Ingame timer or Realtime [locked]

Edit history:

Darazanjoll: 2014-04-23 12:22:04 pm

Darazanjoll: 2014-04-23 11:26:32 am

2014-04-23 11:17:54 am

2 posts

So this thread is meant to be a discussion about what timer to go by when you're speedrunning RE2.
Some people want to use realtime and some want to use ingame timer and I would like to hear everybody's
opinion about this. First off I'll tell you what I think.

I think that everybody should use realtime because the ingame timer for RE2 isnt correct at all. I'm not
talking about that its slower compared with realtime but that its random and the time will differ almost
every time, this occurs all the time when people are racing on SRL (were everybody have to go by realtime).
The person who wins the race will not always have the best time for the ingame timer.

I will provide you with some pictures that I've taken from a few runners and compare them with each other and
also themselves. Everybody is using the program for the ingame timer/Health meter and a realtimer so all I had
to do was to look what the ingame timer was when everybody been playing for 50min in realtime.

Runs:

ArcadiaSkies: 45:52
ArcadiaSkies: 45:58
Darazanjoll: 46:40
Darazanjoll: 46:55

Fiercekyo: 46:24 http://www.twitch.tv/fiercekyo/b/478366533
Fiercekyo: 46:38 http://www.twitch.tv/fiercekyo/c/3560429
Deathnite11: http://www.twitch.tv/deathnite11/c/3480336

As you can see everybody had been playing Claire A for 50min in realtime and started the timer at the same time.
What we can see from the pictures is that the ingame timer doesnt only differ between other players but
also themselves. The biggest difference was one of Arcadias run compared with my own run. He was 1 minute
and 3 seconds ahead of me according to the ingame timer but in realtime we where right next to eachother.

Going by the ingame timer becomes a really big issue for those who want to improve there PB's and they will
have no chance in the world to break a WR with their current setup, using realtime though would solve everything.

Lets not forget the current WR's and all the PB's. I dont think that everybody would like to switch to realtime
because then we would have to time every run and change it from ingame time to realtime. My suggestion is that we
can have 2 WR for each scenario. one for realtime and one for ingame time.

Lastly I just want to make something clear. I do NOT complain about the actual program that allow us to see the ingame
timer and health meter which is made by Dchamps and Mortician. As far as I'm concerned it doesn't effect the ingame timer because I've done runs without it and
got the same result as I've been explaining.

So please tell my your opinions and let the discussion begin ;)

Thread title:

So this thread is meant to be a discussion about what timer to go by when you're speedrunning RE2. 
Some people want to use realtime and some want to use ingame timer and I would like to hear everybody's
opinion about this. First off I'll tell you what I think.

I think that everybody should use realtime because the ingame timer for RE2 isnt correct at all. I'm not 
talking about that its slower compared with realtime but that its random and the time will differ almost 
every time, this occurs all the time when people are racing on SRL (were everybody have to go by realtime). 
The person who wins the race will not always have the best time for the ingame timer.

I will provide you with some pictures that I've taken from a few runners and compare them with each other and 
also themselves. Everybody is using the program for the ingame timer/Health meter and a realtimer so all I had 
to do was to look what the ingame timer was when everybody been playing for 50min in realtime.

[url=http://postimg.org/image/yb02xnu0x/][img]http://s24.postimg.org/yb02xnu0x/sample.jpg[/img][/url]

Runs:
[hidden]ArcadiaSkies: 45:52 https://www.youtube.com/watch?v=d9f1zGJKJT4
ArcadiaSkies: 45:58 https://www.youtube.com/watch?v=EjIJ7Uq5w6c
Darazanjoll: 46:40 https://www.youtube.com/watch?v=36rfaG1UFpo
Darazanjoll: 46:55 https://www.youtube.com/watch?v=_SMj7VRaYfk

Fiercekyo: 46:24 http://www.twitch.tv/fiercekyo/b/478366533
Fiercekyo: 46:38 http://www.twitch.tv/fiercekyo/c/3560429
Deathnite11: http://www.twitch.tv/deathnite11/c/3480336[/hidden]

As you can see everybody had been playing Claire A for 50min in realtime and started the timer at the same time. 
What we can see from the pictures is that the ingame timer doesnt only differ between other players but 
also themselves. The biggest difference was one of Arcadias run compared with my own run. He was 1 minute 
and 3 seconds ahead of me according to the ingame timer but in realtime we where right next to eachother.

Going by the ingame timer becomes a really big issue for those who want to improve there PB's and they will 
have no chance in the world to break a WR with their current setup, using realtime though would solve everything.

Lets not forget the current WR's and all the PB's. I dont think that everybody would like to switch to realtime 
because then we would have to time every run and change it from ingame time to realtime. My suggestion is that we 
can have 2 WR for each scenario. one for realtime and one for ingame time.

Lastly I just want to make something clear. I do NOT complain about the actual program that allow us to see the ingame 
timer and health meter which is made by Dchamps and Mortician. As far as I'm concerned it doesn't effect the ingame timer because I've done runs without it and 
got the same result as I've been explaining.

So please tell my your opinions and let the discussion begin ;)

Fiercekyo

2014-04-23 12:28:05 pm

12 posts

I really don't see the point of switching from Game Time to Real Time. if someone's timer is running slower than others because they have a less powerful computer then it will probably affect the real timer as well because it's just not running the game that well(in my opinion anyways).

Dchaps

Edit history:

Dchaps: 2014-04-23 12:51:52 pm

2014-04-23 12:47:37 pm

74 posts

http://outbreakf12.blogspot.ca/

First of all for the record, my program only "reads" from memory , it does not write in anyway (altering the game) it is simply a viewing window into the memory location that holds the timer, it is converted into time format with code, because otherwise "00:3:00" would just be "180" in memory, meaning in memory it counts by seconds, so its simply a time conversion function that changes it from "seconds" to real time format. Everyone's game is programmed to use the same location for the timer, which means when you start the game(not at the menu) the seconds timer resets and starts ticking and the game considers it "finished" when you reach the end of the tunnel and that's how it goes for everybody, now about the picture i dont really know what to think of it but do you seriously think every one of those players were mirrored? those are totally different runs , people probably got bit at different times , made errors in some form, people screw up the door skipping thing all the time, so its impossible everyone's skipped simultaneously, someone probably ate a few black screens another didn't . So the time difference in those pictures is due to peoples mistakes, its not like there minutes off, there within like 20 seconds of each other, 1 wrongly skipped door can eat 2 seconds make sure your thinking about all the possibilities here.

in conclusion , why do i think there's nothing wrong with the game timer ? "splits" off the in game timer and these match up everytime i play, aslong as i dont make mistakes, the in game timer to me seems to be a flawless guideline of when to be at point x, and other players have got to point x at the same times as long as they were on the same pace. When ive been in record wars with someone like Elyzsabo we both come to the same conclusion that "6:01" is a great time to push the 2nd statue, thats just an example of how we all seem to reach the same points around the same time as long as we were both mirrored.

