Posts for FractalFusion

Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
Translation of asutoro's post: ---- On "Dolphin[desync-fixes]3.0-382-dirty_x86", Mario Galaxy 2 has many desyncs. As well, there are many savestate/loadstate errors. Since it happens 1/10 of the time, making a TAS for Wii games is very much difficult. I hope that, with savestates being improved on "Dolphin_dtm-config_x86", I can (eventually) make TASes on Wii.
asutoro wrote:
Dolphin[desync-fixes]3.0-382-dirty_x86の マリオギャラクシー2でdesyncが多発しました。 またステートセーブやステートロードによるエラー発生が多いです。1/10の確率でエラーが発生し、WiiでのゲームによるTASはほとんどが困難です。 私はステートが改善されているDolphin_dtm-config_x86の方で、WiiもTASが作れることを望みます。 (Sorry,i cannot speak English.)
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
I had a look at this run. The branch as stated ("computer repeated jail sentences") is vague. The run does not have any specific goals that are impressive, just "TAS and see what happens and at least I can get the CPU to hit the go-to-jail square every so often". That seems way too easy to do, especially since there are a million ways that it can play out. Why not get the CPU to go to jail with three doubles? I know it requires advanced luck-manipulation but I am putting forth that possibility. Why not deny the CPU any properties so it can't mortgage its way out? Why not bankrupt the CPU on the jail fine? Or, if not that, why not on a card causing loss of money? The run just bankrupts the CPU on house rent, nothing which other runs don't do. A playaround TAS of something like Monopoly needs innovative goals, with results that occur from diligent planning. Just doing "TAS and see what happens" isn't enough. JWinslow23, you have been submitting too many subpar TASes and no form of constructive discussion has taken place. Do not submit any more movies for the next week or else we will have to take away your submission privileges. You have the opportunity to discuss any game in the relevant game discussion thread so there is no need to make any submission for the next little while.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
adelikat wrote:
... let's continue with some truths: ... ● Is awesome ● TAS of the year ● Is worth a minute of your time
In my opinion, those are lies. Sorry. I don't see why this should be published. Next to nothing happens in this TAS (granted, there are TASes which are even closer to nothing, but that's not saying much). The wait at the end to call home is brutal. I believe there are better games to TAS on the Atari 2600 than this. Also, the amount of self-deprecation in the submission text is ridiculous. It's like trying to convince others that a TAS is good by saying it is terrible in the hopes that one sees it as parody and concludes the opposite.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
Link to video Mediafire download (<2MB): http://www.mediafire.com/download.php?h9iq6yupa3bvg0a The recent monopoly discussion reminded me of this TAS I made six years ago. There are faster runs that abuse trading and other stuff, but this one is funny because the CPU goes bankrupt one minute after end of input. The CPU even buys houses and hotels!
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
feos wrote:
FractalFusion wrote:
This run is the NES Battletoads TAS that never was.
What does it mean?
I was just joking. I meant that it resembles NES Battletoads, except that this run is far less broken so it shows off a lot more of the game.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
This run is the NES Battletoads TAS that never was. There are a few parts which aren't very interesting to me, but since this run beats the previous, then of course it should be accepted as an improvement. The most interesting part for me was Stage 2, where everything (other than the two players) is killed very fast. Some of them are killed multiple times. Also, the first time I reached the battle with the Dark Queen, I had no clue what happened until I read the submission text.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
Here's a 60fps encode for download, in case anyone wants it: Mediafire (~100MB): http://www.mediafire.com/download.php?5q00d522o1884j1 (Note to encoders: The auto-twitch that the author uses during boss battles is affected if shown in 30fps. As far as I know, 30fps effects don't appear in any other part of the game.) I found this run interesting. I don't know why everyone (including TASeditor himself, it seems) says that the tutorial levels are boring. Those levels are a good indicator of the difference between how the game is supposed to be played, and how much the game can be broken. It sets up the rest of the run easily. There isn't a lot of information in the submission text, but anyway here's how the run goes: - Tutorial levels. - Chapter 1 story mode. - Challenge mode for tutorial levels and 1-1 to unlock chapter 2. - Chapter 2 story mode. - Challenge mode for 1-2 to 2-4 to unlock Chapter 3. - Chapter 3 story mode. - Challenge mode for 1-1 and 3-1 to 3-3 to unlock Chapter 4. - Chapter 4 story mode. Challenge mode is basically achieving challenges in levels to unlock the next chapter(s). Each level has 3 or 4 challenges. Some levels have to be done more than once to achieve all its challenges. Though, doing challenge mode for level 2-5 (which was not done in the run) may be faster than repeating levels for the 50 acorns challenge (tutorial-3 and 4, 1-1 and 3-1, if I remember correctly) but TASeditor would know better than I do. All the boss-level challenges are survive challenges and are out of the question. The style of which the TAS speeds through the run is game-breakingly fast. It's unfortunate that the boss fights are very long (the "final boss" fight is almost 4 minutes long) and are completely contrary to the selling point of this TAS. Most of the run is great though, so I find this run acceptable regardless of anything that happened in the final 4 minutes.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
turska: It's great that you got it to sync. Though I noticed your encodes have these horizontal transparent lines in the clouds in Luste's stage, which I think are not supposed to be there. Anyway, I like this run, though I prefer that the amount of time saved over this Grolla run be given somewhere. Guessing from the end times of the two uploads, it is about 2 minutes.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
I noticed a bug (tested in both BizSNES and BizGB) where, after pressing a controller button, sometimes it remains held down even when I release it. Then again, BizSNES runs 10-15fps for me (half lsnes speed) and BizGB also 10-15fps. So it is probably my fault and not any fault of the emulator.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
Sorry for the delay. After watching it carefully again and referring to this map, I have to say it is well done. You even found a secret Konamiman door that was not listed in the map (where 36 is on the map, ~16:17 in your encode). I think though that two of the door visits (~5:10 and ~5:28 in your encode) could have been switched. It seems that enough people like this and are vocal about it, so I will accept the run.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
LYF, do not spam the forum. Thread locked.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
hegyak wrote:
Has anyone bothered, to actually watch the run? I can't get it to work. WTF at Version 2.10?
