The game Alien Syndrome does not work properly in Nintendulator. The game uses the Tengen board 800037 with the RAMBO-1 mapper. After the round select, the gfx go bad and then the screen just goes blank. The only other RAMBO-1 games I have are Klax and Skull & Crossbones, and they both seem to work fine, only difference I've found is that those 2 use the seemingly identical 800032 board. One other odd thing is the image I had previously had the mapper set to 4 instead of 64 for some reason, but it displays the same behavior when set to 4 as well.
Next, Impossible Mission II (NINA-01 board), works fine, but Nintendulator throws a lot of warnings each time you change rooms. The warning says something about it trying to use both BNROM and NINA-01 functions.
On a sidenote, how the *hell* did BNROM and NINA-01 get assigned to the same mapper number? They are completly different... If anything, couldn't BNROM be assigned to mapper 7 instead?
The difference between B*ROM and A*ROM is that B*ROM uses horizontal or vertical mirroring while A*ROM uses mapper-controlled 1-screen mirroring. But couldn't B*ROM be merged with Color Dreams without much problem?
And are you sure you're using good dumps?
tepples wrote:
The difference between B*ROM and A*ROM is that B*ROM uses horizontal or vertical mirroring while A*ROM uses mapper-controlled 1-screen mirroring.
Yep I was aware of that, but seeing as a BNROM game would know nothing of the extra register in A*ROM, I would think it would work without a hitch.
As for using a good dump, pretty sure I am unless there is something different about the mapper on the 800037 board compared to the other RAMBO-1 board that would cause it to dump improperly. The contents of my dump match the existing one floating around.
Mapper 7 is single-screen only (no horizontal/vertical mirroring support whatsoever), so BNROM is not compatible with mapper 7. Also note that A*ROM has only one register - it's just that one more bit is significant (bit 4), and neither possible setting for that bit matches BNROM behavior.
Ah ok. I was aware that AxROM only had 1 register, I guess I was a bit ambiguous there. AxROM has the 1 extra bit (not register) for mirroring, which seems odd now that you mention that it uses 1-screen mirroring. Would setting that bit have any useful purpose then?
The LAST thing we want is to use the H/V mirroring flag to distinguish between two different mappers. It's already been done with the 4-screen VRAM flag and mapper 78, and it's really annoying to have to specifically redefine the 'global' flags for specific mappers.
Isn't that basically the case with BNROM and NINA-01 already? Why do they share? There is no common ground between them. They don't interfere with each other, but still it seems stupid to have grouped them together. This must 've just been an accident or something, where someone didn't know it was already asigned to something else.
Oh yeah, Alien Syndrome works in Nestopia, so I'm not sure what the deal is there.
Alien Syndrome uses a special mapper hookup which happens to be equivalent to iNES mapper #118 (TLSROM/TKSROM), NOT #64 (RAMBO-1) - the reason Nestopia supports your ROM image is probably because the mapper number is wrong and the internal database is correcting it for you. Assign it to mapper 118 and you'll find that it works fine in other emulators as well.
Actually, better yet - run the Japanese version (which just uses the MMC1), and you'll get an even better game (complete with cutscenes, even with English text)!
One possible reason why NINA-01 and BNROM were grouped together is that they were each used for only a single game, and their register mappings are mutually exclusive - BNROM uses only $8000-$FFFF (like all of Nintendo's other discrete boards), while NINA-01 uses only $7FFC-$7FFF (thanks to a 13-input NAND gate for complete address decoding).
Thanks for the info. So 800037 would always be mapper 118 then? If it's even used in any other games, that is.
No, just in Alien Syndrome.
Mapper 118 is technically a plain MMC3 with the high order CHR bank bit connected directly to VRAM A10, causing the upper bits of CHR banks at $0000-$0FFF to select nametables at corresponding locations in $2000-$2FFF.
If Alien Syndrome actually uses a RAMBO-1 (which is an "enhanced" MMC3), then it really deserves an additional mapper number; however, if it uses the "MIMIC-1" (which is actually a clone of the Namco 108/118 mapper, a simplified subset of the MMC3) then mapper 118 will work okay, though it should still have its own mapper number (since the Namco 108/118 has no IRQs, hardwired mirroring, no SRAM support, and lacks the extended PRG/CHR bank bits located in the upper 2 bits of $8000).
FCEU reassigned it to mapper 158 a while ago, so I did the same.
Yes, It does use the RAMBO-1 mapper, not the MIMIC-1. That's why I was under the impression it should be mapper 64.
I think it should be assigned a new mapper. Mapper 64 does support the CHR bank-bits nametable selection and Mapper 118 does not support the full RAMBO-1 memory addressing scheme.