I agree with fierce that there is no point in doing this

First of all for the record, my program only "reads" from memory , it does not write in anyway (altering the game) it is simply a viewing window into the memory location that holds the timer, it is converted into time format with code, because otherwise "00:3:00" would just be "180" in memory, meaning in memory it counts by seconds, so its simply a time conversion function that changes it from "seconds" to real time format. Everyone's game is programmed to use the same location for the timer, which means when you start the game(not at the menu) the seconds timer resets and starts ticking and the game considers it "finished" when you reach the end of the tunnel and that's how it goes for everybody, now about the picture i dont really know what to think of it but do you seriously think every one of those players were mirrored? those are totally different runs , people probably got bit at different times , made errors in some form, people screw up the door skipping thing all the time, so its impossible everyone's skipped simultaneously, someone probably ate a few black screens another didn't . So the time difference in those pictures is due to peoples mistakes, its not like there minutes off, there within like 20 seconds of each other, 1 wrongly skipped door can eat 2 seconds make sure your thinking about all the possibilities here.

in conclusion , why do i think there's nothing wrong with the game timer ? "splits" off the in game timer and these match up everytime i play, aslong as i dont make mistakes, the in game timer to me seems to be a flawless guideline of when to be at point x, and other players have got to point x at the same times as long as they were on the same pace. When ive been in record wars with someone like Elyzsabo we both come to the same conclusion that "6:01" is a great time to push the 2nd statue, thats just an example of how we all seem to reach the same points around the same time as long as we were both mirrored.

I agree with fierce that there is no point in doing this

Carcinogen

Edit history:

Carcinogen: 2014-04-23 01:06:38 pm

2014-04-23 01:06:03 pm

204 posts

Quote from Fiercekyo:

I really don't see the point of switching from Game Time to Real Time. if someone's timer is running slower than others because they have a less powerful computer then it will probably affect the real timer as well because it's just not running the game that well(in my opinion anyways).

Yeah, but if so many peoples' timers are that inconsistent but the games run at the same pace, how does that make game time a good metric to run the game by? Treesus proved that you can fuck with the timer a bunch by hitting F8 a lot and maximizing/minimizing the game. My audio fades out sometimes when I skip the door at the lag frame.

It's actually very possible that the game engine, since it is an older PC game, doesn't keep up very well. I mean, this game is already having problems running on peoples' computers as it were.

But like, what I'm seeing is not from the point of view of the mistakes that the runners made at one time, but from the perspective of the two timers. The Real Time is all at 50 minutes, but the in-game timers are everywhere.

That's just weird.

The reason I wanted to see videos was to see where all these runners started their game timers. I'll look over them tonight.

[quote="Fiercekyo"]I really don't see the point of switching from Game Time to Real Time. if someone's timer is running slower than others because they have a less powerful computer then it will probably affect the real timer as well because it's just not running the game that well(in my opinion anyways).[/quote]

Yeah, but if so many peoples' timers are that inconsistent but the games run at the same pace, how does that make game time a good metric to run the game by? Treesus proved that you can fuck with the timer a bunch by hitting F8 a lot and maximizing/minimizing the game. My audio fades out sometimes when I skip the door at the lag frame.

It's actually very possible that the game engine, since it is an older PC game, doesn't keep up very well. I mean, this game is already having problems running on peoples' computers as it were.

But like, what I'm seeing is not from the point of view of the mistakes that the runners made at one time, but from the perspective of the two timers. The Real Time is all at 50 minutes, but the in-game timers are everywhere.

That's just weird.

The reason I wanted to see videos was to see where all these runners started their game timers. I'll look over them tonight.

Jake Tabor

2014-04-23 01:26:46 pm

49 posts

Howdy folks! I'm happy to post in this nice forum.

On to the topic at hand, I think that using RTA would be a bad idea. When you click out of the game by accident the IGT pauses as well. That negates consistency if we were to use RTA. so, there's one pro of using IGT I can think of on the top of my head. Another pro would be that the IGT remains the same regardless if it's on slow PAL mode or regular mode. I agree with Fiercekyo and Dchaps.

Rebecca

2014-04-23 01:36:04 pm

if you use only RTA I think is bad because the speed of the timer depends of the PC. Sometimes timer go fast and sometimes go slow. I don't know too much about pc version but when I play I realize that sometimes the timer changes his speed in my PC

Carcinogen

Edit history:

Carcinogen: 2014-04-23 02:10:55 pm

2014-04-23 02:04:21 pm

204 posts

One other thing that crossed my mind: since the game timer is in its own environment, what if the timer is slowing down and speeding back up because it slows down and speeds up when the game engine itself lags?

That would explain why elyzsabo and dchaps are able to say "such and such time in game-time is a good time to be pushing the statues" and still be at exactly those times in a perfect run.

The case you are making, Darazanjoll, assumes that the game timer runs in sync with Windows system processes, and counts up by your computer's CPU clock... But you can see that whenever you click outside the window, the timer pauses. The game timer runs in sync with the game engine's logic, and pauses when you click outside of it. Since it needs to be programmed a specific way as a console game port, the game timer will run exactly the same whether it lags or not... Of course it's going to be inaccurate vs. Real Time if you were racing, but it's not going to be inaccurate between two records. Let me explain.

Resident Evil 2 PC, as a port of a console game, runs at 30FPS. It can go lower, but it can't go any higher, otherwise the game will overclock and go too fast. It sometimes manages to underclock itself into a sort-of "PAL Mode" pretty badly, for an unknown reason. Consequentially, the game engine must run within its own processing "environment" to be able to continue running consistently, or it will run extremely fast like with an old-school MS-DOS game.

