If you go in the "Options" screen before starting a New Game, IGT will roll as long as you are in the options screen.
So if you stay on the option screen for 10 seconds, you will start the game at 10 seconds IGT.
If you go in the options before starting a run, make sure to soft reset the game so the timer goes back to 0. (L1+R1+Start+Select)
The following load times count towards IGT:
Fire Escape Key behind bars (after selecting Yes) Investigating the hole near butterflies Elevator Game (Trick or Treat) Rope Pull Piano in Hotel Can of Light Bulbs (after skipping scene)
I might have missed some.
On PC, you can alleviate these time losses by triggering these scenes before starting a run. This will put them in the Windows file cache, which will decrease load times.
Just a test. I think it looks better, but emulator will be at the top. Please continue to submit runs in the Level Leaderboard.
New feature allows for any combination of sub categories (100% Kills, Tonfa/M93R Only, PE/No PE, etc.). Not many runners atm, but in case this game gains more attention and people want to run these categories.
We could keep the level leaderboard, since sub categories work there as well. It would keep Emulator, Console, and PSP separate, so emulator will not always be at the top in Full-Game.
Letters in the initial configuration of the briefcase never match the solution's word.
For example, if the initial configuration is "lvwn", then words starting with L are not possible solutions (love, lose, luck, lock). Words in which the 2nd letter is V are also not possible solutions (over), and so on... From "lvwn", you can find that (luck, lock, love, lose, over, down, town, damn, and open) are not possible solutions, deducing 9 of the 19 possible words are not the solution.
The initial configuration of the briefcase is randomly generated when you go to the 2nd floor hallway (the hallway just before you use the fish key), so you can test that this information is true by saving and loading just before it is generated.
Knowing this, an 'o' in the 2nd slot is best, since there are 6 words with 'o' as the 2nd letter. If the first letter is 'd', that will rule out 5 words. There are several letters that correspond to 4 words or less. 'x' is the only letter (as 2nd) corresponding to 0 words.
These are my analysis notes, which sorts the words 4 times alphabetically. Make sure to view with a fixed-width font, or the words will be mis-aligned.
Any thoughts or possible strategies?
I think setting a better mandatory emulator for runs would be a good thing for this game, such as psxjin or BIZHAWK.
Both of these have VSYNC counters which essentially act as an IGT. An emulated VSYNC represents a unit of game time way better than RTA, especially for emulation.
I wouldn't like to discount the current times set in place, but I think this would benefit future competition, so as to avoid having to use certain plugins that produce better RTAs (especially slight differences in FPS).
BIZHAWK is pretty slow for scoring times, and I hear some people have trouble running it at a good framerate. psxjin runs pretty fast at the cost of emulation accuracy. The current TAS uses it anyway.
Note this also means using the vsync counter to time runs more IGT-like; ((endframe - startframe) / 60) to get the seconds. A livesplit script could automate the frames to real-time conversion. Livesplit wouldn't be required, we can just playback the video and find the frame of hitting "New Game" and the last hit on final boss.
What do you think ?
The current times won't be removed, but future runs would have to use the specified emulator.
Hello. Autosplitter for Dolphin (64-bit 5.0) PogChamp Supports Leon & Claire scenarios; A & B Also supports Game Timer.
Works for version 5.0, but can be easily modified for other Dolphin versions with brain. It uses the same split and variable names as the PC autosplitter, so it is compatible with PC splits.
Make sure to use the latest livesplit! There are two settings that specify Claire and B scenarios.
DOWNLOAD: https://drive.google.com/open?id=0B0vjbgX1QQWTT0JYNUFZNmdxV3M
For "A" scenarios, the IGT ends as soon as the credits begin. For "B" scenarios, the IGT ends as soon as the true final FMV begins. These splitters will end the timer earlier, so you could add an extra split and do it manually.
If you want to modify it for your own version of Dolphin, I will give you a brief rundown, but not for newbies. Start the game (so you are at the first screen with zombies) then search for byte 200. Take damage and decrease until you find HP value. Subtract the address of the HP value by 0x1CEEBF and search for that address. (This gives you the beginning of GameCube memory) Find the static address that exists within the Dolphin.exe module. Calculate the address as an offset of "Dolphin.exe" Change "vars.DOLPHIN_BASE" to that offset in the asl script ( vars.DOLPHIN_BASE = 0xABCDEF ).
Many credits to Dchaps, StevenMayte and Darazanjoll for supplying a foundation for this script!
I feel there should be a standard frame timing method in place for closely tied WRs. "Timer STARTS at character selection and ENDS when you open the warehouse safe room door" isn't specific enough.
I would say that "the frame the text disappears on the character selection screen" to "the first frame that the timer freezes at the warehouse door" is more appropriate. Subtract these two frame counters to get the total frames spent in the run.
For example: Frame 300 - Text disappears in the character selection screen (you have selected this character) Frame 4000 - The timer value at the top is identical to the previous frame 4000 - 300 = 3700 frames. This would be the counter used for timing. Divide the frames by 30 to get the time in seconds.
What do you think?
Can I record the game with FRAPS and use the frames passed for timing? Why is on screen timer required anyway?
I've previously and jokingly (thought not really a joke) said that 320x240 will be faster by just a few microseconds, but this might prove otherwise that it is faster by frames/seconds.
I've figured since the game's native resolution is 320x240 that background stretching would be a bottleneck on the load times, especially in a resolution that isn't an integer scale (800x600, 1024x768, ...) However, there seems to be identical results even with 640x480 or 1280x960, so a floating point scale may be used in every case except 1.0 (320x240). I hypothesize that 320x200 or 160x120 will be slower than 320x240 because of down scaling.
You can argue that there are FPS dips on my machine during 60 FPS moments, but the screen transition test may prove otherwise.
I've tested all these cases at least 3 times except menu open. It seems opening the menu was faster in 800x600, but it is unknown if the controls have influenced the appearance of "Yes".
What do you think? Is it just my FPS dip or is the game really loading faster in 320x240?
