CPS A-Board Differences

    This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

    • Apocalypse wrote:

      rtw wrote:

      Here are the two CPU dumps from daimakai and daimakair:


      GAME( 1988, daimakai, ghouls, cps1_10MHz, "Capcom", "Daimakaimura (Japan)" // Wed.26.10.1988 in the ROMs
      GAME( 1988, daimakair, ghouls, cps1_12MHz, "Capcom", "Daimakaimura (Japan Resale Ver.)" // Wed.26.10.1988 in the ROMs // 12MHz verified

      They were created by modifying MAME to dump the main CPU memory region after the ROMs are loaded in the same fashion as with the CPS2 and F3 generator I wrote.

      The files are clearly different.

      Note:
      The fill value is 0xff on daimakai and 0x00 on daimakair but even if we ignore this the files are different. For example at offset: 0x6

      Source Code

      1. daimakai: 0x2e 0x19
      2. daimakair: 0x96 0x1a
      daimakai.zip
      I know that, the ff instead of 00 doesn't affect code neither the other differences you found which are used to trace the version.As most game boards and consoles CPS1 has a certain amount of CPU time between 2 vblanks. This period is fixed and doesn't change with the CPU speed. If all code can't be executed in this period a frame drop will occur resulting in a slowdown for the user. Code isn't modified for 12Mhz CPU, if there weren't any slowdowns when playing the game on a 10Mhz mainboard then there won't be any difference on a 12 Mhz motherboard. If there were slowdows then they may disappeared or be less frequent on a 12Mhz motherboard.
      Just looking for some clarifications:
      -I have a 89626A-3 A-board (short - 10Mhz crystal) - 91635B-2 B-board - CPS-B-21 C-board

      -It's been converted with the daimakair set (Daimakaimura - Resale ver.) that was supposed to be on a 12 Mhz board

      I understand the principle of sync to display refresh, so even though daimakai and daimakair differ, can we maintain that both roms play exactly the same, independent of which A-board they are on? Or simpler: are there any slowdowns in daimakai that are not in daimakair because of the 12Mhz clocked CPU?
    • geekmiki wrote:

      can we maintain that both roms play exactly the same, independent of which A-board they are on?
      I personally never felt like I did enough side-by-side testing to make that statement.
      Initially it appears to be the same, for me the concern is how would they compare after many hours of gameplay/operation?

      So in the absence of a proper answer, I simply paired all my B-boards with what would have been the original A-board speed (10 or 12 no mix and match).
      Darksoft: CPS3, ST-V, CPS2, F3, MVS
      RGB: RECO v2, HAS v3


      [genesis] [segacd] [saturn] [dreamcast_ss] [nes-101] [snes] [n64] [gamecube] [pce] [pcecd] [psx]
    • sammargh wrote:

      So here's something interesting. I bought a SF2 "CE" off eBay that someone has had in their closet for years, even still had a static bag and looked entirely stock.



      ..Except it's actually SF2 HF and it came with a 10MHz A board

      i.imgur.com/wozSnrP.jpg

      (linked because huge image)
      I wonder if we got it from the same guy. I bought a HF board off eBay and it came with a 10mhz shortboard. I didn’t even know there was different boards at the time. The game worked, but something always seemed off. I ended up getting rid of the B and C boards as I played Super Turbo and Alpha 2 way more on my CPS2. I kept the A board to use with Final Fight.

      When I later found out about there being different boards, I was convinced I had a 12 seeing how it came with HF. Thanks to a few people here I found out that I have 10mhz short. I’ll post some pictures tomorrow night just to double check which one I have.

      What bugs me is that the background in the auction pictures (an electronics workbench for crying out loud) leads me to believe the seller was not ignorant of the mismatch...
    • mukurofish wrote:

      sammargh wrote:

      So here's something interesting. I bought a SF2 "CE" off eBay that someone has had in their closet for years, even still had a static bag and looked entirely stock.



