if anyone's interested, I wrote some tools to practice the game, sharing in case it's useful for someone else. It's an ASL script for LiveSplit and the JPN version of Fatal Frame 3 ran on PCSX2. It has the following functionality:

1. Shows the IGT as "Game Time", the same one you see in the pause screen and results menu
2. Starts automatically when you select the difficulty and resets if you exit or load state into main meu
3. Has the option to autosplit by hour (eventually I can make it all-doors if there's interest)
4. Stores useful variables you can see without having to pause the game: points and amount of each film you have FOR TRAINING PURPOSES ONLY. Obviously this is an unfair advantage over console, but I've been finding it useful to route points (for example: I learned that you gain 1000 points on many pictures that are revealed in daytime even though that doesn't show up on the camera)

ASL file

Instructions on how to use it on LiveSplit

If you have any suggestions, potentially useful functionality, or you tried it and it's not working for you, I'm happy to hear any feedback! :)

So, yes and no I think.

This is basically trading fast and risky for slow and safe. Yes, you are way more likely to get an amazing eve, but at this point we know that it is possible to kill eve with 4 pepper sprays if you get a good damage roll. See here and here.

This is a very, very rough estimate but I think you can get away with killing eve with 4 pepper sprays about 1 every 4 times, which isn't that bad, but it is unfortunate that you roll the dice at the very end of the game. Yes you can get very nice PBs with the extra pouch, but in the long run, as your time gets better, you will have to let go of the extra pouch and start hoping for a good fight, kinda like starting to yolo Carlos on RE3 and letting go of always knowing the hospital password by risking a 30s time loss to save 6s to check what it is.

Also on where to get the pouch, you also have to wait for Pierce to stop moving to talk to him and proceed with the game, so this waiting time will happen no matter what. In Dryfield you have to take a side detour to check the trunk, so I think it's about the same where you get it.

Last thing to consider: If you want an extra pouch, there's one that maybe takes the same amount of time later in the game (~13s): Press the 1st, 2nd and 4th light at the garage where you get the jeep after the GOLEMs.

10x speedup is slightly faster. I can only speek for nightmare hundo and NG any%. For hundo 10x speedup is about 15 mins faster, and for any% about 2.5 to 3 minutes faster than ePSXe.

Sertyan started doing it trying to be sneaky , broke WR with it and then Sena got upset they beat his time and "ohh well I'll do it too" because I want my name on #1 situation,

THIS WHOLE ISSUE is literally between Sena and Setyan, 2 people only using speed ups to get ahead of the board for the "glory" of WR while everyone else is using standard preset values on settings.

Just for clarity's sake, can you explain again how upset and heartbroken I was when my time was taken, how badly I want the glory of the WR on the 25 year old game that hardly anyone plays, and how big of a cheater I am for speeding up the load times which is very, very obviously clearly cheating? I don't think everyone caught that yet.

Jfc it's like talking to a 5th grader.

Reminder to please not forget to vote 1 (keep 10x speed up) or 2 (use "None" for speedup). At the very least to keep a documented set of opinions, for whatever it's worth. Only wolf, gren and I stated where we stand so far.

@Amai_Fox those are great examples.

• SH1 and RE1: mentions clock speed control, read speedup and frame rate, not seek speedup. In my runs for example, I complied with all rules mentioned there. Note that these games are measured in IGT, so disk speedup would only make RTA faster and the games more enjoyable to run, provided they don't break emulation.

• Spyro: Ok fair, they said to disable seek speedup.

Important point for SH1 and spyro: The console and emu leaderboards are unified, so there needs to be extra attention to ensuring whatever timing method carries over between these. This does not apply to this game, where categories are separated.

Here's another example: MGS1 (image shown below).

Yes, seek speedup is a case-by-case situation, which requires testing to ensure the game doesn't break. And in fact, using infinite speed up I did have PE2 crash occasionally, but I've been playing at 10x for months, both NG Any% and Nightmare 100%, with no issues whatsoever. Apparently Setyan too, and if someone would come in and say they had crashes, I'd be right there with you and agree that no one should use it. This is why these things beget discussion and input, which is literally all we're hoping for and never get.

