I believe so, yes.

As of today 19/11 we are releasing an improved version of the autosplitter we use for STEAM runs. Unfortunately, we cannot 100% assure that this will be the last time it is modified, because of how truly difficult it is to balance and make fair for everyone. If you find any issues with this autosplitter contact the mods and we will look into it. So why are we doing this again?

Because the previous timer:

1. did not remove all loads in their entirety (it stopped pausing before loadings were finished)
2. it had an un-finished feature called framecounter which attempted to compensate for frames (FPS) lost in-game converting them into time but unfortunately this led to the infamous checkpoint bug where the timer would sometimes go nuts and go backwards a couple seconds or even entire minutes after resetting a checkpoint anywhere and on another hand it created inaccuracy between runners since the time "compensated" by this feature was not consistent at all.
3. it wasn't global, meaning it didn't work for things like No Merchant and Easy/Amateur without having to create a whole new autosplitter file changing some things.

This newer timer features:

1. full load-removal (as far as we could test). If you notice a loading isn't entirely removed, bring it up with the mods and we will try to fix it with the people that created the script.
2. works globally, for any category you'd like to run
3. adds key item splits which will split whenever you pick up a key item that you can select and customize through Edit Splits -> Settings
4. less lag due to more concise and readable code

This new autosplitter should appear automatically on Edit Splits if you have RE4 Steam selected as the game. If you don't want key item splits or if you wanna keep using your previous splits, you can keep using the regular doorsplits and uncheck key item splits on the settings window.

This will not invalidate runs done with the previous LRT.

If you need them, you will find splits and the ASL file on the resources tab of this site.

If you have any questions, leave them down below.

All credit goes to Wipe, Yuushi and Mysterion for their work on this newer autosplitter!

As Trichael says, it's not mandatory.

Quick update since this is a very important situation for the board and the community.

Our plan right now is to first of all make our load remover truly remove 100% of the loads in any hardware, and once we have that, removing the framecounter, since it is clear it is too random, causes too many issues, and we cannot make it work properly and make it fair for everyone. By removing this but fixing all the loads, the times should be a bit faster than what they are right now with the current timer, instead of slower as it was going to be with the initial new LRT, which is a big positive. However, we will need to have a serious discussion as to what to do with actual framedrop lag (for example dropping to 58, 57, or even lower FPS for several seconds which slows down the game significantly thus creating a disadvantage) specially for high level runs from here on out since we won't have the FC.

I appreciate everyone that shared their thoughts here by the way, and I'm sorry if I came across as too stubborn.

As I was saying to Mysterion on Discord and since I wanna avoid misinformation as much as possible, I'll correct something.

[Mysterion_06_] Having the new LRT as a our timing method is far more fair than the current one; the new one removes EVERY load in the game (Which destroys the P2W gap and only leaves the frame dropping open (like EVERY other speedgame also has)(So they unfair and P2W too??)), while the current one just returns a broken time and its still questionable how so many of you can support a timer that returns a false pb and then also be proud of that. (Its like running in a marathon, but finding a shortcut and abusing that, even though its not allowed).

As I said before, the first statement is wrong, since neither of the two timers we have stop until where they should, and even though this wouldn't be a big problem in most games, IT IS on this game due to how optimized the speedrun is and the ridiculous amount of loadings everywhere, creating the big gap between hardware (20-40s). This also explains why the IGT difference between good and bad PCs is usually 1 minute or more, cause you have those 20-40s PLUS the other half of the loading that IS removed on both LRTs we have. The reason people are "happy" with this timer is simple, it gives VERY similar times to EVERYONE using it with a difference of frames, while also allowing people with bad loadings to still compete. Despite how flawed it is, it's still the closest thing we've had to the game not being pay to win. We can be discussing this forever, but it really boils down to this: the main reason the framecounter is still relevant is a direct cause of our load removal method/memory address being bad and inaccurate.

[Mysterion_06_] "20sec timeloss with new LRT cuz bad pc" is also false, it wouldnt even be any timeloss unless you like for real drop under 50fps for a longer period of time, cause losing time on less fps is only then inefficient when your game literally freezes OR when you notice a slowdown of your game and that wont happen unless you are like below 50fps even less maybe (Ive ran myself on a shitty pc so I know).