The recent release of Resident Evil 4 UHD on Steam had a similar issue. You can see that as a port of a console game, it can only process the game at 30FPS or 60FPS logic, but not any higher, when most native PC games render frames to keep in sync with the system clock. When RE4 UHD on Steam was run at 60FPS prior to the patch, you had video that was still running at 30FPS logic (the CG movie files needed to be remastered with 60 frames per second in order to make them run at 60FPS), and audio that ran at 60FPS logic (audio isn't something that renders in frames), so it desynced.

So, it refreshes data in a specific number of cycles in the game engine... Consequentially, equivocal to the number of frames rendered. The game lagging means the engine is lagging, and so is the timer. During the door skipping, when the game lags, so does the timer, but it always stays in sync to the 30 frames per cycle rule.

That is why more reset-heavy players like Jake, Dchaps, and Fierce manage to see "consistency" when they reach certain points in their runs all the time, because the game timer is keeping up with them as they play the game, not in real time. The game timer counts load times, it counts cutscenes, but it doesn't count lag... instead it pauses when the game engine pauses when it is trying to figure out what the next thing is that it needs to process.

I think you have a very real chance of competing with Insane's time. I'll watch them tonight, and if I can tell you where you can save 47 seconds, then you should try it.

One other thing that crossed my mind: since the game timer is in its own environment, what if the timer is slowing down and speeding back up because it slows down and speeds up when the game engine itself lags?

That would explain why elyzsabo and dchaps are able to say "such and such time in game-time is a good time to be pushing the statues" and still be at exactly those times in a perfect run.

The case you are making, Darazanjoll, assumes that the game timer runs in sync with Windows system processes, and counts up by your computer's CPU clock... But you can see that whenever you click outside the window, the timer pauses. The game timer runs in sync with the game engine's logic, and pauses when you click outside of it. Since it needs to be programmed a specific way as a console game port, the game timer will run exactly the same whether it lags or not... Of course it's going to be inaccurate vs. Real Time if you were racing, but it's not going to be inaccurate between two records. Let me explain.

Resident Evil 2 PC, as a port of a console game, runs at 30FPS. It can go lower, but it can't go any higher, otherwise the game will overclock and go too fast. It sometimes manages to underclock itself into a sort-of "PAL Mode" pretty badly, for an unknown reason. Consequentially, the game engine must run within its own processing "environment" to be able to continue running consistently, or it will run extremely fast like with an old-school MS-DOS game.

The recent release of Resident Evil 4 UHD on Steam had a similar issue. You can see that as a port of a console game, it can only process the game at 30FPS or 60FPS logic, but not any higher, when most native PC games render frames to keep in sync with the system clock. When RE4 UHD on Steam was run at 60FPS prior to the patch, you had video that was still running at 30FPS logic (the CG movie files needed to be remastered with 60 frames per second in order to make them run at 60FPS), and audio that ran at 60FPS logic (audio isn't something that renders in frames), so it desynced.

So, it refreshes data in a specific number of cycles in the game engine... Consequentially, equivocal to the number of frames rendered. The game lagging means the engine is lagging, and so is the timer. During the door skipping, when the game lags, so does the timer, but it always stays in sync to the 30 frames per cycle rule.

That is why more reset-heavy players like Jake, Dchaps, and Fierce manage to see "consistency" when they reach certain points in their runs all the time, because the game timer is keeping up with them as they play the game, not in real time. The game timer counts load times, it counts cutscenes, but it doesn't count lag... instead it pauses when the game engine pauses when it is trying to figure out what the next thing is that it needs to process.

I think you have a very real chance of competing with Insane's time. I'll watch them tonight, and if I can tell you where you can save 47 seconds, then you should try it.

Carcinogen

Edit history:

Carcinogen: 2014-04-23 02:31:44 pm

2014-04-23 02:29:06 pm

204 posts

Funny... Actually, this also manages to explain how PS1 RE games run on emulator and PS1 RE games run on PSP can compete just fine as well. In that version, as long as the loading remains consistent (read: all the game's data is loaded before the door closes) then not only do the fade-ins sync, the game timers run consistently in the same places too.

It's painfully obvious that emulators and PSP load games fast, but while the timers within the game engines remain consistent, they are still slower if there is a hardware bottleneck like disc drives affecting the loading times... loading time and lag being mutually exclusive.

C

Carcinogen

Edit history:

Carcinogen: 2014-04-24 06:37:45 am

2014-04-24 06:36:33 am

OK, so this is going to prove my hypothesis for why the game timer is actually perfectly fine for running the game.

RTA Times (From Difficulty Select/RESIDENT EVIL... TWOOOOOOO to Escape Passage Fadeout)
Darazanjoll: 54:17 (50:52 IGT)
Insane: 54:17 (50:05 IGT)

Seems fishy, doesn't it? I'm using line trajectories and movement fluidity to determine the time differences, instead of split timers. This is how anyone is supposed to find time gains/losses if you are not using a split timer.

OK so, this is where I see all of Insane's time gains.

=Street=
Darazanjoll: Unlucky Zombie Spawn in street area near Gun Shop forces imperfect movement.
Insane: Clear route to Gun Shop; hits the door some fractions of a second earlier.

=Basketball Court=
Darazanjoll: You delayed your zombie grab-shove a little bit until the second zombie was lined up (1:07 at gate)
Insane: Perfect 1-bite Basketball Court with no flinching (1:05 at door)

=Dumpster Alleyway=
Darazanjoll: You got hung up in the dumpster alleyway waiting for the zombie to move (1:27)
Insane: Absolutely ballsed-out in front of the last zombie in dumpster alleyway (1:25)

=Entering Bus=
Darazanjoll: 1:33
Insane: 1:31

=Bus=
Darazanjoll: No QS, 3/2 (1:43)
Insane: Uninterrupted QS, 3/2 (1:40)

=Entering Police Station=
Darazanjoll: 2:00
Insane: 1:57

That's 3 seconds game time. Insane obviously had a cleaner run there. Next post will be 1st half of police station to Ada Gunshot.

OK, so this is going to prove my hypothesis for why the game timer is actually perfectly fine for running the game.

RTA Times (From Difficulty Select/RESIDENT EVIL... TWOOOOOOO to Escape Passage Fadeout)
Darazanjoll: 54:17 (50:52 IGT)
Insane: 54:17 (50:05 IGT)

Seems fishy, doesn't it? I'm using line trajectories and movement fluidity to determine the time differences, instead of split timers. This is how anyone is supposed to find time gains/losses if you are not using a split timer.

