how currently tracks on DC are separated:
all audio sectors in one file; data sectors in other
even though i think it's incorrect regarding INDEXes - i could not check
but i believe gap from Audio->Data it's still likely to be 3:00 seconds large difference, not 2:00.
well but i kind of regret now keeping 3:00 gaps for PCE - where data file would start with audio silence.
only application reproducing correctly CD from such cue sheet is CDRWIN anyway
and even though INDEXes are correct - probably Q-TNO aren't.
so 2:00 in cue sheet, reproduced on different application, might get incorrect Q-Index but correct Q-TNO.
it's as good as 3:00 but much better looking, IMHO.
so it's really more a philosophical thing.
to avoid confusion:
previously i've mentioned some PCE CDs having this gap set to 2:00, not 3:00 and that i think they should be 3:00.
well what i meant - i think they should not be different from the rest CDs, whichever pattern is chosen.
but i personally prefer Audio->Data being 2:00, so that audio and data sectors would end up separated one from another.
currently dumping by layout would be difficult and tedious unless such gap managment is implemented in imaging application.
so this is my understanding of problem.