Ok so we have some kind of unofficial clearance from adelikat for the psx version, me and Uroboros are teaming to optimize this run like it deserve to be, I except this to be done relativly fast given we have a very good draft to work from, Im hoping for sub hour movie if everything goes right.
Edit: 102 frames saved on main title screen + character select (end of first room...), so well yeah, im sorry to say this was very sloppy indeed
AFAIK it's fair to vote No to something that's entertaining but that wouldn't be watched again
And thats not fair to vote yes in that case ? i voted yes to many runs i havent watched again
Note that i wasnt complaining about your comments, im _sure_ they are fair (except maybe the gradius/tetris part since thats the kind of run that cant be matched regardless, especially in a fighting game), its just there was that part where you was speaking of entertainement and obviously yeah we was entertained, and thats subjective, if i was pro at mortal kombat its likely it would have been less fun already, yeah.
If anything the movie did sucessfully confused and amused me (maybe not the begin tho, but surely after 8 minutes in).
The cookie is lie! Booo x.x
The entertainement is supposly subjective, personally i dont know the games well enought to have a realistic technical comment against the movie, but seriously, how hard is it to admit that it can amuse people to watch this kind of movie in all fairness? You really think this is doable realtime?
I wouldnt like to take a dozen every day for breakfast but once in a while its still fun to me.
Now if there could be less mortal kombat troll submisions it would be a breath thats sure.
Did i won a cookie ? :p
I realized how to open doors correctly and how the menu thing worked around mid-way in the TAS (Caverns Area)
Those kind of things if you dont mention it in submision im not mind reader i cant guess it, i wouldnt have checked otherwise, but anybody replaying the movie frame by frame would spot it tho (well that you turbo'ed and didnt exactly went for precision)
Well i guess me or Uroboros will take back the burden when we are done with re2.
Official mind on the version matter would be _much_ appreciated, like i was saying allowing chris psx because chris ds isnt made yet doesnt make sence, so we have to be clear wether or not the psx version is ok for tasing, and if it shall cross plateform obsolete ds or just get published alongside.
Edit: read it as, i want a judge answer regard adelikat judgement over the ds movie, plz, the psx run exist despite its canceled somebody has to address the issue.
Well im sorry you didnt got more feedback on the forum, this can happen, you cant blame us to be slower than your pace, thats not reasonable, we arent bots :p
Now of course if it was mario64 or OoT or stuffs like that yeah you would get more feedback as more people are interrested, I was kinda interrested as well but no time to deal with it busy with other stuffs.
10 seconds range sound like realistic, wether its a bit more or less is kinda a detail, but yeah thats those 10 seconds that can make a huge difference to say you are really doing something considered superplay, i remember for castlevania adventure those 10 seconds prooved to satisfy a lot of the borderline feedback, while it was already about 3minutes faster than the rejected run, so in the end yeah you would hardly see the difference in the run itself but technically its more intensive work and more satisfacting overall.
Edit: ok so i finally got a hand on the movie, it doesnt end correctly (skip ending...), also i took randomly a menu access to test it, it was just about equiping the bazooka (not long after you pick it up) and i won 12 frames just for optimizing that menu, it might be well over 10 seconds actually, you apparently applied turbo strategy everywhere, so theres not any precision, while this doesnt forcably is a problem in many cases (the game is often all about running dumbly), the menu is the perfect exemple where you should have had extreme precision and not just bashing keys alternativly on frame advance, i wouldnt have underlined this if you was 1 or 2 frames off, 12 for such a trivial thing as the menu is kinda not within our standarts.
This isnt a huge criticism, but i was hoping there would be more precision, the run is a bit easy to obsolete in its current state (doesnt require new tricks or knowledge).