Ok here we goooo I can't believe I gotta say this again like I had for RE3 Modding but NOONE here mods or verifiers here are getting compensation of any form for modding this game or any other game

I know. that's literally what I said if you read above : "I have the utmost respect for people who take the time to verify runs because I know it's unpaid work that people do out of the love for the game".

like how dare I not check every forum for every game I mod like everyday huh? How dare I have a life and a family and work and such a terrible mod xD

No one's asking you to check the forums every day, we all understand it takes time to verify runs and make these decisions. But yeah taking 7 months to address the issue, when, like you said, there are only a few active players, all of whom also "have a life and a family and work" (and none of whom use this tired, offensive excuse) but are somehow able to take a few minutes to discuss this, yeah, that's terrible modding.

Amai informed me about this months ago we discussed it and it was an easy answer from me "just don't cheat"

Then, again, you guys (Amai specifically) shouldn't have accepted the runs, or at the very least explicitly stated it in the rules. As described above and will again be described below, this isn't as "common sense" as you're making it out to be, and if it is, don't accept runs and give players the wrong message.

The whole "Oh I cheated because he took my word record by cheating" is immensely silly , this whole issue started because a world record got taken on a 25 year old game that has literally 2-3 active runners , it is a GAME people it supposed to be for fun

I agree, but if we're competing, even for fun, there have to be ground rules. Playing a match without the arbiter to call foul kills the fun out of it, that's the problem.

I don't think noone here is getting money for been #1 on this speedrun site,

You're definitely disconsidering my above point that a lot of people gain followers and, in the long run, subs and donations through src. You're above this in your streaming career, basically all of us aren't, and yes that includes a 25 yo video game that lots of people check because they liked the game back then and want to check on the speedrun status. Trust me, I literally see this every day on my stream when people ask when I'm playing PE2 again.

secondly if you want those fast times JUST RUN EPSXE and get those sub 1h15ms again, that board is there it never left.

Or, like you said, we could use a great emulator and get the best of both worlds: Easy to configure, good graphics, and on top of that we can make it faster than ePSXe. How is this a bad thing if it's the same loading times for everyone? (34s to the door). You're tripling down on this being cheating. Is it cheating to play on PS2 instead of PS1 because load times are faster? No. What we do is we split categories. But this is somehow different just because you select something on a checkbox, no external tools or anything? And this is already available to everyone from the get go?

so THIS WHOLE ISSUE is literally between Sena and Setyan,

This whole issue is because you guys didn't address the problem when we brought it up- to us, not amongst yourselves -, and created this problem to begin with when accepting Setyan's run (You'd think at least the run claiming to be 1st place would be verified somewhat thoroughly). Then @Amai_Fox told me on discord, with no uncertain terms, that this should be treated the same way as "a PC game receiving new updates", and then you come in 7 months later and says the exact opposite, then rejects my run and my run only (yes I know Setyan's is rejected now but too little too late).

2 people only using speed ups to get ahead of the board for the "glory" of WR while everyone else is using standard preset values on settings.

I'm not even gonna comment on how offensive that is not just to me but literally everyone whose board you mod. The only person using the word "glory" here (ironically obviously) is you. That speaks volumes about what you think of us, but I will again point out that people either stream for a living or are trying to grow their stream, and you're making a mockery out of that.

My vote is towards #2 which 90% of runners are currently doing and default settings.

I vote 1. For anyone else who's willing to chime in, consider that

1. This basically makes the load times slightly faster than ePSXe

2. This makes 100% a much more welcoming category to newcommers (reduces about 30mins of game time), since right now we have 0 submissions

3. This adds up A LOT in the long run. If you need to reset early (e.g. No 9 gas bomb) you get there much earlier.

4. In ePSXe you have to do a plethora of configurations to run. This is a one time checkbox change that then is done forever, much like, say, in RE3, where you have to disable the CGI cutscenes prior to playing

5. This is the future-proof solution. If you keep regular seek speeds, new updates can roll out and still tamper with this, giving a disadvantage to all previous runs as it was the case 1 year ago. The loads here are essentially instant at 10x, and you can't really be much faster than 0s :)

Ok, but why?

