The following are some considerations and discoveries after dumping more than 200 Saturn discs:
1) Write Offset.
It seems that ring codes have something to do with write offset in the following way:
- when there's a ring code ending with A, B or C the w.offset is always +18
- when there's a ring code ending with C xx the w.offset normally is +0 (By the way in this case we may have many alternatives audio tracks as it seems there was more than one master copy). Some ring codes are missing, by the way in this case w.offset can be also +222 or else (like http://redump.org/disc/1431/ and http://redump.org/disc/2767/).
- when there's a ring code ending with V we can have different values for w.offset, but the most common are +390 and +684.
So the most common w.offset for Saturn discs seems to be +0, +18, +390 and +684
----------------------------------
2) Problems and anomalies in dumping process.
I've dumped all my games both with PerfectRip and EAC (and IsoBuster where's needed) to check and avoid bad tracks dumping.
PerfectRip sometimes gives bad results, the most common are the following:
----------------------------------
A) Data track with errors in log and after checking with CDMage.
Solution 1: Redump data track alone without audio tracks
Solution 2: Slow down reading speed
Solution 3: a- Take bad dump, dump the same data track with Isobuster (subtracting track02 pregap)
b - check bad track with CDMage and import sectors from data track dumped with Isobuster (this is only if someone wants to check it, the data track dumped with IB normally is good).
----------------------------------
B) First audio track (normally "track02") may have garbage in pregap (you can find this if you dump tracks also with EAC as PR doesn't report any dump or C2 error).
If you open the track with an Hex editor you'll find 2 or more -12 bytes- code strings = | 26 48 26 48 26 48 26 48 26 48 26 48 | in the first sectors.
Solution: Simply redump track with EAC.
This problem was found in the following games:
- Bubble Bobble also featuring Rainbow Islands (E) - http://redump.org/disc/3907/
- DecAthlete (J) - http://redump.org/disc/3252/
- Magical Night Dreams - Cotton 2 (J) - http://redump.org/disc/3889/
----------------------------------
C) Wrong "track02" pregap, bigger pregap and so smaller data track and sectors of data in "track02" pregap.
Solution: Forget PR and dump the old way IsoBuster+EAC.
This problem was found in the following games:
- Puyo Puyo Sun (J) - http://redump.org/disc/3766/
- Waku Waku Puyo Puyo Dungeon (J) - http://redump.org/disc/3751/
Moreover the following games have wrong pregap detection both on PR and EAC with a Plextor:
- Panzer Dragoon (J) - http://redump.org/disc/2636/ and http://redump.org/disc/3243/
- SteamGear Mash (J) - http://redump.org/disc/3890/
[s]PS: there are other cases where point A, B and C are mixed.[/s]
----------------------------------
D) No pregap for some tracks in PR, but pregap with EAC.
- PR adds pregap to previous track.
- EAC move one sector to previous track if it finds mode=0x02 sectors (look down for Dremora explanation).
This is the most strange problem found with PR/EAC
Solution: Redump track with 0 pregap with EAC, check subcode (dump CD with CloneCD or Alcohol and open .SUB) with sca for 0x02 mode sectors. If first sector of one of the problematic tracks is 0x02, move this sector to previous track.
Here is the list of dumped games, and tracks pairs, with this problem (there could be many others).
Strikers 1945 (J) (Track 1-2)
http://redump.org/disc/4898/
SUB ---> T-14407G
Analysis: Track02 first sector is ok, last sector of data track is 0x02 so here this sector is put at the end of track by EAC.
*** I saw P_star dumped a pair of EU-US games with same strange gaps, with exception of anomalies above, both EAC and PR give the same serpentine pregaps structure.
----------------------------------
3) Games with INDEX > 1
4) Other games with serpentine pregaps
The games below have serpentine pregap but don't have the problem of point 2.
King of Fighters '95, The (J) (Track 04-05, 11-12, 14-15, 17-18, 20-21, 24-25)
http://redump.org/disc/1874/
SUB ---> T-3101G
Analysis: Actual pauses are exactly 150 sectors, but according to the Q-channel, those pregaps are 149 sectors long.
5) Other anomalies
The games below have only one track with different pregap from 2.00 or 1.74.
Kyutenkai (J)
http://redump.org/disc/3716/
SUB ---> T-1801G
Analysis: Track pregap is really 2.01, please beware that EAC reports 2 sectors more pregap for every track (2.00 ---> 2.02 and 2.01 ---> 2.03).
Whizz (J)
http://redump.org/disc/5164/
SUB ---> T-36102G
Analysis: Seems the opposite of Gunbird, sync errors that put real pregap to 2.02. Subcode reports 2.00, but if you take this pregap the last 2 sectors of data will be void (fffff...) and px_d8 reports sync error while checking w.o. on those sectors. To get a w.o. you have to go back to 2.02. Moreover IsoBuster reports errors while track extraction.