OK so, this is where I see all of Insane's time gains.

=Street=
Darazanjoll: Unlucky Zombie Spawn in street area near Gun Shop forces imperfect movement.
Insane: Clear route to Gun Shop; hits the door some fractions of a second earlier.

=Basketball Court=
Darazanjoll: You delayed your zombie grab-shove a little bit until the second zombie was lined up (1:07 at gate)
Insane: Perfect 1-bite Basketball Court with no flinching (1:05 at door)

=Dumpster Alleyway=
Darazanjoll: You got hung up in the dumpster alleyway waiting for the zombie to move (1:27)
Insane: Absolutely ballsed-out in front of the last zombie in dumpster alleyway (1:25)

=Entering Bus=
Darazanjoll: 1:33
Insane: 1:31

=Bus=
Darazanjoll: No QS, 3/2 (1:43)
Insane: Uninterrupted QS, 3/2 (1:40)

=Entering Police Station=
Darazanjoll: 2:00
Insane: 1:57

That's 3 seconds game time. Insane obviously had a cleaner run there. Next post will be 1st half of police station to Ada Gunshot.

Carcinogen

2014-04-24 08:14:17 am

204 posts

=Entering Police Station=
Darazanjoll: 2:00
Insane: 1:57

=Licker Hallway=
Darazanjoll: Took a wider angle around the licker and had to correct trajectory a bit to get to the door (4:18)
Insane: Did a one or two frame turn and hit the door at a straighter angle (4:14)

=Fireplace=
Darazanjoll: 4:47
Insane: 4:42

Pretty sure this is extra second gained is from overall loads and movement... Lag is mutually exclusive from loads with IGT logic. There is no millisecond timer IGT, so I can't tell. It's safe to assume that the game counts frames, and trying to count game frames, outside of being extremely superfluous for what we're trying to do, is near impossible because of the game rendering frames vs. OBS/twitch recording the game. The framerate output of the game (which the game timer operates on) vs. the recording itself differ completely in framerate.

Similar issues exist with Zelda: ALttP Speedruns
https://forum.speeddemosarchive.com/post/important_analog_capture_frame_rates_and_mysterious_gained_seconds.html

=West Stairway=
Darazanjoll: 4:58
Insane: 4:53

=Statue Hallway=
Darazanjoll: 5:25
Insane: 5:20
Both statue pushes were the same, but Insane did a weird little backpedal there. Jikkachan's PS1 TAS would disagree:
http://www.nicovideo.jp/watch/sm19728039?ref=search_key_video

=Hands=
Darazanjoll: 7:11
Insane: 7:05
I saw you do a little D-Pad jiggling in that hallway. I think there was another gain from loads and movement, probably from the statue pushing.

=Return from Licker Hallway=
Another slightly skewed curve. Insane had a straighter trajectory.

=Unicorn Medal=
Darazanjoll: 7:40
Insane: 7:33
Interesting, hitting it in the middle. I'm pretty sure that added into another second gained, when you turned right to move toward the door after grabbing the key, instead of turning left.

=Licker Hallway=
Darazanjoll: 7:52 (170 Health)
Insane: 7:46 (148 Health)
Insane tanks a hit from the licker, loses 1 second.

=Crank=
Darazanjoll: 8:11
Insane: 8:04
You push the ladder 6 times, which is the correct number, and Insane pushes 7... How he gained a second here, I'm not entirely sure, but that sure was an interesting pivot.

=Licker Hallway=
Darazanjoll: 8:15
Insane: 8:09
Nevermind, it seems to sync back up. Probably just overall loadtimes and frames/milliseconds.

=Hands Hallway=
Darazanjoll: 8:26
Insane: 8:19
Insane had much cleaner movement here, which pretty much fully establishes a 1 second gain. 7 seconds ahead now.

=STARS Hallway=
D: 8:54
I: 8:45
You stopped here for safety. That's a 2 second loss.

=Library=
D: 9:35
I: 9:25
Insane had cleaner movement. He was one frame away from hitting the camera angle change before the panel slid down.

=2F Main Hall=
D: 9:51
I: 9:40

There's going to be full-second discrepancies between rooms. Every time I pause the video, due to the framerate discrepancy I mentioned earlier, the timing is not going to be accurate between rooms and you're going to see a gain and a loss here and there. If I leave it blank, then that's probably why, but if there's a gain, then it's probably from movement... Insane's movement is very very clean most places, so you're going to see why he gains seconds where he does. I'm only going to post times between groups of rooms now, and do writeups only for obvious errors... There will be two-second gains where those are.

=Art Storage Room=
D: 11:32
I: 11:20
Consistent. Neither of you use the Japanese version-exclusive Misty dodge to run behind her after getting the Valve Handle.

=Re-entering Save Room=
D: 12:00
I: 11:46
You hugged the wall and paused to clear the zombie. Insane just blew around it without hesitation. 14 second split.

=Before West Stairwell Hallway=
D: 12:57
I: 12:43

=Marvin=
D: 13:32
I: 13:16

You hit the wall and wasted movement while waiting for the other Misty to walk forward to run behind her. You also waited for the middle zombie in the evidence locker to hit the desk before executing the dodge. Insane didn't. He also exited the Marvin room faster.

=Ada Gunshot=
D:14:17
I: 14:01

16 seconds ahead now.

=Entering Police Station=
Darazanjoll: 2:00
Insane: 1:57

=Licker Hallway=
Darazanjoll: Took a wider angle around the licker and had to correct trajectory a bit to get to the door (4:18)
Insane: Did a one or two frame turn and hit the door at a straighter angle (4:14)

=Fireplace=
Darazanjoll: 4:47
Insane: 4:42

Pretty sure this is extra second gained is from overall loads and movement... Lag is mutually exclusive from loads with IGT logic. There is no millisecond timer IGT, so I can't tell. It's safe to assume that the game counts frames, and trying to count game frames, outside of being extremely superfluous for what we're trying to do, is near impossible because of the game rendering frames vs. OBS/twitch recording the game. The framerate output of the game (which the game timer operates on) vs. the recording itself differ completely in framerate.