Why do we have to use the all doors autosplitter? Since the game is RTA and frame rate is shown on screen, we technically wouldn't even need a timer. Any mod could simply subtract the eve killing time from the new game pressing time and retime a submission. Takes 15 seconds and some basic math. I literally had a run rejected because an old duckstation autosplitter started at the first cutscene and ended at the blank screen, when the easiest thing in the world would be to just retime. I've had runs that were retimed just to increase my time because mods (and by mods I mean @wolfdnc ) couldn't make up their minds about whether the time stopped at the last shot at eve or when screen went blank.

Some people don't like using all doors splitters to keep their golds more realistic. Some people don't even like using splits at all. Now these people can no longer compete? What's the point of this splitter? Have you considered all of these things? Have you asked the runners what they do or even checked literally every single run on ePSXe except for mine to realize most people don't play like that?

Why do we have to set the seek speedup to none? Did you know that over half the runs in the duckstation leaderboards have seek speedups faster than "none" because one of the duckstation updates imposed 2x speedup? Are you going to reject all of those too even though players were basically using default rules at the time? This logic of "to avoid advantage" makes absolutely no sense considering a fixed speedup, even 10x, would give the same loading time for everyone and would make 100% runs infinitely better to run. Have you guys considered that in your decision making? Have you reached out to the players? Or even taken 2 minutes to share a freaking google poll which is the easiest thing ever to do for any decent mod? Have you read the previous forum post where actual runners were discussing possibilities as soon as this speedup update rolled out? If this was a problem, why did you guys accept several runs with speedup and leave it on the board for months? How else are we supposed to interpret this besides "it is allowed"? You said you were verifying runs but literally the only one you rejected was mine, while the current 1st place on the landing page of this board has speed up, only beat my time because of speed up, but you guys haven't realized it yet because you think being at the 1:19:00 ballpark is because he didn't speed up, not because his run was bad. Because you don't understand the game or duckstation and are ill fit to make decisions.

Of course you guys don't know these things, because you don't give two shits about this board or what people are doing, except when you wake up and decide to do so. The way wolf manages this board is: He wakes up one day, pulls a bunch of rules out of his ass, dumps at us and then disappears. Dude doesn't even run the game any more, much less keeps track of duckstation updates or, by his own words, doesn't have time to verify runs. Why even be a mod anyways? All mods are essentially MIA and the one mod who is active in managing the rules is terrible at it and pretends players don't exist.

The way this board bas been manage is an exemplary case on how to make the wrong decision at every turn. Duckstation was decided to be the main category without anyone's consent. No consideration was given about the nuances of software updates, people were left to do whatever and retroactively punished because "common sense" dictates speed up is cheating?! How is it cheating if the category is literally called "emulator" and everyone has access to the same settings? No one is modifying the game, no one is increasing the game speed, no one is starting with 5000 exp and 10000bp to buy better weapons. None of this is common sense

This is personally frustrating and infuriating to me because I've been running this game for almost 5 years now, literally started out on twitch trying to do nightmare 100% and owe the vast majority of my viewership to PE2. Until a week ago I was literally the only active runner on NG any%, and all my efforts to draw attention to these ongoing changes to duckstation fell into deaf years. This mismanagement has bled out any motivation I had to grind and route this game, and it's been consistently bad at this point. I have the utmost respect for people who take the time to verify runs because I know it's unpaid work that people do out of the love for the game, but there's no love here. If you don't like or know what you're doing, there are several players who would be happy to take on the responsibility.

tl;dr I' done with this. Thanks @IfAndWhile @SirRetro91 @xGrenadier @ToxicTT @Marlindafish and all the homies from whom I've learned so much, but this is unbearable. @wolfdnc is a terrible mod and should step down.

Pog

That's huge if it can be done consistently! We would lose the free reload but it may be worth it since we can reload in the already-open inventory before using the cross-shaped key when fighting the four radiation dudes.

I think Symm's suggestion is ideal. There are only 2 runs that bumped the seek speedup: Me and Setyan's. I don't mind removing my run at all and I'm willing to guess Setyan wouldn't either, just fyi.

We'd only need to confirm if seek speedup 1x on current versions is the same speed as the previous ones, which I'm 99% sure it is based on playing it on 1x speedup with my previous splits from God knows how long ago.

