1 2
6 7 8 9
Warepire
He/Him
Editor
Joined: 3/2/2010
Posts: 2178
Location: A little to the left of nowhere (Sweden)
Dang it :(
Joined: 6/22/2012
Posts: 81
Sorry reborn this discussion, but searching Ram the address 7E00F3 represents the value from lottery. For the win I get Value = 1 Previous = 20 Changes =96
Player (206)
Joined: 2/18/2005
Posts: 1451
So I finally came around to finish a long time considered TAS for this game, using the original English PAL version: Link to video It's an any% run, breaking the 3 hour mark with a 2:56 ingame time at Dark Gaia's defeat, which is universally comparable to any version of the game except of text delay differences. Recorded on Snes9x 1.43. Enjoy!
See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2823
Location: Northern California
Saturn wrote:
Enjoy!
I'd enjoy an input file far more than a Youtube movie.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
Joined: 5/8/2010
Posts: 177
Location: Entropy
I'm glad to see Terranigma is still being worked on, I never got to play it on the SNES back then but loved it when I found this hidden gem on a rom site. Saturn run seemed acceptable tho is there a reason why the intro isn't being quickly skipped?
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2823
Location: Northern California
It's absolutely not acceptable. It was made on Snes9x 1.43, which is horribly inaccurate and not accepted on the site anymore. The intro was left in. No credit was given to anyone in this thread, even though he uses tricks discovered and talked about here. It's done on the (E) version, which is MUCH slower than the (J) version due to PAL timing and slower text. Lemon's (J) test run earlier in the thread is actually faster than this by... Offhand, I wanna say 10 minutes or so? And Lemon wasn't even using the diagonal boost movement trick. An optimized (J) run would be around 3 hours real time, not just "a few minutes" shorter like Saturn said in the comments on the Youtube vid. And the worst part is Saturn's acting like it's a huge favor to the community (typical of gigantic narcissists like him), which it isn't. Nothing can be gleaned from it: It's just a shitty 30fps Youtube video. There's no input file attached, nothing we can study for a full run, no list of memory addresses or any notes on the run at all. He literally just came in and said "Hey, look at how awesome I am! I and I alone did this run everyone wanted to see! Praise me!" A good run of this game will happen someday, don't worry.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
ALAKTORN
He/Him
Former player
Joined: 10/19/2009
Posts: 2527
Location: Italy
^Good post.
Joined: 8/28/2012
Posts: 11
Location: Germany
Allright I found out a way to get OOB in Terranigma Link to video Whenever you enter a stair your sprite will be tempoary loaded under the regular sprites. If you manage to stop Arks movement while he is going up the stairs he will be stuck below the sprites untill you leave the area or enter another room. You are able to completly move arround without anything stopping you. Looks like the glitch from ALTTP. Maybe you guys can find a way to get something out of it. Sorry for the lazy edited video, I was just doing it with OBS to get it done quick
wzl
Joined: 5/20/2015
Posts: 1
Gunty wrote:
These are some of the Lua scripts: terranigma_collision_map.lua terranigma_exp_timer.lua terranigma_dmg_graph.lua
Thanks for these. I took the freedom and modified the collision script somewhat, i hope you don't mind. Used it in order to get more familiar with the scripting and i refined the info output somewhat to yield more useful information. Also i made it a bit more readable and translucent so you can have it running without losing too much focus on the game I added a tile id legend to the script file although i dont grasp what some tiles are for. Legend: '?' => 0x18 mostly walls, but can be chests or pots Red => Blocked off completely Blue => Water Lightblue => Passable LightGreen => Highground - as far as i can tell while on highground you cannot attack or be attacked from low ground and vice versa, there are some places where those tiles are not seperated by a wall so that might proof useful someplace Interactible terrain Made seperate visualizations for vines, claw walls, pushable objects (which were previously ignored), and staircases on which you cannot start to run using Y I might have forgotten some more, let me know if you find something. terranigma_collision_map_mod.lua Hope this is useful to you. Further more i'd like to find exit spots and trigger locations and render them, but as of yet i have no clue how to actually find them. Gotta do some research and ram watching i suppose.
Player (199)
Joined: 12/3/2006
Posts: 151
Nice additions :) I dug up some more scripts I made a while ago to help analyse/predict the RNG: terranigma_rng.lua terranigma_rng_parasite.lua terranigma_rng_crits.lua The first script is a generic script that shows a table of future random numbers that the RNG will generate. The second script predicts which random numbers will make the Parasite boss spawn its spikes. The last script predict which random numbers will result in a critical hit, although I haven't tested this one too much.
Post subject: Here we go again...