      ..Except it's actually SF2 HF and it came with a 10MHz A board

      i.imgur.com/wozSnrP.jpg

      (linked because huge image)
      I wonder if we got it from the same guy. I bought a HF board off eBay and it came with a 10mhz shortboard. I didn’t even know there was different boards at the time. The game worked, but something always seemed off. I ended up getting rid of the B and C boards as I played Super Turbo and Alpha 2 way more on my CPS2. I kept the A board to use with Final Fight.
      When I later found out about there being different boards, I was convinced I had a 12 seeing how it came with HF. Thanks to a few people here I found out that I have 10mhz short. I’ll post some pictures tomorrow night just to double check which one I have.

      What bugs me is that the background in the auction pictures (an electronics workbench for crying out loud) leads me to believe the seller was not ignorant of the mismatch...
      I've bought a few sets now as I intend to convert them to other boards. It seems 10 MHz swaps are highly common on the SF2CE/HF boards likely due to operators at the time not knowing the difference. Could've been accidental due to the original A boards dying or just board swapping in cabs landed that way. Of the 3 I bought 1 legit was 10 MHz, 1 "boot" was 10 MHz and another legit was 12 MHz. The two legit sets were from non-major distributors and were sold untested because the person owning them didn't have any arcade hardware at all to confirm them.

      That said I'm fine with this setup as all the games I want to convert to are 10 MHz originally so this is all fine to me.
    • I've done what I think is an interesting test: I've ran two SF2' (same region, same version) side by side, one with its OG 12MHz motherboard, the second one with a 10Mhz motherboard.

      Game speed is exactly the same during attract (including demo fights), there are only small desynchronisations during "loading" (for instance between the disclaimer and the intro then between the intro and the demo fight) with an advantage to the 12Mhz board (logical, access times are shorter).
      However in game, game is 20% faster on the 12Mhz board!
      No this raises the question: how is that possible that fight speed is identical during attract but different in game?
      It's worth having a look at the code in details.
      Selling files for conversions/desuicide on Sega System C2/Sega System 24/Sega System 32, Capcom CPS1, Irem M92/M107 and more (just ask)
      Money is reinvested in boards, parts and tools in order to release more stuff.
      Contact me through my blog: arcadefixer.blogspot.co.nz/
    • Apocalypse wrote:

      I've done what I think is an interesting test: I've ran two SF2' (same region, same version) side by side, one with its OG 12MHz motherboard, the second one with a 10Mhz motherboard.

      Game speed is exactly the same during attract (including demo fights), there are only small desynchronisations during "loading" (for instance between the disclaimer and the intro then between the intro and the demo fight) with an advantage to the 12Mhz board (logical, access times are shorter).
      However in game, game is 20% faster on the 12Mhz board!
      No this raises the question: how is that possible that fight speed is identical during attract but different in game?
      It's worth having a look at the code in details.

      This is really interesting as there's a 6 year unresolved old bug in MAME regarding SF2 and its correct speed.

      mametesters.org/view.php?id=00408

      Please keep us updated on the matter.
    • @rtw
      I've also done the following tests:
      - OG Ghouls'n Ghosts (10Mhz) compared to Daimakaimura resale (12MHz): no difference in speed
      - Daimakaimura resale on 12Mhz motherboard compared to Daimakaimura resale on 10Mhz motherboard: no difference in speed
      There were only slight desynchronisations during "loadings" due to faster access times on the 12Mhz motherboard.
      Selling files for conversions/desuicide on Sega System C2/Sega System 24/Sega System 32, Capcom CPS1, Irem M92/M107 and more (just ask)
      Money is reinvested in boards, parts and tools in order to release more stuff.
      Contact me through my blog: arcadefixer.blogspot.co.nz/
    • The 'hyper fighting' speed (vs championship edition) is due to frame skipping. CE also was supposed to be played on a 12MHz board. World Warrior was 10Mhz, and there's not much difference (if any) speed-wise between WW and CE, afaik. Other than the supposed 'loading' speedups, and presumably when the game slows down?

      I also seem to recall doing tests of HF 10 vs 12 through attract mode and being surprised at how similar the speeds were...

      It's possible that the 10Mhz boards can't deal with the extra computation necessary during the skipped frames.
      There is no knowledge that is not power

      Wanted: Mortal Kombat 2, NBA Jam TE
    • fluxcore wrote:

      The 'hyper fighting' speed (vs championship edition) is due to frame skipping. CE also was supposed to be played on a 12MHz board. World Warrior was 10Mhz, and there's not much difference (if any) speed-wise between WW and CE, afaik. Other than the supposed 'loading' speedups, and presumably when the game slows down?

