MMC5 exact timings

Discuss emulation of the Nintendo Entertainment System and Famicom.

Moderator: Moderators

User avatar
rainwarrior
Posts: 8734
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: MMC5 exact timings

Post by rainwarrior »

The Paperboy problem I think is only known to happen in other Mindscape games. I only know of these 6 games which have the same problem:
  • Captain Planet
  • Dirty Harry
  • Infiltrator
  • Mad Max
  • Paperboy
  • The Last Starfighter
Bad Dudes looks like a more normal controller read at first glance in a debugger. In general I would not expect it from a Japanese developed game because expansion port controllers were common, and reading both D0 and D1 in a way that also looks at the open bus area seems unlikely to me. I have found many non-Japanese releases that only check D0, but haven't yet found a Japanese one that fails to check both D0 and D1.

I'm wondering if instead you've discovered that the first controller read after reset (or certain situations?) might be unreliable (or reliably corrupted?) on some carts? Is 1 extra poll consistent for the games where it is needed, or do they only have a chance of working?

Low G Man does have a known bug that reads somewhere in the open-bus WRAM region at times (during boss music, and I think when using the boomerang weapon, not sure when else). In the bug I know about, as long as the read value is above a certain value it should have a consistent result (just checks if under a threshold and otherwise returns, I think)... though it's possible there is something else to do with this that might affect a controller read. The extent of my investigation was just to get it to not crash or visibly mess up, I wasn't as rigorous as a TAS verification. (I had made a patch that filled WRAM with $FF at startup, which seemed as good as open bus for the purposes I tested.)
Alyosha_TAS
Posts: 173
Joined: Wed Jun 15, 2016 11:49 am

Re: MMC5 exact timings

Post by Alyosha_TAS »

Oh I didn't realize there was a list of such games, maybe I'll get one or two of those and see if I get similar results.

It's been a while, but yes I remember it being very consistent. Paperboy never worked without the extra poll, for example. I never got very far in figuring out what mechanism might be the cause. Nothing special seemed to be happening.
Post Reply