Player (72)
Joined: 5/28/2013
Posts: 99
https://goo.gl/eIgh7P So that's the current .tasproj file I've been working on for this game. It's been kind of kept off of here for a while because I wasn't sure if I would really get into it or not like I did with Secret of Evermore forever ago. That being said, thanks to the people in my discord channel for Terranigma's TAS I've been going at it pretty regularly throughout the past month or so. Currently we're in Tower 2, but there are some very obvious differences between what we're working on what Saturn's provided in his TAS. His is most definitely a more smooth looking movie during specific situations. The most notable so far has been the Tower 1 fight at the end with the 4 red huballs in which he attack 5 times and gets a god-like RNG set to work with. I spent an absurd amount of time trying to replicate this, going so far as adjusting the optimized routes we came up with through Tower 1 to alter the RNG set we got to work with. Every single time I did this (all the way back to having 10+ frames lost to adjustments) I would end up with 1 damage short of an enemy that could not be manipulated into a favorable location, both of which shut that scenario down hard. Another smaller example is in Tower 2, where I double slicer the 2 blue huballs. This was faster than the adjustments i'd have to make prior to that room (or during it) to enable a single slicer attack like in Saturn's TAS. Given these things, I don't think Saturn's TAS is necessarily "better" or "worse" currently, because it's very hard to judge from my level of understanding of the game itself (I've never played the game before, just watched speedruns of it), but it's something to keep in mind if you look at the project thus far. I highly encourage anyone to play with the project file itself and see if they can get something quicker in Tower 1 or where-ever. I WILL go back and change it. All that being said, I'm super happy with the movement in the TAS so far. It feels optimized, good, and fluid (though i'm sure some can be shaved off in Tower 2 with a bit more refinement/2nd pair of eyes looking at it). I think the general route through everything is more or less going to be the same as Saturn's is, because it's kind of pointless to detour away from that route unless we find something broken somewhere. Info for people who might be interested in joining our discussions/helping optimize things: Discord channel: https://discord.gg/KDhsRYj Project File: https://goo.gl/eIgh7P Current RTA WR: https://www.twitch.tv/crawek/v/57425069 wzl's speedrunning guide: http://terranigma.wzl.gd/ ^ might be outdated? Not really sure personally... Saturn's TAS: https://www.youtube.com/watch?v=imK_BtsHncQ Lemon's tricks and stuff: https://youtu.be/yiM9NmQvH2M Vaan's notes: http://pastebin.com/4Wb3EZXW That's basically my info dump, most of which is already in this thread. I'm working on slowly compiling a gdoc that's on par with my Secret of Evermore one, for anyone familiar with that. It will be a while though. Hope to hear from you guys!
My Stream: http://www.twitch.tv/theangrypanda1 Mostly SoE TAS and casual playthroughs.
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2823
Location: Northern California
Like I said on Discord, the differences between Saturn's TAS and this one could easily be chalked up to emulation differences. Looking back, he made his TAS on a completely deprecated and unacceptable version of Snes9x due to how inaccurate and garbage it is (I thought it was at least 1.51, but nope, 1.43). No matter what happens in terms of losses compared to Saturn, his run as a whole is still 100% invalid in our eyes, enough so that comparisons probably wouldn't even be made at all during the judgement process. If you have to sacrifice a few frames for something, it's not a problem at all.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
Joined: 4/23/2013
Posts: 59
Samsara wrote:
Like I said on Discord, the differences between Saturn's TAS and this one could easily be chalked up to emulation differences. Looking back, he made his TAS on a completely deprecated and unacceptable version of Snes9x due to how inaccurate and garbage it is (I thought it was at least 1.51, but nope, 1.43). No matter what happens in terms of losses compared to Saturn, his run as a whole is still 100% invalid in our eyes, enough so that comparisons probably wouldn't even be made at all during the judgement process. If you have to sacrifice a few frames for something, it's not a problem at all.
Not to mention the fact he played on PAL when a Japanese NTSC version exists.
Player (138)
Joined: 9/18/2007
Posts: 389
the RNG script you linked didn't work, so i fixed the bugs. if it didn't introduce new bugs, get the update here. I derived a small damage test from it, and here are some of my thoughts (which need to be proven right or wrong) A good RN is necessary for critical hits. These do only appear if you attack the "right" spot. If you hit, but just barely, they do not appear. However, it makes a huge difference which speed you have during an attack and how long you have been running in a certain direction. For the same RN and attack but different speeds and positions, my damage varied between 8 and 10 for the blue enemies and doubled for critical hits.
Fortranm
He/Him
Editor, Experienced player (880)
Joined: 10/19/2013
Posts: 1122
Glad to see someone working on this game! Has any useful application of the OoB glitch been found? Apparently you can't save on world map in this game like in FF games, but sometimes I wonder if using sub-frame reset on lsnes can lead to useful memory corruption.