Similar issues exist with Zelda: ALttP Speedruns
https://forum.speeddemosarchive.com/post/important_analog_capture_frame_rates_and_mysterious_gained_seconds.html

=West Stairway=
Darazanjoll: 4:58
Insane: 4:53

=Statue Hallway=
Darazanjoll: 5:25
Insane: 5:20
Both statue pushes were the same, but Insane did a weird little backpedal there. Jikkachan's PS1 TAS would disagree:
http://www.nicovideo.jp/watch/sm19728039?ref=search_key_video

=Hands=
Darazanjoll: 7:11
Insane: 7:05
I saw you do a little D-Pad jiggling in that hallway. I think there was another gain from loads and movement, probably from the statue pushing.

=Return from Licker Hallway=
Another slightly skewed curve. Insane had a straighter trajectory.

=Unicorn Medal=
Darazanjoll: 7:40
Insane: 7:33
Interesting, hitting it in the middle. I'm pretty sure that added into another second gained, when you turned right to move toward the door after grabbing the key, instead of turning left.

=Licker Hallway=
Darazanjoll: 7:52 (170 Health)
Insane: 7:46 (148 Health)
Insane tanks a hit from the licker, loses 1 second.

=Crank=
Darazanjoll: 8:11
Insane: 8:04
You push the ladder 6 times, which is the correct number, and Insane pushes 7... How he gained a second here, I'm not entirely sure, but that sure was an interesting pivot.

=Licker Hallway=
Darazanjoll: 8:15
Insane: 8:09
Nevermind, it seems to sync back up. Probably just overall loadtimes and frames/milliseconds.

=Hands Hallway=
Darazanjoll: 8:26
Insane: 8:19
Insane had much cleaner movement here, which pretty much fully establishes a 1 second gain. 7 seconds ahead now.

=STARS Hallway=
D: 8:54
I: 8:45
You stopped here for safety. That's a 2 second loss.

=Library=
D: 9:35
I: 9:25
Insane had cleaner movement. He was one frame away from hitting the camera angle change before the panel slid down.

=2F Main Hall=
D: 9:51
I: 9:40

There's going to be full-second discrepancies between rooms. Every time I pause the video, due to the framerate discrepancy I mentioned earlier, the timing is not going to be accurate between rooms and you're going to see a gain and a loss here and there. If I leave it blank, then that's probably why, but if there's a gain, then it's probably from movement... Insane's movement is very very clean most places, so you're going to see why he gains seconds where he does. I'm only going to post times between groups of rooms now, and do writeups only for obvious errors... There will be two-second gains where those are.

=Art Storage Room=
D: 11:32
I: 11:20
Consistent. Neither of you use the Japanese version-exclusive Misty dodge to run behind her after getting the Valve Handle.

=Re-entering Save Room=
D: 12:00
I: 11:46
You hugged the wall and paused to clear the zombie. Insane just blew around it without hesitation. 14 second split.

=Before West Stairwell Hallway=
D: 12:57
I: 12:43

=Marvin=
D: 13:32
I: 13:16

You hit the wall and wasted movement while waiting for the other Misty to walk forward to run behind her. You also waited for the middle zombie in the evidence locker to hit the desk before executing the dodge. Insane didn't. He also exited the Marvin room faster.

=Ada Gunshot=
D:14:17
I: 14:01

16 seconds ahead now.

Carcinogen

2014-04-24 09:45:38 am

204 posts

Second Half of Police Station

=Gained control of Ada=
D: 18:56
I: 18:38
He got lucky with the spider, and it got out of his way.

=Exiting Parking Garage=
D: 21:11
I: 20:51
He started moving immediately when he got into the spider room. Poison be damned, I guess. Some fractions of seconds saved on movement and box pushing.

=Licker Dodge + Club Key=
D: 21:27
I: 21:06
You swerved around too much to try to get the licker to jump. Both you and Insane got attacked here. Insane's movement had fewer deviations though.

=Before East Office=
D: 21:50
I: 21:29
21 seconds ahead

=East Office=
D: 22:01 (148)
I: 21:37 (126)
You tanked a bite here. That cost you 3 seconds. 24 seconds ahead.

=Blue Hallway=
D: 22:13
I: 21:49

=After Gold Cog=
D: 23:03
I: 22:38
Movement (25)

=Red Hallway Exit=
D: 23:15
I: 22:48
Movement (27)

=Reenter Library=
D: 23:46
I: 23:17
You had an input flub-up and had to scroll more text at the ladder. He also held up-left and started running immediately at the licker, where you turned left before you started running. 29 seconds.

=Before G-Imago=
D: 27:18
I: 26:49
Consistent (29)

=G-Imago Fight=
D: 28:09
I: 27:37
He QSed the boss from standing position, you attempted the Jikkachan 6-shot position. Both of you took 7 shots, only he waited until after he was done shooting to shake off the embryos. You had to shake them off in the middle of the kill, interrupting your shooting. Now that I think about it, I'm not even entirely confirmed how that 6 shot trick works myself, but I'm pretty sure I saw Trevor Seguin do it in the old SDA run too... And he didn't use Quickshot. As evidenced in the TAS for RE1, there are some animation frames where bosses take critical hits, but I'm pretty sure bosses have different hitboxes that register different amounts of damage. I should test it with Dchaps' new Boss HP Tool when I get the chance. Either way, you lost 3 seconds with suboptimal execution of what could be an optimal strategy if you know the mechanics of it. He exits the Police Station 32 seconds ahead.

Second Half of Police Station

=Gained control of Ada=
D: 18:56
I: 18:38
He got lucky with the spider, and it got out of his way.

=Exiting Parking Garage=
D: 21:11
I: 20:51
He started moving immediately when he got into the spider room. Poison be damned, I guess. Some fractions of seconds saved on movement and box pushing.

=Licker Dodge + Club Key=
D: 21:27
I: 21:06
You swerved around too much to try to get the licker to jump. Both you and Insane got attacked here. Insane's movement had fewer deviations though.

=Before East Office=
D: 21:50
I: 21:29
21 seconds ahead

=East Office=
D: 22:01 (148)
I: 21:37 (126)
You tanked a bite here. That cost you 3 seconds. 24 seconds ahead.

=Blue Hallway=
D: 22:13
I: 21:49

=After Gold Cog=
D: 23:03
I: 22:38
Movement (25)

=Red Hallway Exit=
D: 23:15
I: 22:48
Movement (27)

=Reenter Library=
D: 23:46
I: 23:17
You had an input flub-up and had to scroll more text at the ladder. He also held up-left and started running immediately at the licker, where you turned left before you started running. 29 seconds.