      I also seem to recall doing tests of HF 10 vs 12 through attract mode and being surprised at how similar the speeds were...

      It's possible that the 10Mhz boards can't deal with the extra computation necessary during the skipped frames.
      Hyper Fighting Attract mode and the credits roll demo after completing the game run slower than the gameplay.
    • @Apocalypse, just to make sure, you're talking about SF2', aka SF2 dash, aka the later and revised Street Fighter II Champion Edition that Japan had. Right?

      There are so many versions of SF2 floating around that it's really hard to keep track of them.

      Can we just make a naming convention to avoid ambiguity, something like:

      SF2 = Street Fighter II World Warrior, first release, afaik it's the same in all regions. Pretty sure it was supposed to work on 10Mhz boards only.

      SF2CE = Street Fighter II Champion Edition, second release with the grand masters playable and several balance changes. This one was released in US and Europe (maybe part of asian market as well?). Bear in mind that while we were playing CE, in Japan were stuck with World Warrior. I *think* that this was on 10Mhz as well.

      SF2' = Street Fighter II Dash, third japan only release which had some changes to CE (some balance stuff). No clue about 10Mhz or 12Mhz, but thanks to Apocalypse, as I understand it, it should be played on a 10Mhz board as a 12Mhz one would make the game too fast (remember that the speed should be the same across SF2, SF2CE and SF2' ).

      SF2HF and SF2T = Street Fighter II Hyper Fighting and Street Fighter II Turbo: Hyper Fighting should be 100% the same (first one for US and EU, the other one for Japan). We know for sure those ran on a 12Mhz board.

      There's another thing to check: there's a dipswitch which is labeled as "unused" in MAME which *seems* to make SF2 run unthrottled. This should not make any change on SF2 -> SF2' running on a 10Mhz board but in the link I provided earlier, a user said that it made the game run faster on his 12Mhz board.

      Yeah, that's one big can of worms, hopefully we'll manage to find some answers.

      As of now, the questions which need to be answered are:

      What happens when SF2 runs on a 12Mhz board?
      Same with SF2CE and SF2', by checking the unused dip switch as well.
    • Don't worry, unless you're a hardcore fighting game player you wouldn't have noticed those :D

      There are also other differences mentioned in that video's comments, it looks like SF2CE played a little bit faster than SF2 Dash.

      It would be great if someone could make in-game comparisons (not attract mode). Does anyone have Aje_fr's multi here?
    • I don't know anything about special Japan only versions of CE or ' or whatever.
      All I know is CE when it was released over here (USA) it came stock with the 12mhz A-board.

      The story I had been told about this... Original World Warrior had glitches and slowdown, we all know this/you can see them.
      Capcom fixed the game play glitches in later revisions of WW but the slowdown always remained, affecting some gameplay elements.
      When making Champion they up'd the clock speed to finally remove/address the slowdown issues.
      Darksoft: CPS3, ST-V, CPS2, F3, MVS
      RGB: RECO v2, HAS v3


      [genesis] [segacd] [saturn] [dreamcast_ss] [nes-101] [snes] [n64] [gamecube] [pce] [pcecd] [psx]
    • donluca wrote:

      Nope, they are different. Dash came after CE and they took the chance to balance some stuff around. 100% sure.

      EDIT: proof
      i think you are getting mixed up the game revisions versus the game versions.

      Capcom liked to produce multiple revisions of game code, but differentiates between the actual upgrades. E.g.

      1.0 = World Warrior
      2.0 = Champion Edition
      3.0 = Hyper Fighting
      4.0 = Super SF2
      5.0 = Super SF2 Turbo
      etc.
      etc.

      So different revisions of Champion Edition (which is known as SF2 Dash in Japan due to the apostrophe on the title screen that is ignored in the USA) may indeed have different gameplay aspects, animations, etc. However, it won't have major changes like you would see between major version differences.

      Also - that video isn't able to provide any proof since they never showed the boot-up screens to denote they are using the same code revision but only different regions, nor did they show that this was run on actual arcade PCBs. This could be a video of two different ROM Revisions from various regions being run in an emulator. That wouldn't prove anything really.