Experiment: phantasy star collection

Phantasy Star Collection's disc uses the regular ROM images of the Japanese editions of Phantasy Star 2 and 3. 1 and 4 were remade however. I'm going to rebuild the disc with the English ROMs of 2 and 3 in place of the Japanese ones and see if it runs or crashes. If it crashes I'll try editing the header to Japanese reigona nd fixing the checksum. If that fails. Screw it
I'm pretty sure that Arakon already tried this and that they all crashed. I think it was determined that none of the ROM images are actually identical to any known cart version. Also, I'm pretty sure that PS1 is an actual ROM image, just not padded to a standard size. I'd have to check though.
Disc got done burning and yep. Everything crashes. I'm going to look into the ROMs a bt more. Maybe try hacking them on the PC then putting them back onto the disc to see if they run rather than outright replacing them.

Edit after Arakon's post:

Is it a checksum issue or data just being in the wrong place? Also is it looking for a checsum that matches the expected ROM or just a valis checksum? I suppose you could always monkey with the US version until you get it to match the checksum of the version on the disc.
ExCyber: The roms are indeed identical to known cart versions. After running them through GoodGen to see what versions they were detected as, it came out with

Phantasy Star 2 (J) [p1].bin

Phantasy Star 3 - Generations of Doom (J).bin

I assume PS1 wouler perform similar in GoodSMS. So I guess it mus tbe a checksum thing like Arakon was saying. And since the US games would have good checksums, it must be looking for a specific checksum and not just a valid checksum. I'll look into hacking the US roms to match the Jap ROMs checksum and take another shot at it. If that doesn't work, ah well
It was fun for a few minutes and I got actual use out of CDMAGE
"p" generally means that the dump was from a pirate cart. Wouldn't that suggest that the dump was from a pirate cart that took the ROM image from the PSC CD? Also, now that you mention it, I do remember PS3 being the same. PS4 wouldn't be for obvious reasons, and ISTR someone saying that the PS1 ROM was modified somehow (possibly with magic illegal instructions).
Yeah, the pirate cart bit makes sense. I know something is different int he PS1 ROM since it doesn't work in emulators but I'm not sure what. I need to track down the Japanese rom and run a bit/bit compare on them. It doesn't look THAT different though but something was definitly changed since it won't work on a Master System or on a Master System emulator.
What emu are you using? I'm 99% sure that I ran it in Dega at some point, though not past the intro I think.

Anyway, by "magic illegal opcodes", what I meant is that they may have emulated certain things in a game-specific way with native Saturn code, and inserted opcodes into the ROM that aren't defined on a real Z80 (but probably have some oddball poorly-documented behavior on the real thing) to use as triggers in their interpreter/compiler. This seems to be the most likely explanation to me, but it's certainly possible that it's something else.
ok, now that I'm a bit more awake..

it appears that the games are not simply emulated. music is stored seperately, and the rom is possibly only used for parts of it, i.e. graphics and text, but not the game code (which is a similar way to what they did with the FF and chrono trigger etc SNES->PSX RPGs). which would also mean, since the data is not in the same places in the roms in the jp. and US versions, that simply replacing the roms, even after defeating any checks, would still cause garbage or crashes.
What if the PS# executables in CDROM:/ handle each game as far as sending it to the 68000 for code processing? Then it could be run without emulation. And since there's no YM2612 in the Saturn anyway having audio samples makes sense.
What if the PS# executables in CDROM:/ handle each game as far as sending it to the 68000 for code processing?

It's possible that they do this for some routines, but you can't just throw unmodified code at the 68000 because it doesn't have an MMU that you can use to provide a compatible memory map and e.g. trap register writes.
And that doesn't get into the problem that the 68000 is the saturn's sound processor and th egame music is all stored in saturn audio format. If the 68000 were doing program execution what would be playing the sound. I just find it really weird to believe the main code binary for Phantasy Star 2 is only 180KB.
Isn't the PSC PS1 actually the Genesis version? This way they would have less problems in running it, by "emulating" only the genesis, and not the SMS.
there is no genesis version, technically.

all the genesis cart is is a SMS rom in a cart wired up as a SMS adapter.

a dump of it would give you a 100% SMS rom, NOT a genesis rom that any genny emu could run.
Yeah, it just passes stuff through.

I don't know if you own a SFAM dumper but try dumping a Super Gameboy cart through the super gameboy and all you'll come out with is the SGB BIOS and no game. Or when you try to dump a NES cart through a Super-8 plugged into an SFAM copier. You don't get anything.
Just a related question. Just to check if we're looking for a real checksum problem, could anyone change a byte in the ROM and see if they still run? I'm pretty sure it this was done before, but we never know... ^o^;