=Before G-Imago=
D: 27:18
I: 26:49
Consistent (29)

=G-Imago Fight=
D: 28:09
I: 27:37
He QSed the boss from standing position, you attempted the Jikkachan 6-shot position. Both of you took 7 shots, only he waited until after he was done shooting to shake off the embryos. You had to shake them off in the middle of the kill, interrupting your shooting. Now that I think about it, I'm not even entirely confirmed how that 6 shot trick works myself, but I'm pretty sure I saw Trevor Seguin do it in the old SDA run too... And he didn't use Quickshot. As evidenced in the TAS for RE1, there are some animation frames where bosses take critical hits, but I'm pretty sure bosses have different hitboxes that register different amounts of damage. I should test it with Dchaps' new Boss HP Tool when I get the chance. Either way, you lost 3 seconds with suboptimal execution of what could be an optimal strategy if you know the mechanics of it. He exits the Police Station 32 seconds ahead.

Carcinogen

Edit history:

Carcinogen: 2014-04-24 11:02:50 am

2014-04-24 10:45:51 am

204 posts

=Enter Sewers=
D: 28:09
I: 27:37
32 seconds

=Enter Bridge Room=
D: 32:33
I: 32:04
You gained 3 seconds in movement here, somehow. The most obvious place I can think of where you gained the most time was unequipping the handgun upon controlling Ada... Insane did it in the water mid-run, so it took him more time to restart his movement. He may have missed a lag frame on the door, too.

A little secret by the way - the fastest place to do that is before the drop-down into the water, mid-run. Notice how you have to stop running in order to drop into the water? Entering the menu and unequipping the handgun as you hit the edge cancels that transition and gains a few frames so you can hit X and drop down quicker.

Also, this is the sewer. It's impossible to fuck up the sewer, so I'll fast-forward a little bit. Assume most errors are based on movement. I already did you a square there by telling you about that.

=Entering Cable Car=
D: 36:26
I: 35:54
32 seconds. You got an extra text box by hitting the power button for the cable car too many times.

=Claw=
D: 37:15
I: 36:42

You hesitated until you confirmed it was moving, when you can actually shoot it a frame after it's done coming through the hole. The claw plunging and the claw moving are two different animations, and the game is coded so the 3rd animation of the claw retracting doesn't execute until it takes a hit during the second animation. I usually position myself to make it come through one hole so I don't have to rotate or auto aim and I can hit it on the first frame it comes down. As long as you don't hit a camera transition, you won't lose any time by moving hole to hole as long as you hit it within a frame or two of the claw_moving animation.

=Factory Passage 1=
D: 37:31
I: 36:57
34 seconds. 3 second gain. This is the only other spot in Leon A where tanking a zombie hit is quicker.

=Door to G2=
D: 38:49
I: 38:12
You picked up an FAS here. 37 seconds ahead.

=G2 Fight=
D: 39:39
I: 38:58
Either run to a side, tank the damage and pop an FAS if you need it, or QS the boss. All that running around wastes so much time. I tell people this every time they think they're saving time by not taking damage. Insane messed up executing QSes here, tanked one hit, and got the bad kill, but still gained 4 seconds because he didn't spend so much time rotating Leon and running around. 41 seconds ahead.

=Enter Sewers=
D: 28:09
I: 27:37
32 seconds

=Enter Bridge Room=
D: 32:33
I: 32:04
You gained 3 seconds in movement here, somehow. The most obvious place I can think of where you gained the most time was unequipping the handgun upon controlling Ada... Insane did it in the water mid-run, so it took him more time to restart his movement. He may have missed a lag frame on the door, too.

A little secret by the way - the fastest place to do that is before the drop-down into the water, mid-run. Notice how you have to stop running in order to drop into the water? Entering the menu and unequipping the handgun as you hit the edge cancels that transition and gains a few frames so you can hit X and drop down quicker.

Also, this is the sewer. It's impossible to fuck up the sewer, so I'll fast-forward a little bit. Assume most errors are based on movement. I already did you a square there by telling you about that.

=Entering Cable Car=
D: 36:26
I: 35:54
32 seconds. You got an extra text box by hitting the power button for the cable car too many times.

=Claw=
D: 37:15
I: 36:42

You hesitated until you confirmed it was moving, when you can actually shoot it a frame after it's done coming through the hole. The claw plunging and the claw moving are two different animations, and the game is coded so the 3rd animation of the claw retracting doesn't execute until it takes a hit during the second animation. I usually position myself to make it come through one hole so I don't have to rotate or auto aim and I can hit it on the first frame it comes down. As long as you don't hit a camera transition, you won't lose any time by moving hole to hole as long as you hit it within a frame or two of the claw_moving animation.

=Factory Passage 1=
D: 37:31
I: 36:57
34 seconds. 3 second gain. This is the only other spot in Leon A where tanking a zombie hit is quicker.

=Door to G2=
D: 38:49
I: 38:12
You picked up an FAS here. 37 seconds ahead.

=G2 Fight=
D: 39:39
I: 38:58
Either run to a side, tank the damage and pop an FAS if you need it, or QS the boss. All that running around wastes so much time. I tell people this every time they think they're saving time by not taking damage. Insane messed up executing QSes here, tanked one hit, and got the bad kill, but still gained 4 seconds because he didn't spend so much time rotating Leon and running around. 41 seconds ahead.

Carcinogen

Edit history:

Carcinogen: 2014-04-24 01:05:24 pm

2014-04-24 01:03:54 pm

204 posts

=Entering Licker Hallway=
D: 42:47
I: 42:04
Movement and splits (43 seconds). One thing you definitely need to pay attention to is how he manages to get menus to work frames faster by being quicker on inputs. Menus don't seem like much, but they're actually a big deal, and Insane has better menus.

=Licker Hallway 1=
D: 43:00
I: 42:17
Insane gained a second here. How, exactly, I'm not entirely sure. Really interesting, considering your execution was good. You got through without taking damage, Insane got through with one hit. There's a possibility I mistimed it and he gained splits on loads (I have no way of timing ms in IGT), and the extra swerve caused you to waste time, but I'm actually not sure what's going on here. :|

=Entering P4 Lab=
D: 43:16
I: 42:33
Well, by now, it's definitively 43 seconds ahead. Probably just a weird timing anomaly then, because the no damage dodge seems to be equally fast as the dodge with damage.