This is not true, even if you don't drop a single frame during GAMEPLAY, if you drop frames during/right after loadings your loading times will be slow or rather slower than on a top-pc, and you WILL lose time on the new timer on each loading and that WILL add up to the crazy differences we have seen. The argument that ALL or the MAJORITY of the time that you lose using the new script vs old one is due to framecounter rewinds OUTSIDE of the loading time rewinds is completely false. Again, if this were true anyone that drops a lot of frames would have by far the fastest gold splits.

Take this clip and compare it to this one I lose literally half a second to the new LRT while shima loses .16, all because of the loadings. Since FPS drop during/right after loadings, the current LRT gives me back most of the stupid time I lost on those loadings due to the load remover being inaccurate, while this gets completely ignored on the new LRT.

Here's another comparison, this time with a split that is literally a straight line to make it more noticeable: no framedrops whatsoever except for the ones that always happen after loadings, yet I lose time. You can also consider the fact that Shima got 61fps there, which means his current LRT time should've been faster, in which case he probably would've been +-0.00 compared to the new LRT.

Basically the point is.. loading times are tied to framerate, and everybody drops frames during/after loadings, except the bad PCs drop way more, so even if you don't drop any frames during GAMEPLAY, you will still be at a massive disadvantage with the new LRT if you drop more frames during/right after a loading (which translates to a longer loading) than someone else. Meaning that the new LRT isn't accurate either unless EVERYONE has the EXACT same loading times (which unless everyone has an i9 and 3080 is never gonna happen), and even then it probably still wouldn't be accurate, due to the fact that the timer starts counting even before the top-pcs finish the loading, AND the fact that even the inventory load removal itself is inaccurate/doesn't start and stop at the right times, which, considering we open and close the inventory hundreds of times during the run, already creates time differences. It is the same issue the IGT always had, it's "accurate" IF and ONLY IF everyone in the comparison has the fastest hardware. The only way a timer we use would be "accurate" since we use this word so much, would be a timer that stops on EVERY single blackscreen during the ENTIRE duration of it, from its first frame to its last.

I'm also still waiting for Yuushi to provide proof for his claims.

[Yuushi] - There are low-spec PC runners who are advantaged by the current LRT and high-spec PC runners who are disadvantaged.

Can you show examples with video proof of this? As I mentioned previously, we have received no proof of this whatsoever. The only "proof" we've seen is a canyon comparison clip which was totally inaccurate because 1. Canyon is a long, execution based and rng based section 2. the rng wasnt the same in the two clips shown 3. the videos were edited 4. canyon is optimal on KBM which also makes the comparison inaccurate already

[Yuushi] - but If you intentionally cause frame drops, the time will be even more inaccurate because "fake time save" are more likely to give. "LRT bug gives high-spec PC runners an disadvantage." This is an unusual situation not only for us, but for new runners as well. Because most of them will be preparing a high-spec PC to run this game, but as it is, it's pointless.

I don't know where you get the idea from that new runners would be preparing a high spec PC to run a 2014 port of a 2005 game? It is the opposite actually. This game should NOT require you to have a good PC to run it properly. Conversely to what I said before, we also don't have proof of high spec PCs being at a disadvantage, EXCEPT potentially the 61fps thing. Derek has a better PC than Shimarisu and almost as good of a PC as you and not only did he not report feeling at a disadvantage, he also has the fastest segments nearly everywhere, AND on controller (which is generally not 100% optimal). How do you explain that?

[Yuushi] - As you can see, formulas are written. (frameCount * 16.6666) What do you think would happen if you rewrote these numbers? As you may have guessed, that would mean rewriting the timer method itself. Even a non-programmer like me can manipulate time by just rewriting a few of the numbers. The biggest problem is, this is hard to find out, and I think it's worse than any of the cheating with trainers.