Ive also determined the pause trick is slower than just shooting, in REAL time (not game time), what this mean is unless you would be hit by a boss, or cant manipulate the zombie AI, this must be avoided at all cost (and you seem to use it everytime you can) 10-20 frames range saved (depend how you count since you also seem to turbo those, ie no precision)
You dont have to be perfect, re3 aint perfect and it was my first tas, it has to be as close as possible, if i didnt made a test run first im pretty sure it would have been lot more difficult for a judge to accept second version, things like the yes/no trick or speedup with one handed guns can only be founded while frame whoring or trial and error for exemple, and thats what ppl except from you here, what i mean is even if this get accepted you wont always get away from lesser technical quality (well actually in general you never get away) and greater game knowlegable, 7 days for a 1 hour movie is barely any trial and error for a game like that, usually this is call test run, note that im not saying here it shouldnt be accepted, its likely improvable if its the case (can be osboleted).
Now onto the version question that mukki did bring up, i think obviously theres cons and pro, so its not a easy call, you also have to keep in mind there should be (well imo) a chris run alongside whatever is published, what i mean is for exemple it would be silly to reject this and accept a psx chris because nobody made chris on ds so far.
Im more for the original version but the advantage of the ds one is the cutscene skip obviously, it also mean its quite different timing, tricks and maneuver all way trought, not a easy call definitly, unless we where to allow to have both versions published.
Arakudo runs has around 56xx rerecords and is 7 minutes longer mine has around 43xx.
Wrong my run was around 80k rerecord (this is in submission so you missed it), my movie was destroyed near the end and while hexing input i lost rerecord, everybody know it took me 3-4 months to produce it (test run+ published run), its on the forum topic.
Also more importantly, its not because somebody get low rerecord published it mean you are allowed to make things sloopy, very very very bad argument here (thought i understand it and i think its natural, but remember here we are all about doing godly stuffs lol)
Im inclined to vote yes because i believe the ammount of errors will save only a couple of seconds, and not minutes, this isnt whored to death (theres small errors) but i think its well made and should be borderline publishable because theres efforts and obviously the route and tricks are well planned. I can easly understand ppl voting no or meh, coz usually even a few seconds can be enought to refuse a run (this happened to me as well), im just on the other side of the borderline.
I havent voted yet coz i didnt had time so far to watch it fully from begin to end.
For a first time at tasing this is good, that im pretty sure, so dont get demotivated and take the comments about optimization on the bright side, it actually have nothing to do with the game itself, but what the site except from a taser, wich is to test everything several times to unsure its the fastest, you rerecord behavior is obviously far far away from what they except, this isnt disqualifing as a whole, it will comes naturally over time while tasing.
The reason for jap version to be faster is simple, it take less character to say the same thing, so the texts are shorter (quicker...), im not sure about this one but in re3 boss have less hp in jap version.
It is usually prefered to use the usa version (see rules on the site)
9-2 is 4 rooms, but i guess you speak of the vertical corridor at end, they are out of reach for the trick (you have to be able to go upward the ring, there you barely reach it with your head)
Pretty sure scrimpeh tryed ninja jumps but it wasnt faster
Sound like 30/60 fps issue, if the game is really 30fps then even if you input any frame its same result, since the game is still processing and not polling anyway.
It would make little sence to change the rng routine in between versions so that you get less headshoot, and you wouldnt have any way to figure this unless decompiling or something like that.
In resident evil games usually hp boss are lower in japanesse versions, i dont know if that is followed for this game.
The reason to take pal version is glitchs, afaik it allow to skip certain boss or stuffs, i dont think its a problem to run the game at 60hz because this is _in_ game setting, the rule is about to run a 50hz game but at 60hz, resulting in smaller real time video, this isnt the case here. (ie you didnt changed emulator speed, the game did)
Y a peu de chances que le sprite du perso soit sur un layer totalement independant, c possible mais pas obligatoire, de plus sa risque de commencer a etre compliquer pour l edition de la video, genre ajouter la transparence ou des couleurs different (sur metroid les couleurs de samus marche bien pour voir les differences), ajoute ensuite la complexité de mixer 8 runs a la fois, je crois honnetement ca devient baleize.
Mais oui sa semble fessable sur certain emus (nes, snes, genesis) de desactiver les layers.
