I'm resuming here my analysis and current knowledge about the original releases of Tetra Quest.
There are two known releases/versions at this time.
One is the Microdeal US release. A sample of the box is here: http://www.atarimania.com/game-atari-st ... 10517.html
The other version is from the Tri-Star Warriors II compilation. This was dumped very recently by Kodak80. See here: http://www.atarimania.com/game-atari-st ... 23780.html
The Microdeal release is very common, but unfortunately it is not fully working. It has a mastering defect and all known copies have the higher tracks (73-79) unformatted. The high score data and graphics are supposed to be located at that area. The game boot and works fine, but it misbehaves at the end when the high score is supposed to shown. This release has no actual copy protection except for a custom track layout and some dependency on disk load timing.
The only dump we have so far of the Warriors II release doesn't seem to work. Initially we thought the disk was damaged or corrupt. But further analysis shows that the disk is definitely not damaged and the data seems to be correct. It doesn't boot because there is a timing issue that I'll elaborate below.
But what resulted very interesting is that this release is fully copy protected. It has a a very special type of weak sector on track 0 that also is causing problems to Pasti emulation. This is in addition and not related to the timing problem that affects the original disk itself on a real Atari. We also found that the Microdeal release has still the same copy protection check on the code. The Microdeal release was slightly altered to ignore the result of the protection check. In other words, the Microdeal release was actually deprotected by the publisher itself. This means that the Warriors II release was based on an earlier master, probably the first original master of this title!
Except for the protection both releases seem to be almost identical. It is possible to copy tracks 73-79 from the Warriors II release to the Microdeal release and this "patch and mix" of two different versions seems to be fully working.
About the timing issue of the Warriors II release: The game code is partially encrypted. It is decrypted after the first stage loader. The decryption key is obtained by measuring the loading time of the first stage. This is identical in both releases. Identical loading, identical decryption routine and identical decryption key. The Microdeal release has the correct load timing and then decrypts correctly. The Warriors II release has a slightly wrong load timing. Then the decription key obtained is wrong and the game crashes.
The difference in the loading time between both releases is produced by a slightly different sector alignment on a couple of tracks. For the technically oriented, the inter sector gap is different. As a consequence of this different alignment those tracks take an extra disk revolution to load. This couple of extra revolutions is what makes the loading time to be correct or wrong.
If I take the Warriors II dump and realign the sectors on those tracks to match the alignment on the Microdeal release, then the decryption is correct and the dump is fixed without changing not even a single bit.
It is not clear why the alignment of those tracks is "wrong". It might be again, a mastering or a duplicating problem. It is possible that this depends on the RPM speed of the drives. A disk recorded at a different RPM speed would move the sectors to a slightly different angular position. And this might be enough to "correct" the timing.
Either way it seems that they "fixed" the alignment on the Microdeal release. But then it is strange they used a known "bad" master at the compilation release that presumably was much later.
Continuing in the next message ...