Posts for FractalFusion

1 2
31 32 33
128 129
Posted: 6/14/2015 6:02 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
For reference: MarI/O is a neural network/genetic algorithm program that completes a Super Mario World level, even when given nothing more than the location of objects within the level and how far Mario is to the right. Youtube video: https://www.youtube.com/watch?v=qv6UVOQ0F44
Posted: 6/14/2015 4:48 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
What codec ("compressor") are you using to dump AVI in BizHawk, and what settings?
Posted: 6/13/2015 7:47 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
I feel that both the shoryuken route TAS and the best ending no upgrades TAS are byproducts of ancient history and might as well be obsoleted by the publication of this newest TAS. Although, the best ending no upgrades TAS can be kept around until a 100% TAS is made, which achieves "best ending" by default. I mentioned before that the easter egg of Zero appearing on the highway only occurs if you beat the game without any Light capsule upgrades (as in best ending no upgrades). Apparently this is false, because the easter egg occurs in this TAS as well, even though you get X-Buster upgrade. (Maybe you have to beat it without any heart/sub tanks?) So yeah, this TAS almost completely outclasses the best ending no upgrades TAS; the only thing that best ending no upgrades does that this TAS doesn't and that is actually better is the increased usage of Strike Chain and the "best ending" along with the extra fights along its route (neither of which are that much better) and pretty much nothing else. By the way, "best ending" (all Zero parts) really isn't that much; all it does is you don't fight Zero near the end, and the very end "Presented by Capcom" has a correctly-colored Zero instead of a black Zero. Edit: As for tier, I think this TAS should get a star. If the shoryuken route TAS is not obsoleted, then take away its star.
Posted: 6/13/2015 1:41 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
I use Avisynth+VirtualDub to edit video pre-encode (VirtualDub is pretty much just a gateway; there may be better options but that's what I'm used to). Video encoding is done using x264.exe by command-line/batch file and audio encoding with something else (usually neroaacenc or whatever it's called). I use Avidemux if I need to trim or split the video post-encode; there are options to copy input stream (since there is no point re-encoding obviously) but you can only split or set start range on a keyframe (I-frame).
Posted: 6/5/2015 3:35 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
Note: The Youtube video is to be watched at 720p60 mode which has full 60fps; the other modes are bad because they are 30fps. Some more encodes: Downloadable encode at MediaFire(~98.9MB): http://www.mediafire.com/watch/x339i2ijk7vzk49/mmx2_hetfield90.mp4 Nicovideo: (account) Part 1: http://www.nicovideo.jp/watch/sm26417384 Part 2: http://www.nicovideo.jp/watch/sm26417342 (no account) Part 1: http://www.nicozon.net/watch/sm26417384 Part 2: http://www.nicozon.net/watch/sm26417342
Posted: 6/3/2015 7:03 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
In general, the Japanese term for a video game glitch is バグ (bagu). Note: Although a バグ tag exists, not everyone uses it so don't search バグ by tag.
Posted: 5/27/2015 1:56 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
For reference: This TAS is 196 frames faster. However, the actual gameplay improvement is 26 frames. 170 of the frames saved are merely from turning off the music option. This TAS uses a different layout than the one used in the published TAS. By the way, the improvement posted by the author makes it 231 frames faster (gameplay improvement is 61 frames). I don't agree with turning off the music option, but a gameplay improvement is an improvement. I was hoping though that the author spent more time with making this (like searching over many possible layouts and optimizing them).
Posted: 5/24/2015 12:21 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
Hm. Seems that gui.DrawNew("native") is rather interesting. (gui.DrawNew("emu") is default, so it doesn't do anything new). Apparently, gui.DrawNew("native") not only can draw in the black space around the emulated system's screen, but all drawings persist there even if the script advances a frame or is turned off. The only way I know to clear the drawings is to call gui.DrawNew("native") each time. Also, frame advances automatically switch the mode back to "emu". So the question of being able to draw in the black space is pretty much solved now. By the way, there are a few extra gui methods that don't even appear in the documentation. They are: gui.defaultForeground(color), which sets the default drawing color for drawing methods, gui.defaultBackground(color), which sets the default fill color for drawing methods, and gui.createcanvas(width,height), which, despite being in the code, doesn't even work.