=Entering Licker Hallway 2=
D: 43:54
I: 43:10

D: 44:09
I: 43:23
You swerved to try to get the licker to jump then took two hits. That's actually pretty random, because I just run straight forward and he might jump or he might not. I tried finding a pattern in the RNG by equipping or unequipping my magnum before entering the room. Although, watching Insane's run shows me that you could make it consistent by pausing after 5 steps, because I hear the clicking pattern that matches what a licker sounds like when it's rotating on its hind legs, checking for footsteps, and getting ready to jump... He does it again for the second licker, too. Interesting how audio cues help that much. This is something you wouldn't be able to gather from just watching, you would need some understanding of Licker mechanics, too.

Insane is 46 seconds ahead.

=Entering MO Disc Room=
D: 44:36
I: 43:50
Insane stopped a second to equip the Magnum. He lost 2 seconds by doing so. You need to do so when you're doing the fuse section. Either way, 44 seconds ahead now.

=MO Disc Room=
D: 44:59
I: 45:09
He regained 6 seconds by executing the MO Disc room perfectly here. You tanked hits and used an FAS. ~50 seconds ahead.

=Entering Licker Hallway 3 (with Zombies)=
D: 49:00
I: 48:11

=Licker Hallway 3=
D: 49:22
I: 48:31
He actually gained time here, I think, by shooting the first zombie in place, then optimizing his movement by auto aiming into the next one.

=Birkin G3/G4 Dead=
D: 50:25
I: 49:38
Insane ran into a case of bad positioning. He went into the menu to reload the Magnum, and G4 was jumping around a bunch. He lost 4 seconds here.

=Final Time=
D: 50:52
I: 50:05

So, it does add up to 47 seconds. All the extra hits you tanked and times you stopped couldn't possibly give you a time advantage when considering how often you had stopped or turned.

When you and Insane finished at exactly the same time in real time, the only conclusion is that your game and Insane's game run at different speeds. This is because you have different computer hardware. But the timer adjusts itself to compensate for those differences, because the engine runs processing cycles in its own environment, just like it's supposed to on console... You can actually see it in the final Birkin fight for instance, how the timer lags every so often when the game lags, because the engine is lagging. Pretty sure the facts have presented themselves, so the game timer is actually perfectly fine.

=Entering Licker Hallway=
D: 42:47
I: 42:04
Movement and splits (43 seconds). One thing you definitely need to pay attention to is how he manages to get menus to work frames faster by being quicker on inputs. Menus don't seem like much, but they're actually a big deal, and Insane has better menus.

=Licker Hallway 1=
D: 43:00
I: 42:17
Insane gained a second here. How, exactly, I'm not entirely sure. Really interesting, considering your execution was good. You got through without taking damage, Insane got through with one hit. There's a possibility I mistimed it and he gained splits on loads (I have no way of timing ms in IGT), and the extra swerve caused you to waste time, but I'm actually not sure what's going on here. :|

=Entering P4 Lab=
D: 43:16
I: 42:33
Well, by now, it's definitively 43 seconds ahead. Probably just a weird timing anomaly then, because the no damage dodge seems to be equally fast as the dodge with damage.

=Entering Licker Hallway 2=
D: 43:54
I: 43:10

D: 44:09
I: 43:23
You swerved to try to get the licker to jump then took two hits. That's actually pretty random, because I just run straight forward and he might jump or he might not. I tried finding a pattern in the RNG by equipping or unequipping my magnum before entering the room. Although, watching Insane's run shows me that you could make it consistent by pausing after 5 steps, because I hear the clicking pattern that matches what a licker sounds like when it's rotating on its hind legs, checking for footsteps, and getting ready to jump... He does it again for the second licker, too. Interesting how audio cues help that much. This is something you wouldn't be able to gather from just watching, you would need some understanding of Licker mechanics, too.

Insane is 46 seconds ahead.

=Entering MO Disc Room=
D: 44:36
I: 43:50
Insane stopped a second to equip the Magnum. He lost 2 seconds by doing so. You need to do so when you're doing the fuse section. Either way, 44 seconds ahead now.

=MO Disc Room=
D: 44:59
I: 45:09
He regained 6 seconds by executing the MO Disc room perfectly here. You tanked hits and used an FAS. ~50 seconds ahead.

=Entering Licker Hallway 3 (with Zombies)=
D: 49:00
I: 48:11

=Licker Hallway 3=
D: 49:22
I: 48:31
He actually gained time here, I think, by shooting the first zombie in place, then optimizing his movement by auto aiming into the next one.

=Birkin G3/G4 Dead=
D: 50:25
I: 49:38
Insane ran into a case of bad positioning. He went into the menu to reload the Magnum, and G4 was jumping around a bunch. He lost 4 seconds here.

=Final Time=
D: 50:52
I: 50:05

So, it does add up to 47 seconds. All the extra hits you tanked and times you stopped couldn't possibly give you a time advantage when considering how often you had stopped or turned.

When you and Insane finished at exactly the same time in real time, the only conclusion is that your game and Insane's game run at different speeds. This is because you have different computer hardware. But the timer adjusts itself to compensate for those differences, because the engine runs processing cycles in its own environment, just like it's supposed to on console... You can actually see it in the final Birkin fight for instance, how the timer lags every so often when the game lags, because the engine is lagging. Pretty sure the facts have presented themselves, so the game timer is actually perfectly fine.

tekkie

2014-04-24 07:12:03 pm

3 posts

SH3 (and I believe 2) in-game timers work that way.

Carcinogen

Edit history:

Carcinogen: 2014-04-24 07:26:11 pm

Carcinogen: 2014-04-24 07:23:21 pm

2014-04-24 07:18:34 pm

204 posts

Apparently Arcadia just made a video, educating people who don't know any better. All of my hard work being nice and explaining things, gone to shit, because FUCK CARCINOGEN, RIGHT GUYS!? RIGHT GUYS!?

If you use Real Time though, you're going to put someone who doesn't run the game with the same processor at a disadvantage. Fair warning.

Eleval

2014-04-24 07:29:42 pm

16 posts

Way to make everyone happy : Ask an experienced BH2 modder if he can check in the game' disassembled code how the timer behaves and we're settled.

I'm pretty convinced it works the way you've shown it Carci, timer being incremented each 30 frame.

C

2014-04-24 07:40:38 pm

Actually, that's a great idea.

zenix

2014-04-24 07:43:47 pm

22 posts

After reading some explanations by carc and checking the SH3 PC case it makes sense in many ways and its a good defense of the IGT over RTA (and it's true, it would be so unfair to those with the game lagging randomly because you know, PCs) Maybe someone could make RE2 lag a lot and test this like the SH3 video to show more proof?