Player (138)
Joined: 9/18/2007
Posts: 389
Well, no, I don't think I am in the mood to start a TAS for this great game, mainly for the following reasons - The game is really long, it would be about three hours - All the movements (even running in a straight line!) have to be done frame-by-frame - There is a lot to test, and some of these require planning about two hours in advance weapons, grinding, death warps, potion and crystal collecting, maybe even glitch abuse
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2823
Location: Northern California
partyboy1a wrote:
Well, no, I don't think I am in the mood to start a TAS for this great game, mainly for the following reasons: - The game is really long, it would be about three hours - All the movements (even running in a straight line!) have to be done frame-by-frame - There is a lot to test, and some of these require planning about two hours in advance: weapons, grinding, death warps, potion and crystal collecting, maybe even glitch abuse
- TheAngryPanda is already working on it.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
Player (138)
Joined: 9/18/2007
Posts: 389
Yes, that's the reason why I tried the Lua script in first place.
Player (72)
Joined: 5/28/2013
Posts: 99
I'll check out the lua script as soon as I get the chance to (probably later this afternoon) and appreciate the help a ton. :) As far as useful things for the OoB's, I'm only at the end of Tower 2, and thankfully have had Samsara, Vaan, and Lemon's help to optimize/plan out most things throughout the process. So we're making steady (though slow) progress for now. I do have a job and other projects, so this will be a long(er) project of mine compared to most, but hopefully not as long as Secret of Evermore (2.5 years). That being said, the RNG in this game is so poorly designed it actually hurts me to think about it in many cases. It's still "good" from a RTA perspective, but for TAS'ing, you can run into significant problems that are literally impossible to correct without backtracking an immense amount, which can daisy chain into itself. This happens when the boss/enemies are few in number, but important for progress/techniques. Also, if anyone is ever interested in keeping up with and/or helping optimize/further the TAS, the links to the .tasproj and .bk2 are in my discord's topic for the channel, which you can find here: https://discord.gg/KDhsRYj I do my best to keep those 2 links updated after any real progress is made. I also don't mind going back to re-optimize movement where possible, so if someone notices anything, please say so. I want this TAS to be what everyone wants it to be, amazing and worth watching. :)
My Stream: http://www.twitch.tv/theangrypanda1 Mostly SoE TAS and casual playthroughs.
Fortranm
He/Him
Editor, Experienced player (880)
Joined: 10/19/2013
Posts: 1122
TheAngryPanda wrote:
As far as useful things for the OoB's, I'm only at the end of Tower 2
If there is any major sequence break, it might affect the route beforehand. It's unlikely though. The only potential one came into my mind is the gate between France and Spain... I mean Loire and Sylvain Castle. I don't know if you have it already, but here is an editor that can help checking game data: http://www.romhacking.net/utilities/1072/
Player (206)
Joined: 2/18/2005
Posts: 1451
Samsara wrote:
Like I said on Discord, the differences between Saturn's TAS and this one could easily be chalked up to emulation differences.
Wrong. To make it clear, the run is fully reproducible on new emulators (tested on Bizhawk), since the RNG in this game is entirely based on in-game actions, thus not influenced by emulation (which differs only in lag during screen changes) at all. That was actually a criteria I was looking for during run-production, with the intention to eventually convert it to the Bizhawk format when it was finished. I later decided to let it be, since it still shows the same quick way to beat the game, and the at most few seconds of added lag over the run wouldn't be noticeable, nor would they influence the ingame timer for which I aimed for anyway. It should be noted however that since Panda is using a different ROM, the RNG will likely deviate at some point due to shorter text anyway, from which some of the strategies would have to be readjusted to the given conditions.
See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun
Fortranm
He/Him
Editor, Experienced player (880)
Joined: 10/19/2013
Posts: 1122
Nice to see you again. Would you like to upload your input file?
Player (206)
Joined: 2/18/2005
Posts: 1451
What's the point? Snes9x 1.43 isn't accepted anymore, as well as the English ROM choice it seems.
See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun
Editor, Skilled player (1825)
Joined: 8/25/2013
Posts: 1200
Saturn wrote:
What's the point? Snes9x 1.43 isn't accepted anymore, as well as the English ROM choice it seems.
Maybe so we can figure out how you did certain tricks without reverse-engineering it from a two and a half year old video? Seems pretty obvious to me why we would want it. Also no shit the English version isn't accepted, it adds another hour to the game because of how much slower the framerate is. If the framerate was the same it would be fine.
effort on the first draft means less effort on any draft thereafter - some loser
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2823
Location: Northern California
Saturn wrote:
What's the point? Snes9x 1.43 isn't accepted anymore, as well as the English ROM choice it seems.
Wrong. The fact that the (E) version is in English has nothing to do with anything. You're completely overselling it if you just call it an "English ROM choice". What matters is that it's a 50fps PAL release when the (J) version is a 60fps NTSC release.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
1 2
6 7 8 9