nesdev.org

In reply to this post:

I don't think you can call CV3 as a hack. A hack would be to tricky replace all mapper writes by JSR commands so that another routine is called. I think Konami just ported their game to MMC5 mapper for some reason. (Nintendo seems to not allow Konami to use their VRC outise of Japan ? For some reason Sunsoft was able to use their FME-7 for Batman - Return of the Joker).
They would have to change the music anyway, or include adaptaters to input on the bottom NES connector, wich I don't think Nintendo allowed either.
A lot of other games were ported from older VRCs to UNROM, and some (like Contra) were degraded. However, Konami could use Contra as-is with the MMC3 and very few mapper hack (I did that), instead of downgrading the game and compress graphics to hell to fit a smaller UNROM cart. Or maybe they did just the UNROM version first, and just added cutscenes in the japanese version right before the release ?

Bregalad: It comes down to timing, so think about the history of these NES games.

In 1990, when CVIII was released, Nintendo had a policy of manufacturing games only by themselves, using their own boards. In 1991, when Batman ROJ was released, Nintendo had relaxed their manufacturing policy, letting US companies use their own mappers and boards. (I dunno if this means they let the companies make NES games themselves.)

When Contra was released in early 1988, was the MMC3 even available?? Does anybody know what the first MMC3 game was?

When Contra was released in early 1988, was the MMC3 even available?? Does anybody know what the first MMC3 game was?

I think the MMC3 was first used in SMB3 in October 1988. Contra being released in february 1988, the MMC3 wasn't available yet I think.
My Castlevania III has 1992 noted on it, but I think that's the european release date. Note that Konami could use VRC6 only in europe, but since nobody in japan cares about europe, they just scrammed the USA version and adaptated it to PAL timing (and took 2 years doing so).

Bregalad wrote:...Or maybe they did just the UNROM version first, and just added cutscenes in the japanese version right before the release ?

Correct. The US version of Contra was released before the Japanese version. At least that's what I heard, and it's the only explanation that makes sense to me.

Bregalad wrote:Note that Konami could use VRC6 only in europe, but since nobody in japan cares about europe, they just scrammed the USA version and adaptated it to PAL timing (and took 2 years doing so).

It would have been completely infeasible to use the VRC6 version in Europe, since it would have required either retranslating the Japanese version to English again (and subsequently re-removing all of the VRC6 sound, since the European NES still doesn't support expansion sound), or they would have had to take the MMC5 version and then hack it back to VRC6 and 'fake' all of the IRQ scanline counts (since the VRC6 counts scanlines as exactly 113+2/3 CPU cycles).

In the end, it was far easier for them to just take the USA version, change a few timing values in the game engine, make similar adjustments in the music engine, and make a few cosmetic changes (update the titlescreen, etc.).

In the end, it was far easier for them to just take the USA version, change a few timing values in the game engine, make similar adjustments in the music engine, and make a few cosmetic changes (update the titlescreen, etc.).

Sure, but remember that the MMC5 was probably atrociously expensive. Maybe they could sell carts less expensive if they were to use their own VRC6 by porting bankswitch code from japan, and just script, music and title screen code from the USA (and I prefer japanese difficulty by far).
Overall, it's not bad at all like this in MMC5, just a bit less fun because it would be cool to have one more rare board and released only in europe.

tepples wrote:I wonder where PocketNES's inefficiency lies.

Pocketnes doesn't yet speedhack this type of vblank wait:

WaitNMIEnd:
tay
lda NMIStatus
bne +
jmp WaitNMIEnd
+

Well it was about a week since I last used my Wii.. (at my parent house over xmas) but just before I went here I tried zelda and I thought some sfx sounded a little wrong.. but it might just be my memory.. it's really a long time since I last played zelda on a NES

Bregalad wrote: Sure, but remember that the MMC5 was probably atrociously expensive.

But Nintendo could've ordered a lot more MMC5s to be made than Konami would've for VRC6, even though MMC5 may have been more complex it could've been the same cost or cheaper because of the quantity. Also there's all kinds of factors when it comes to making chips, like yield % where some amount of the ones made are unusable. Could totally wreck the cost (or not).

Konami sure was hardcore when it comes to custom chips though.. I've noticed in MAME some of their arcade games ran on Konami CPUs..!

ccovell wrote:In 1990, when CVIII was released, Nintendo had a policy of manufacturing games only by themselves, using their own boards. In 1991, when Batman ROJ was released, Nintendo had relaxed their manufacturing policy, letting US companies use their own mappers and boards. (I dunno if this means they let the companies make NES games themselves.)?

Konami must have been allowed to use their own boards prior to this period even in 1988 already, many of their games have custom boards that were released early on. It can't be pinned on just later manufacturing runs either, because I've found the boards in carts with the "old-style" round seal, which they stopped using early 1989 and also some of them have earlier CIC versions like 6113 or 6113A.

Hey, can we have this thread back for discussion of the Wii virtual console's accuracy? I'm very interested in reports about it, but not interested in a digression about memory mapper availablity or Konami's custom chips. Thanks.