I planned on doing some runs of Spyro 2 with a MiSTer FPGA. It's a hardware emulator that recently had PSX ported to it, which runs on an FPGA board. It may not run at 100% speed as far as I'm aware of, maybe a tad bit slower. I timed most of the loading screens of Spyro 2 on MiSTer vs PS2, and it seems to be slower than PS2 by about 1-4 seconds depending on the zone.

I figured I'd bring it to someone's attention to see if it can be considered legitimate, since its "technically" a PS1, but it's not using software emulation either. Kind of a weird gray area. I'm sure you probably think people are crazy for wanting to run on a slower console, but I figured I'd check.

Thanks.

Hey! this is a late reply, and I'm not part of the mod team or anything, but I wanted to reach out because this did get talked about a bit in the discord server a few months back when you posted this! ^^

Wanted to say personally that I'm a fan of the MiSTer project, and I really hope that its hardware emulation/replication will eventually reach effectively perfect accuracy in the future. As for now, from the discussions had, it seems as though it wont be allowed for leaderboard submissions, however I fully encourage you to run on it because having more documented data and recordings of gameplay on a MiSTer to compare with original hardware and software emulation for accuracy is huge. If you're saying it is slightly slower than PS2, and you haven't noticed any major bugs/glitches that would significantly affect gameplay, then in my opinion it should be considered and allowed for leaderboard submitted runs similarly to accepted emulators.

There should definitely be enough research done though to ensure there aren't any hardware specific differences. You mentioned that it might not run the games at 100% speed, though if you're able to get clean recordings (especially with an input viewer), then we can compare with other runners as well to check for its accuracy. Of course many folks in the community who are more knowledgeable about things like this can help as well, and I'd be willing to compare with my emulator runs on Duckstation at the very least for various levels! I'm not sure how useful it would be as well, but if you can get some video feed with an input viewer working as well for various levels (particularly with Spyro 1 as its a bit more deterministic overall) then we could even compare the results to a TAS recreation in BizHawk, which would help with checking for any potential lag discrepancies throughout the game.

Sorry if I'm asking for a lot, or if this sounds like a lot of work <>;; I know I also dont have the time right now to dedicate to testing things like this at length, however like I said, I'm all for supporting and accepting accurate hardware emulation/replication for speedruns. After all, it is effectively as accurate as you can get to original hardware at its theoretical best!

If you want to talk more about all this, either with me or even with others who are a bit more active with things like this in the community, definitely join the speedrun discord, and feel free to DM me!