On NES timing is a huge, huge deal I think :/ I'd hate to try going for high compatibility there. On the other hand, it doesn't have all of the horrible VDC timing variables PC-Engine does. Castlevania III could be mapper related in addition to timing things.
Mappers do make things ugly... probably not slow though. Just tricky for the developer getting it just right.... as far as mappers go fortunately most of them are now well understood. The biggest hangup for me again is audio accuracy. Fortunately the NES has such a simple audio generator, now that it's been so thoroughly reverse engineered there's not much excuse for bad NES audio anymore.
QUOTE
I haven't used it but from what I've heard (and a video I've seen) calling it impressive would be a gross understatement. I think it does show that a lot of PCE games aren't using the CPU even close to its full potential. Idle loop detection could probably work a lot of the time.
I would wonder what games such performance was exhibited in... the PC Engine games I own work the system's CPU quite hard, Air Zonk is the biggest culprit.
QUOTE
The nice thing about PC-Engine is that the graphics are actually very simple. With a well optimized renderer it should be much faster to render than Genesis or especially SNES. In fact, it should only be somewhat more intensive NES. Only one layer, color depth doesn't really matter. There's a lot more sprite real estate though.
I agree. The CPU is basically the NES' at a higher clockrate (it's also CMOS, IIRC, not NMOS)... and the PPU isn't that complicated to my knowledge. And no mappers. It's too bad the PC Engine, like other less popular consoles (I'm looking at you Sega saturn, Atari Lynx, and many more) were really left out in the cold emulation wise so the developers who take on such projects have a fresh road ahead of them, not like NES, MD, SNES, etc. where there are lots of people who went before and can provide info on how to implement things in an emulator for that platform effectively.
QUOTE
In the first release definitely not. If I can get around to it I might make an attempt to do it on the x86 build. Unfortunately I can't test that on a real GP2X because I only have NAND to work around in. In fact, I can't do much testing in general so I'll need some dedicated testers anyway.
I do hope you can get around to the CDROM games. Either way, I've got quite a bit of PCE hardware now, both a PC-Engine Duo and a PC-Engine CoreGrafx II, and a large variety of games. I'd be more than happy to compare the graphical and audio accuracy when you have test builds together..... and of course when we have these shiny new handhelds in our hands.
notaz said:
Stealth Bagel said:
But glaring flaws like the lack of horizontal interrupt-based palette switching on MD emulators do bother me.
You bother on things you don't bother making sure of.
Using the gp2x I don't have anymore, that would be a difficult thing to 'make sure of'.... I respect the work you've done, it's quite impressive from what I've seen, and I'm not trying to deny your work either.... I was making generalized observations on accuracy in general and that was an example off the top of my head. You don't need to get rude about it.