I've watched the run. Use version 2.1.4a or later with New PPU.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
devanwolf: I watched the movie, and I saw nothing interesting. I suggest doing a game that is not as ridiculously easy to optimize as any of the Fisher Price games. You can try a NES or SNES platform game that hasn't been done yet.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
It was uploaded just a while ago. August 2012 TAS Ranking: Link to video The entry barrier for 30th place for August 2012 is over 30K. A sign of growing activity again.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
Probably due to all the No votes. I was worried that having a poll for this submission would encourage users to come in, vote no, and then make a bad post to justify voting no. I think now that perhaps it doesn't solve anything to delete the poll, but I am unable to put back the poll due to a site glitch. A submission poll is not the be-all and end-all for judging on this site, so it is not a huge loss. (There were 7 no-votes in the poll and none of any others when it was deleted). By the way, it was the Lester the Unlikely poll that I deleted by accident. Edit: Ilari restored the poll.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
I watched the run (it requires New PPU to sync). While this run is accurately done for this type of game, it seems too easy to optimize, since it is merely a concentration-type game. Gameplay is also very slow. A TAS of a concentration-type game would need the gameplay to go very fast (as in, way beyond human speed) in order to merit publishing on this site, something that very much depends on the game. devanwolf, this run is a good start to practice TASing, and I hope you try a TAS for some other game.
CoolKirby wrote:
What...happened to the poll? I don't see it anymore.
I deleted it.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
Sorry, I accidentally deleted the poll. I have set a new one up again. Feel free to vote whatever you voted before. By the way, there is a run of this game on nicovideo. However, it is both slower and suffers from A/V problems: Account: http://www.nicovideo.jp/watch/sm7919203 No account: http://www.nicozon.net/watch/sm7919203
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
I made an encode which displays the game over screen correctly: Mediafire download (~3.5MB): http://www.mediafire.com/download.php?a2mh542054vp3fv Youtube: http://www.youtube.com/watch?v=0MVv21pokzM Nicovideo (account): http://www.nicovideo.jp/watch/sm18803635 Nicovideo (no account): http://www.nicozon.net/watch/sm18803635 Youtube video uses TASBlend, but at strength 0.75 instead of 2.0/3 .
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
N. Harmonik wrote:
Paused wrote:
/ad bot
...What?
Paused is just saying that he is an ad bot. You can ignore him. I've looked at a gameplay video for this game, and I wonder how long a TAS of this game will be. I'm not a fan of Advance Wars-type games but there should be a number of people who enjoy this type of TAS.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
This game looks really slow. If there is no way to speed up Rayman, then I'd say it is a bad game choice.
got4n wrote:
i Aim More faster possible but i try to take the more lums possible.
That is a conflict of goals. Just focus on being as fast as possible. Do not waste time to take lums. A 100% goal may be used (such as collecting all the lums) but to do that, you must collect all the lums, no less.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
I don't think this 100% run adds much to the site. The only thing different is that it is a lot longer, and that the jumping shoes item is used. Not much else.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
McBobX wrote:
How I can put the subtitle,logo and the video together
That has to be done before the x264 encode. There is a way to edit x264-encoded videos using Avidemux, but it is much better to do all video editing beforehand. I remember you asked if encoding process of x264 making the system slow is normal or not. The process uses a lot of CPU power, so for me it is rather slow. I am using a slow laptop, by the way.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
I never thought a TAS of this game would be this fast. Good job, Cooljay, and yes vote for carnage everywhere.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
Kaphotics kindly sent me a link to a smogon.com post that details the pokerus function in Gen 3: http://www.smogon.com/forums/showpost.php?p=4376835&postcount=959 I managed to make a movie catching Pokerus in Pokemon Ruby: http://dehacked.2y.net/microstorage.php/info/1944550717/0907%20-%20pokerus.vbm Pokerus is determined at the end of battle, after the screen fades to black, but before the overworld appears. It appears to be determined regardless of how the battle ended (so run away or white out works). There is a frame with 4 RNG cycles; the 3rd value determines Pokerus. Let's say the RNG value is the whole 4-byte number xxxxyyyy (hex digits). Then if xxxx equals 0x4000, 0x8000, or 0xC000, then a random Pokemon in the party is infected with Pokerus. This is a 3/65536 chance. By the way, the disassembly in the smogon link above is for Emerald, with the Pokerus routine beginning at 0x806DCB4. The same Pokerus routine in Ruby/Sapphire begins at 0x8040048.
Editor, Experienced Forum User, Published Author, Expert player (2159)
Joined: 6/15/2005
Posts: 3303
Sorry, I forgot. Here is the line I used for the x264 build I use: start /wait x264 --crf 20 --keyint 600 --ref 16 --no-fast-pskip --bframes 16 --b-adapt 2 --direct auto --me umh --merange 64 --subme 10 --trellis 2 --partitions all --fullrange on --rc-lookahead 250 --no-dct-decimate -o video.mp4 in.avi Note "--crf 20" is responsible for video quality and file size. More information here: http://tasvideos.org/EncodingGuide/Encoding.html
McBobX wrote:
FractalFusion,if I use password in a 100% run,this run should be accepted?
I can't say for sure, but if you use password, I think a lot of people will dislike it.