BIZhawk and ePSXe are listed as the approved emulators but I'm wondering if there is any sort of information as to how different settings compare to real hardware (primarily fast disc/region variants) and what may be considered unfair or illegitimate.

I used ePSXe with the default recommended settings for my run because it seemed to perform better on my machine than BIZhawk (both of which performed better than my disc on PS1 but not my PS2 with fast disc speeds) but still if there's any information or resources I'd appreciate it!

P.S. Overclocking on ePSXe seems to mimic fast disc speeds up to the audio glitches but again I'm unsure as to the validity of a run that would implement those settings.

Thank you.

As far as I know, we're not yet sure of the accuracy of different emulators - Bizhawk has slower load times than ePSXe, but is often considered more accurate for PS1 emulation across other speedgames. The comparisons I've done of only the loading sections of the game had my PB with Bizhawk 2.3.0 losing exactly 2 minutes to Valkyrie_ZB's PS2 SCPH-75004 run with fast disc speed. With the load times normalized to his console, my PB would be a 57:44. Across the emulators we've used, the gameplay sections don't appear to have immediately noticeable differences, but there hasn't been any extensive testing so far. I can say that Bizhawk feels more responsive to me than ePSXe, but I can't quite pin down why -- I feel like I'm encountering input lag or inconsistent movement when I try to play in ePSXe.

In 2018, Valkyrie ran on an outdated version of ePSXe (2.0.0) that made his load times noticeably faster than our only other baseline at the time, J.Y's PS2 run, so we disallowed that version to ensure emulator wouldn't be faster than console. However, J.Y's run was normal disc speed on PS2, and once Valkyrie started running on PS2 FDS, his load times were much faster, and may have been comparable to 2.0.0, but I haven't gone back to test since then.

Here's the video where I compared the loading sections from each version that was being used at the time: Ideally, we could figure out some sort of load remover, as other games have been able to do using livesplit, but the long fade-to-black at the end of each level and lack of consistent visual cues makes it difficult to do that automatically. Each run would probably have to be manually timed at dozens of different points to normalize the gameplay sections across all versions, which is very time-consuming, even for a small and sporadically active community, so thus far we've just accepted many versions of the game with the understanding that the final time won't be perfectly comparable.

What method are you using to overclock ePSXe? I don't know much about that functionality or how else it could affect the game. I would say from a personal standpoint that I don't think any emulator run with faster loads than console should be accepted, but that would require further testing to figure out.

I tested it with the standard 1.5 times overclock setting on the latest version of ePSXe on their site. Like you said gameplay didn't seem too different other than the BGM playing faster and coming in choppy. If load times are the issue with the run though I'll be sure to check out the differences.

As for accuracy, BIZhawk on default settings for me seems to run notably slower than ePSXe but because I don't have another PC to run the emulator on I ran the game, the stream, and the timer all on the same machine so I don't know how much of that if any might have compromised my performance as well. Would you mind telling me if you emulated on an isolated machine when you ran your times? Or did you also run all of your software on a single machine?