Anyone can change that value if they want, that's true. However, as mentioned above, it would be incredibly easy to detect because 1. we now require framerate to be showing throughout the whole run which lets us corroborate the framecounter rewinds if necessary 2. this would give a consistent time advantage everywhere, including splits that are nothing but a straight line, which would be incredibly blatant. We have enough experience to know what times can be achieved and what times CANNOT be achieved on nearly any split at this point. Getting for example 7.07 on split 61 or 1.20 on split 59 is impossible.

[Yuushi] - When several people were asked to compare the current LRT with the new LRT, the difference was about 10-12 seconds for 3 out of the 5 people tested in the Main Game category. From this, I thought that if you don't have dropped frames (Except for load time), there is a difference of about 10 seconds.

The sample is not big enough to draw conclusions, however, as I said in my post, it is obvious that runners having a top-PC is the exception here, not the norm. The difference of 20-30s (from 10-12 to 30-40) was actually not from framedrops (for the most part) it was from bad load removal, which I explained in my post, in point number 1. Since the timer barely stops at loadings, the good PCs were not disadvantaged nearly as much as the bad PCs. I might have forgotten to mention, but Wipe's new script didn't fix this load removal issue at all. If it did, we would have actually seen how much time is lost from framedrops, which in theory, shouldn't be much.

[Yuushi] - Without wanting to sound rude but just like in any other sport if you can't compete you need to upgrade your rig. You all think that using the new LRT will be a pay to win and that's true, but it's better than IGT and it's not that difficult to create an environment where frame drops don't cause.

I do not support the pay to win argument in any way and I know most people here don't either. You cannot just assume that people will have the opportunity to upgrade their PC to be able to compete, and even if they did have the chance to do it, there's no reason why they should be forced to spend money to compete. As I said before, framedrops are not the biggest issue here, it's the loads. Even if you don't drop a single frame, if your loading times are bad, you will be at a huge disadvantage with the new script. (which is my case for example, I don't drop to more than 58 almost ever)

[Yuushi] - Since the new LRT removes load time, the difference in time between the new LRT and the current LRT is caused by the presence of frame drops. In short, just don't frame drop.

This is incorrect again. The new LRT DOES NOT remove loads properly, just like the current one. The difference is with the current one you at least get some of that time back with framecounter rewinds.

If Wipe or Mysterion come up with a new method for load removal that removes loads ACCURATELY we can discuss removing the framecounter. In my opinion, as it stands right now it is unviable to remove it due to this problem.

Hello everyone. I will try to keep this as short yet as informative as possible, presenting all the facts. To put it on simple terms, the way the current LRT works is it pauses on each loading screen, it pauses when you pause the game with the options menu, and it pauses when you open and close the inventory (while the duration of that short blackscreen). It also has a feature called framecounter that counts the amount of frames (FPS) you dropped in-game and then converts those to time rewinding the timer backwards. As most active runners know by now, this current LRT we use has had and still has several issues which I will list below, along with proof/examples of them.

1. The load removal feature is inaccurate due to the timer un-pausing before the loading has even finished. This happens on every single loading/blackscreen of the run including each inventory open/close blackscreen (there’s several hundreds of both of these combined during the whole run) and we strongly suspect it is related to the in-game memory address for the loading screens/blackscreens being unreliable, so we are unsure if it can be fixed. Examples: I specifically chose myself cause I’ve a “bad” PC (should be more than good enough for this game but we all know the Steam port is a special kind of unoptimized) and Derek because he has a really good PC and also records at 30fps just like me so the comparison is as accurate as possible. As you can see on the last clip, the framecounter counteracts this problem to some degree due to the fact that it counts frames during loadings as well, and as you can see it rewinds roughly a similar amount to the amount you lost after each loading when you pause/open the inventory. You should also consider that Derek might very well not have the fastest loading times possible, so the differences could be even bigger. Or,. if I didn’t have an SSD, the difference would be bigger. You might be thinking losing 0.1, 0.2, or potentially more because of hardware limitations isn’t a big deal, but when you consider that there’s literally hundreds of loadings in the game, it becomes incredibly significant.
2. The framecounter feature of the timer is inaccurate which causes discrepancies of frames randomly (you can either lose or save a few frames). However, since the load removal feature is also inaccurate, this can counteract it by giving you “fake timesave” which balances out the time you lost from inaccurate load removal as explained in point 1). Example:
3. The timer can go backwards for no reason when reloading checkpoints (this is a bug related to the framecounter). Examples: This, although annoying, is easily fixable since the timer pauses during the options menu so all we have to do is manually add that fake timesave to the final time when submitting the run which is what we’ve been doing for a while.

