Posts for MUGG


Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Comparison vid account | free for those that want to see how I was doing in comparison to andymac. While working on Macro1, Pumpkin1(1), Pumpkin1(2) and Pumpkin1(3) I didn't directly compare with andymac's movie which is why I lost 2 frames in Pumpkin1(2) - but I gained some frames in (3). It's not well visually comparable due to the comparison video desyncing near the end (probably due to the problem I mentioned in the "Gameboy TAS times miscalculated" topic. Many frames get recorded in place of one lagframe, thus timings gets thrown off)...
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Dumping to h.264 lossless and then upscaling is very bad idea (unless you can somehow force h.264 into no chroma subsample mode). If you do that, you get pretty horrible artifacts.
Eh... From what I've seen, the results look very good to me and I haven't seen any artifacts. I've been using this method since forever. I think you are right about the artifacts, but that's a very negligible loss of information to me. Quality always turned out to be good enough for my purposes. Would there also be artifacts if I used lagarith?
Post subject: Re: New to Tas
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Rucion wrote:
Hi, I'm very new to making any sort of video. I'm doing a playthrough of Mega Man X. I did a recording with no editing to see how it works. It seems that after 4 mins it makes a new video file. The 4 minute file was almost 2 gigs in size. I'm using Snes9x right now. Could anyone please tell me how i could splice the videos together?
It was that big because you probably recorded it as uncompressed AVI. I use ffdshow (configured to h.264 lossless) for my videos. To improve quality for Youtube uploads, I halve the FPS (to 30) and then upscale the video using avisynth's pointresize.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
MrGrunz wrote:
This is the most brilliant thing I've ever seen. Mugg, you said, that you think there might be a quicker way through the OoB area. Would you mind giving me some useful bytes for this game and the position of the gliched block in the OoB area? I feel like messing around with it a bit.
EDIT: Reposted lua script on pastebin because the forum's code feature cuts off the code... Download this and use it with VBA23. A2D5 is located at x: 3406, y: 3844 level position. I hope this is useful.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
The pipe glitch only works in v1.0. This includes dying inside the pipe, and pausing and quitting while inside the pipe. However, pausing and quitting while touching an object (mushroom, starman, bubble item, bell...) to carry those items into another level works on all versions. Aside from pipe glitching, there are various other ways to enter the out-of-bounds area: 1.) Sprite-spawning glitch. Have an enemy that spawns objects (chibi dracula, fire-spitting piranha plant, pig) stand offscreen for a long time. They will spawn their stuff until all the object slots are used up. When having everything appear onscreen now, the game will glitch around - You can zip vertically through the ground by jumping, but this only seems to work with floaty state (address A287 controls this), so when you are in moon zone 1 or have the bubble item acquired. 2.) Tree Zone 3 (right). In this level you can go out of bounds by reaching the place above Mario, in the beginning of the stage, using the bubble item to fly up or a bee enemy to jump on. 3.) Pause glitch. When the game lags and you are pausing, the game may reset itself or place solid/breakable blocks somewhere on the screen randomly. Theoretically you could set up a spot of blocks that you can slide into, and get ejected into the out of bounds area. This is only theory, but maybe I'm gonna prove it someday. 4.) --- maybe there are more ways, but I can't recall them...
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
I only watched two addresses while making the run. A200 = Oscillation pattern A202 = Speed (in pixels) There are no subpixels in the game, which is why it uses oscillations in order to move at a speed of 1.5 pixels/frame (1,2,1,2,1,2). 1A00~onwards seem to be the possible oscillations the game uses. Aside from this, I didn't use any position-related addresses. I can try looking for them later, because I know there are some that are useful. FYI, the second half of the OoB area (4000~8fff) is switchable ROM. The area changes between two possible layouts by holding "up/down + b" (and not holding it), which is why there are so many possible ways through it.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
That would mean lagframes are not accounted for. Isn't that the wrong way to go? Times in submissions and publications would be inaccurate. Why aren't we using the console-based frame rates? Is it just a special case with VBA? Because Desmume does record AVIs using 59.826 fps, at least in 0.9.2~0.9.4. Also I added info about Dophin to my first post since you posted.
Post subject: Gameboy TAS times being miscalculated in subm./publications?
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
The SML2 TAS I submitted yesterday has 7740 frames on emulator (2 minutes 9.5885479887824 seconds), ~7978 frames in recorded AVI (2 minutes 13.5733121259052 seconds). This is because some lagframes throughout the TAS take longer than the duration of 1 normal frame, ~0.016 seconds. The recorded AVI accounts for that and records multiple frames in place of one. This also seems to be why TASes on Youtube end later than they are supposed to, time-wise. Yet, the submission time says 2:08.98. Why aren't lagframes accounted for in submissions and publications' times? Also, why isn't it 2:09.59? I used framerate 59.7275 (taken from here). 2:08.98 would mean the framerate used for calculation was 60. Dolphin currently has a related problem. Recorded AVIs don't account for lag (as far as I know, a recent revision fixed it for Direct3d11 (?), so recorded AVIs that use Direct3d11(?) add multiple frames in the place of 1 lagframe). Should future gamecube TASes have those miscalculated times too? It would mean that times would be many seconds lower than they actually are.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
...and no, I don't know why it got obsoleted.
Because the glitched run uses the pipe glitch just like the previous fullgame run (by andymac), just to a greater extent. Therefore it makes no sense to have seperate categories. At first I complained about this reason for obsoletion, but later accepted it. All that's left is either a "glitchless fullgame/all levels run" or a "no-pipe glitch fullgame/all levels run", the former makes more sense since the pipe glitch is actually a "touch an object, quit out of the level at the same time" glitch which there are several varations of, for example touching the bubble item or the bell (to skip half of Moon Zone 2). But I think both categories would be fine. Also, I just noticed I messed up while editing the VBMs, so the submitted VBM lists andymac as the author, yet the rerecord count is correct. I think this can be ignored.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
My first SML2 TAS using the new pixel trick! I didn't spend a lot of time doing this, so maybe I lost 1 or 2 pixels somewhere, maybe andymac can have a look at it. This is 19 seconds ahead of the published run. WIP (VBA23, 1.0 ROM)
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
I would be interested in seeing it, but I can't tell if it's gonna be an interesting watch.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
This is the best TAS ever.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
MUGG wrote:
Lost levels improvement 1.51, SMA+SMW
encoded so it gets more attention.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Toad King wrote:
Those backshifting problems look like polling issues with the input code. If you use the new TAS Input window, you shouldn't get any of those bugs anymore.
I already pm'ed this to you, but here goes: The TAS input window does not fix any of those input-loadstate desync bugs that occur in Mario sunshine and Tony hawk's underground 1. I just TASed a section in sunshine with TAS input and it desynced right at the beginning.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Obviously not. I saw many TASes on nicovideo for this game, by the way...
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
MUGG (in the Desmume development topic) wrote:
http://tasvideos.org/LuaScripting/Registers.html while true do- loops don't work in Desmume yet. Try other registers.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Sorry, I forgot about the rerecord count. http://dehacked.2y.net/microstorage.php/info/1162051984/spartanx%28J%29Corrected.vbm (J run) Please update the submitted vbm with this one and feel free to encode.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Brandon, didn't you see the messages I wrote to you? I was also interested in your opinion. I think we can safely go with the J version now, however, as many voted for it.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
Found a Bowser battle cutscene skip... almost. The game reset itself.
Editor, Experienced Forum User, Published Author, Expert player (2315)
Joined: 5/15/2007
Posts: 3856
Location: Germany
As long as there's something to click on. But I mainly want to interact with it with keyboard presses.