Posted: 5/21/2015 6:15 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
I noticed this just today. The way that BizHawk compresses information related to a movie file when creating BK2 appears to have changed between BizHawk 1.9.1 and BizHawk 1.9.2. For BK2s created in BizHawk 1.9.1, input is compressed. For BK2s created in BizHawk 1.9.2+, input is not compressed, as can be seen by opening the BK2 directly in a text editor. Apart from making files larger, it also creates the impression that the BK2 can be edited directly in a text editor without any problem, which is false. There is no issue with playing back BK2s as far as I know. Also reported here: https://code.google.com/p/bizhawk/issues/detail?id=426
Posted: 5/20/2015 2:21 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
Pokota wrote:
I'm looking at the lua functions, and there's gui.DrawNew(string name). What sort of shenanigans can I do with this?
I've tried it before when testing functions. I don't even know what it is supposed to do, and it has never worked for me.
Posted: 5/15/2015 11:22 PM
(Edited: 5/16/2015 5:17 PM)
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
ALAKTORN wrote:
Does anyone know how to reverse this formula to find x? y = ax+b (mod M) I was told x = a-1(y-b) (mod M) but it’s not working…
Let me guess: You're trying to find the inverse of X(n+1) = 0x343FD * X(n) + 0x269EC3 (mod 2^31). The reverse formula is X(n) = 0x39B33155 * X(n+1) + 0x2170F641 (mod 2^31). In general, the post that Masterjun linked is a possible way to work out the inverse. Note that if you need to calculate multiplicative inverse (a-1 (mod M)), you can use this applet. By the way, a-1 (mod M) doesn't exist unless gcd(a,M)=1. If gcd(a,M)>1, then f(x) = ax+b (mod M) doesn't have an inverse function, since multiple x's give the same f(x).
Posted: 5/15/2015 7:27 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
I had a look at the TAS. I definitely like the idea of a "glitchless" run of some kind but I agree that the pause abuse in Elec Man's stage (and I think elsewhere) was a problem. I didn't even notice the first time because I was so used to former Mega Man TASes opening the menu all the time. It appears that, even though there is a weapon switch button in the PSX version, you can't use this button to switch to Magnet Beam. Which means that, unfortunately, in a run that doesn't use glitches or pause tricks, pausing is required anyway. Of course, one could just have conditions on pausing, such as never pausing when Mega Man is in damage state, or in any boss battle. While using the PSX version allows for weapon switching without going to the menu, I feel that this TAS never really took advantage of it enough. (Compare to TASes of other Mega Man games which have a weapon-switching button.) There were plenty of opportunities to show off this feature but it was never used much anyway; it doesn't help that MM1 without glitches seems rather dull compared to any other console Mega Man game (even if their glitches were never used either). That's just my feeling.
Posted: 5/11/2015 8:50 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
ALAKTORN wrote:
Done that but I don’t like my implementation. Mainly the fact that I have to use a bitwise operator. 
Language: Lua
function getThirdByte(iRNG)
	iRNG = RNGSteps(iRNG, 2)
	return bit.band(iRNG, 0xFF0000)
end
a = readRNG()
function findItem()
c = 0x840000
b = readRNG()
	thirdbyte = getThirdByte(b)
	if thirdbyte == c then
		print(a .. " to " .. b .. " is " .. getNumSteps(a, b, 10000) .. " steps")
	end
end
Instead of bit.band, maybe use math.floor(iRNG/0x10000)%0x100. This gets the third byte as an actual byte, and not as an 8-bit number times 0x10000. Then you write c=0x84 instead of c=0x840000. By the way, there is nothing inherently wrong about bitwise operators. I'd use them if it is convenient. Only thing is, bitwise operators should never be used on numbers that have more than 32 bits (2^32 or greater).
Posted: 5/11/2015 3:26 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
Here is an encode: http://www.mediafire.com/watch/cr9n1u15jco8989/rockmanpsx_TAS.mp4
Posted: 5/5/2015 5:55 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
It turns out that it is possible to pass through the crystal (second-last one) on BizHawk without dying, as shown in this BizHawk movie. To do this glitch: - Shoot pink buster while dashing. This keeps X's dash hitbox. Then jump and then keep walking around to preserve the glitched hitbox. - You can then walk in the space under the crystal using the glitched hitbox to delay zipping as long as possible. Then, when you are just about to zip to the right, hold dash-jump and left and, if positioned correctly, you should be able to go through the crystal without dying. Snes9x and BizHawk differ on the crystal physics making it harder to do the trick on BizHawk. But at least it is possible on BizHawk, and it has been confirmed on both Virtual Console and SNES. The trick can be seen on Twitch.