And since there are no 100% runs on duckstation yet, I would just add the humble request of allowing this option on these runs since they make them at least 1h faster, for the sake of the runners sanity :P it's much nicer to play PE2 on duckstation, but I refrained from doing nightmare hundo runs on it just because it's so much slower, but that update gave me a glimpse of hope. Not too important anyway, just figured I'd share

This is tricky, because there's no way to retroactively re-time previous runs either, but as it stands I can clearly point out very good runs that easily top 3 in execution/luck but are all the way down on the leaderboards because they were done a long time ago with older duckstation builds (e.g. unionse's)

Duckstation is actively being updated, and whatever becomes the rules now may not apply for the future versions of the emu unless we can find a way to time it using LRT. That said I know it may be a ton of work through cheat engine to find a consistent "loading state", and just like TheDementedSalad's autosplitter breaks on more recent duckstation updates, scriptable timers will also have to keep up with updates breaking the algorithm in the autosplitter. It would be a lot of work for a game without lots of runners, but I think it's safe to say that people that do run it care about the game deeply.

Conversely, ePSXe seems to be a legacy emu now, without lots of changes going on, and despite looking like crap to run at full speed, it has had consistent loading times, to a point that my run and Sigma's were done 7 years apart and the loads seemed identical (I checked them side by side).

I don't really have a solution, just throwing my 2 cents.

Hey hey everyone!

Thank you so much for developing the awesome autosplitter that removes load times and makes competition fair for all machines!

You may have noticed that today the update rolled out that gives Caroline a new costume (yay!) I'm not sure but yesterday the autosplitter was working fine and today it won't recognize when the game starts. Is it possible that the update broke the splitter? If it's only on my end and it's working fine for everyone else please disconsider this message.

Thank you!!

Watching Amai Fox's stream (Aug 11 2021), we've been noticing since the latest update that there seems to be a consistent decrease in load times that add up through the course of the run. Seems like the load speed resembles the ps2 speed, where we get to the church at about the 6 min mark (with optimal lines/menus), still seems slower than ePSXe.

This means that wr is easily beatable if anyone wants to grind :)

@wolfdnc I think what @WanderAgro means is that there is a Duckstation port for Android:

https://github.com/stenzek/duckstation/releases

(see duckstation-android.apk)

I believe this can be used if players run on a chromebook for example. I might be wrong (Wolf and the other mods decide on this) but I think all duckstation ports (windows, linux, android, either SDL or QT) should go on "Emulator", and all other non-duckstation emulators in any platform (ePSXe, PCSX, etc) go on "Emulator (Legacy)". Not sure if there are any known load time differences between duckstations across different operating systems. I tested on Windows and Linux and they seem consistent (Javelin glitch is equally horrible on both), so by extension I'm guessing Android should be similar too.

I saw your code for the ePSXe autosplitters and noticed that the game time is separated into frames (named milisseconds in the code), minutes and multiples of 256 minutes. I tried to do the same thing on duckstation for the JP version by loading different saves (with different elapsed game times), and playing around with cheat engine's increase/decrease scan, but did not have any luck at all narrowing down the correct memory locations, which honestly makes me all that more impressed that you managed to do it for 3 different versions of ePSXe.

I also saw the IGT on your replay 100% run, it looks like the timer pauses in door transitions, the battle "heartbeats", the battle results screen and, importantly, in the pause menu. This means that if IGT was the measurement the strategies would probably look very different, e.g. you could attach some items to make the mandatory battles shorter on any%, and triggering battles wouldn't hurt IGT as much as it hurts RTA. I guess it makes sense then to use RTA, which will also favor players with good menuing and battle dodging skills.

So are runs in other emulators no longer accepted moving forward? No problem at all for me to run using duckstation from now (I agree it shouldn't be pay to win) but it is still true that ePSXe is the fastest way to run the game right? (considering we use RTA as measurement) so people might still be interested in it, especially for 100% runs where the time difference will be pretty noticeable with so many door loads (and javelin glitches which are also slower on duckstation) and definitely more noticeable for the player himself (i.e. how long it takes to finish a run).

Just a random thought regarding the issue: It would be really cool if we could find where the IGT is stored (e.g. using cheat engine) so load times are no longer a problem and people had more freedom to choose the emulator without worrying about these things. The difference between RTA and IGT is pretty brutal. When we check for 100% at the end of the game, the IGT is 60-70% of the RTA, and I always wondered what my actual IGT would be, even if the game doesn't show it to the player, and if it actually stops as soon as we beat eve.