One of the main theories that arised from the issues with the timer is the theory that since bad PCs drop a lot of frames, bad PC = free fake timesave. However, we could not find any reliable proof of this. We have a google sheet specifically dedicated to record each and every gold segment of each top runner for analysis-purposes. If this theory was true I (since I have enough skill and unfortunately the worst PC out of all the top runners) would have the best gold by far on many splits, specially laggy ones where I would drop a lot of frames, and similarly, anyone with a bad PC would have the fastest segment on segments that are nothing but a straight line, but neither of these were the case. In fact, nearly every split where I have the best gold can easily be attributed to the fact that those segments are FAR more optimal on KBM, and would be unmatchable on controller no matter the skill (for example all merchants, or splits that require fast shots and quickturns, like village treasures). Moreover, Derek (current 4th place runner) has the majority of the best golds DESPITE playing on controller AND having a good PC AND never dropping a single frame, which again disproves this theory. Currently, the timer is inaccurate (by frames) for everyone, and not just for bad PCs. The only thing related to good PCs being at a disadvantage we strongly suspect of is that getting 61fps (we don’t know why this happens) equals timeloss, which would explain why for example Shimarisu seems to get worse times on many segments (he gets 61fps very often).

In order to fix this messy situation, recently Wipe (the developer of the current timer) developed a new script which modifies how the autosplitter detects/splits on doors, and therefore making it universal for any category you run which is a big positive, but at the same time it straight up removes the framecounter which, although it fixes problem number 3, it massively exacerbates the issue of the inaccurate load remover and it also makes dropping frames in-game a huge issue again. Just a few days ago we tested this new timer with success, however, we quickly realized the massive time differences between good and bad computers, with bad computers losing 20-40s compared to the current timer, and good computers only losing about 10-12s.

The main reason why the LRT came to be in the first place was to eliminate the pay to win aspect of the IGT, and therefore expanding the competition fairly, since not everyone has the possibility to afford a good computer. Not only that, but if we consider the fact that the Steam port of the game is incredibly poorly optimized, with some people with good to amazing PCs having performance issues and also small yet noticeable discrepancies in loading times between good VS top-tier computers, this makes it so on the new script even a GOOD PC won’t be able to compete with a TOP-TIER PC. Someone could bring up the argument that losing 30-40s for example to hardware limitations would not affect someone with a non-top time which is most of the leaderboard, but I personally heavily disagree because 1) this assumes these people will never reach the top which we simply cannot say for sure (in which case this time difference would make them unable to keep competing at that point) 2) it is very unenjoyable to lose time to things you can’t control, in this case hardware limitations, which I believe would make many runners quit the game and 3) the leaderboard was and is made for everyone that runs the game, not just for the top 5 or 10, therefore I believe the logical course of action is to do what benefits the most amount of people in the community.

We think that although the current timer is quite flawed and not exactly accurate, the issues with switching to the new one would heavily outweigh the benefits, and we believe moving back to a timer that is very hardware-reliant and that would only benefit a small amount of people (if we look at the leaderboard and the differences between IGT and LRT it is clear that runners having a top-pc is the exception, not the norm) while screwing over the rest is a huge step-back. However, we are still analyzing what to do with the current situation, and since we have not come to an agreement yet, we want to know what active/former runners of the Steam version of the game think of all this, and if you have any reasonable suggestions, feel free to leave them down below.

^ plus any proposed category would have to be requested by multiple people in the community in order to be added to the boards imo.

Not sure where you read that, but for console runs it is not necessary to use a timer, since runs are timed by IGT which is the time that shows up at the end when you make your one and only save. Using a timer on console is mainly to keep track of timesaves/timelosses easier. Just keep in mind that the default difficulty the game chooses when you don't have the option to do so is Normal, so make sure to submit your run under that category if you decide to go with it. I hope I was clear!