Dchaps

2014-04-24 08:58:59 pm

74 posts

http://outbreakf12.blogspot.ca/

Quote from Carcinogen:

, because FUCK CARCINOGEN, RIGHT GUYS!? RIGHT GUYS!?

Yeah fuck you dictator fuck, how dare you try to give an indepth explanation on how things work.

Quote from Eleval:

Way to make everyone happy : Ask an experienced BH2 modder if he can check in the game' disassembled code how the timer behaves and we're settled.

I'm pretty convinced it works the way you've shown it Carci, timer being incremented each 30 frame.

Sadly even this wont do any justice because its apparent things like that being brought forefront are just plainly being ignored.
If people want to be tunnel visioned about how it works then just let them.

C

2014-04-25 02:37:09 am

Hey Dchaps, any chance you can ask Mortician what he knows about how the game processes timer functions? I would really like his feedback, and it seems like he would know being that he has so much knowledge of what's going on within the game engine.

insanebb

2014-04-25 06:58:16 am

2 posts

Hello to everybody. I'll tell you the whole story that happened to me at night.

1 of all - I wanna say special thanks to Carcinogen for his analys of Personal bests Leon A re2 PC (my and Dara'z) (You spended much of time for explaining all of things, BIG RESPECT TO YOU FROM ME, AGAIN! :P)

2 - I was really surprised, when i was kicked from channel, without any explain, when i asked , answer put me in shock, really.

3 - Darazanjoll , i want to say something you , personally - im COMPLETELY DISAGREE with your opinion by complaining my pb and you own, thats completely wrong example that you tried to attach in general problem (Carcinogen already explained that, and i can only add my opinion on this, but it can only harm you. Well as i pissed off atm , i will write this, and ofc you can hate me after this - you have no rides, to complaining your pb with my own, cuz ur best time is 50:52, my is 50:05, u can watch again whole my run, and see by yourself, that your skill little bit worse then mine, thats for shure, and because of that u still on 5-th place. True - wsplit time can be different then ingame time, cause of power PC, if your PC haven't much power to run in needed speed, your wsplit and game timer will differs a lot. But when you try to say , that your wsplit time and my wsplit time in pb are same, and ingame timer not, well in that case its just a question of skill, cause i did better in my own run, and you NOT! But if you will feel better - i can give you my own PB 50:05 to make you happy.

My advice, personall to you - Stop blaming the timer in your own failures. In Russia we have household term - Нечего на зеркало пенять, коли рожа крива (Needless to blame the mirror if your face curve.) I hope you have learned the lesson from what I said, and apologize to everyone, especially to the Carcinogen, buy new PC , and try to beat WR legally like me, Carc, Jake, Derek, Arc, ElyzSabo, and others who just trying and not crying.

P.S You can hate me after this threat, it is your right. But remember , thats only my opinion on this.

Sincerely yours insanebb.

Hello to everybody. I'll tell you the whole story that happened to me at night.

1 of all - I wanna say special thanks to Carcinogen for his analys of Personal bests Leon A re2 PC (my and Dara'z) (You spended much of time for explaining all of things, BIG RESPECT TO YOU FROM ME, AGAIN! :P)

2 - I was really surprised, when i was kicked from channel, without any explain, when i asked , answer put me in shock, really.

3 - Darazanjoll , i want to say something you , personally - im COMPLETELY DISAGREE with your opinion by complaining my pb and you own, thats completely wrong example that you tried to attach in general problem (Carcinogen already explained that, and i can only add my opinion on this, but it can only harm you. Well as i pissed off atm , i will write this, and ofc you can hate me after this - you have no rides, to complaining your pb with my own, cuz ur best time is 50:52, my is 50:05, u can watch again whole my run, and see by yourself, that your skill little bit worse then mine, thats for shure, and because of that u still on 5-th place. True - wsplit time can be different then ingame time, cause of power PC, if your PC haven't much power to run in needed speed, your wsplit and game timer will differs a lot. But when you try to say , that your wsplit time and my wsplit time in pb are same, and ingame timer not, well in that case its just a question of skill, cause i did better in my own run, and you NOT! But if you will feel better - i can give you my own PB 50:05 to make you happy.

My advice, personall to you - Stop blaming the timer in your own failures. In Russia we have household term - Нечего на зеркало пенять, коли рожа крива (Needless to blame the mirror if your face curve.) I hope you have learned the lesson from what I said, and apologize to everyone, especially to the Carcinogen, buy new PC , and try to beat WR legally like me, Carc, Jake, Derek, Arc, ElyzSabo, and others who just trying and not crying.

P.S You can hate me after this threat, it is your right. But remember , thats only my opinion on this.

Sincerely yours insanebb.

Carcinogen

Edit history:

Carcinogen: 2014-04-25 11:00:20 pm

Carcinogen: 2014-04-25 10:18:00 pm

Carcinogen: 2014-04-25 10:10:55 pm

Carcinogen: 2014-04-25 10:09:22 pm

2014-04-25 12:29:11 pm

204 posts

If anything, this whole thing was a learning experience... Audit every single mistake in your run before you blame the game timer.

It makes sense that the next step in confirming what's going on with this game timer is to play a speedrun with almost 100% the same conditions on the same computer and lag the crap out of the game to see what happens. Would be nice if I could figure out how to get that "PAL mode" running.

As for yesterday - that whole thing was pretty bullshit, but I guess that "divide" in the community needed to happen. I didn't even watch more than one minute of that video before skimming through it, so I have no idea what else has been said aside from the argument posted above...

I warned everyone, but anyone deciding to make RTA the standard without understanding another viewpoint is going to figure out eventually that they will not be able to compete with each other in Real Time if they audited their mistakes. Fuck it, let 'em.

Darazanjoll

2014-04-25 03:34:46 pm

2 posts

I am now convinced that the IGT is correct.

Carcinogen

2014-04-25 11:01:48 pm

204 posts

I'm adding columns for Real Time on the wiki for supplementary purposes.

Game Time is still going to be the primary metric on the wiki, but people who actually care about real time can post that too.

Eleval

2014-04-27 07:55:03 am

16 posts

Anyway nvm my previous posts. I'm going to run a few tests when I get the time to in order to check if the IGT is indeed tied to the framerate.