Posted: 4/30/2015 6:41 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
This is what Flip is referring to. (Note: The numbers on the cards are 25, 50, 75, 100, and two each of 1-10.)
Posted: 4/25/2015 6:13 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
Hm, Morph Moth's first form seems to generate a lot of lag. Is it possible to manipulate to reduce it as much as possible? By the way, Hetfield90 told me that the glitch of going through the crystal in Crystal Snail's stage is an emulation error by snes9x (all versions). This glitch does not work in BizHawk or console.
Posted: 4/14/2015 4:24 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
ALAKTORN wrote:
Sadly wasn’t able to manipulate any life up drops… I wish I knew how drops work in this game.
Is there any currently-known basis for how the RNG is used by the game? My guess is that since it is a LCG, the higher half (upper 16 bits of a 32-bit RNG) is used. Your script seems to indicate that the RNG is 31 bits instead (% 2147483648). Shouldn't it be 32 (% 4294967296)?
Posted: 4/9/2015 7:45 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
Glitcher wrote:
Nicovideo doesn't allow tracking. Can't someone put this on YouTube instead?
Unfortunately, Youtube does some bad things to uploaded movies (especially those involving video game graphics). For example, it reduces video framerate to 30fps, which is a very bad thing to happen to MMX (the flickering problem). The site encoders are going to make a quality Youtube encode anyway which rectifies the problem, so just use that when this movie is published.
Posted: 4/8/2015 9:41 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
Downloadable encode (~72MB) with aspect ratio 4:3: https://www.dropbox.com/s/m64r65hiq6enhpq/tas_megamanx_any%25_hetfield90.mp4?dl=0 Mirror (which apparently doesn't work): http://www.mediafire.com/watch/w3xo0e3f8w0osv5/tas_megamanx_any%_hetfield90.mp4 Also on Nicovideo: (account) Part 1: http://www.nicovideo.jp/watch/sm25986991 Part 2: http://www.nicovideo.jp/watch/sm25986953 (no account) Part 1: http://www.nicozon.net/watch/sm25986991 Part 2: http://www.nicozon.net/watch/sm25986953 Edit: By the way, Hetfield90 informed me that the rerecord count is unreliable. It is not as high as 69k nor as low as 13k (69k-56k, where 56k is the rerecord count for the 100% TAS file that he used to branch off into this one). He thinks that it should be somewhere in between, but he doesn't know what it is.
Posted: 4/7/2015 5:50 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
In GDI+ mode: - Load a ROM. - Open Lua Console. - Load a script such as ButtonCount.lua. - Unpause emulation. Then BizHawk crashes with the following error: http://i19.photobucket.com/albums/b185/jhchan8/gdiluaerror.png Crashes occur in both 1.9.3 and 1.9.4. OS is Windows 7 Home Premium (older computer that couldn't run BizHawk without GDI+). Also reported here: https://code.google.com/p/bizhawk/issues/detail?id=417
Posted: 4/6/2015 7:02 PM
Post subject: BizHawk 1.9.4 Released!
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
BizHawk 1.9.4 has been released! Numerous bug fixes and new features since 1.9.2. For more details, see http://tasvideos.org/BizHawk.html.
Posted: 4/6/2015 6:49 PM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
Link to BizHawk 1.9.4 download.
Posted: 4/5/2015 2:38 AM
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
boct1584 wrote:
So executing c8=Q, then taking it back allows the pawn (now back at c7) to capture horizontally, forcing the mate after Qb7? Am I getting that right?
Yeah, pretty much.
Posted: 4/4/2015 8:32 PM
(Edited: 4/4/2015 9:47 PM)
Quote
FractalFusion
Editor, Experienced Forum User, Published Author, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
Glitchy TAS of 2013 Gameboy TAS of 2013 PSX TASer of 2010
For those who want an encode: Link to video
goldenband wrote:
Chess problem composers love this stuff -- "White takes back his last move and mates in 1" is a pretty standard setup.
Yeah, but this AFD joke relies entirely on a Battle Chess bug. Without knowledge of the bug, it really isn't funny in any way. I found it funny because I already knew what the bug was before I even saw the movie (because I had played around with it before and even made a wiki page). I can't say the same for anyone else though.
1 2
31 32 33
128 129