Je pense perso que par lua/png c un peu plus pratique, meme si j'avoue que la premiere fois ca peut etre compliquer de dumper les pngs, trouver les addresses ram, se faire la main sur le script ect... au moins a chaque nouvelle version du run la ca devient rapide et facile.
If we say that "minimal presses" is a good goal, then it should be "minimal presses ingame" for this particular game, as I don't see any point in waiting for all those menus.
Thats a good point and imo should apply to _any_ games, why force the audiance to watch cinematics and stuffs ? thats often not interresting (ok maybe not always but usually it is irevelant)
Chaque couleur correspond a un run, en fait la video compare 8 runs (!) a la fois , les chiffres sont simplement le nombre de frames en retard, c une indication relativement importante, faut comprendre que si les autres runs etait sur un timing "globale" du coup il serait pas sur l'ecran a cause du retard, pour compenser ca en fait a chaque debut de stage les ghosts sont remis a zero, histoire de voir les ghosts se faire doubler dans le stage en cours, et non pas par rapport au debut du run (les chiffres en haut).
En fait le principe est assez simple, le lua en premier fait un "dump" des position X/Y du personnage pour chaque frame, tu recupere un gros fichier texte bien horrible avec que des chiffres, ensuite un autre script se charge de "simplement" afficher les png du personnage au bonnes positions, en overlay.
La version snes du script de amaurea est vraiment tres bien foutu, et je crois snes9x peut directement faire la capture video de tout ce petit monde.
Pour desmume c un peu plus touchi, d'ailleur c'etait fait avec une vieille version et je suis meme pas sur si ce serait fessable avec la version actuelle, il fallait aussi un soft externe pour encoder.
Sans vouloir trop m'avancer fceux et gens devrait etre capable de faire la meme chose.
You waited 7 years surely can give us a few more days to finish :D
Im in stage9 now, theres a possibility to redo some parts in stage8 but with the "global" timer on plateforms this may not happen (ie useless), stage8 is stupid to tas, well the whole game is amazingly boring/depressive to tas but stage8 especially is pesky.
We are 2426 frames ahead at end of 9-1. (for now)
4-4 is still the biggest improvement for a single block so far. (660 frames)
Une autre approche c le script "ghosts"
Link to videoLink to video
Bien sur c bcp plus compliquer pour produire ce genre de videos, en gros faut dumper tout les sprites des perso du jeux en png pour les overlay sur l emu (par lua), encoder est aussi toute une histoire.
The main issue youll have with this is language, if the runner is speaking english, youll have to sub it, so overall, I think another solution would be to do that by text instead video, for the same result. (aka the real submission text... or any pm)
I know that doesnt address your issue/idea, since you was whilling to make the audiance understand actually there is real people behind the scene and not just bots.
Yet theres is a flawed concept here, if it is bots that are beating the game(s), they cant generate themselves spontanously over the internet, someone must have wrote those bots programs, wich is pretty much the same in the end.
Anyway, im not so sure the audio proof is the real ultimate solution to this, since you guys are 2 presenting the show, maybe one of you could impersonate the runner (just to do the voice acting), that way you would have a guaranted interview for each review.
Ok so lets get something straight NOW, we dont kill all candles or ennemies, this create extra lag in many cases/places (hint: in 4-4 any kill create lag i saved 30 frames just by avoiding killing anything and not destroying candles), also the original wip DOESNT do this, so its totally pointless middle run to add goals like kill all candles or max score, now this point is clear enought i hope.
The dragon fight is faster than the one from Dilman, wich was unmatched before afaik, 1/3 faster than phil movie, you can call it sloppy, whatever angy, yeah you have way too high exceptation because stuff you just sayd is simply _impossible_
(you are irritating when you are saying things like, do this, do that, we are not your waitress dude, either its awesome and your favorite tas, or its wrong and we didnt dealt correctly with that 75# bat or that 29# shit there and here, make your mind otherwise we get contradictory impressions from you, we arent claiming to make THE perfect SCIV movie, but we beat all precedent movies/wip pretty clearly)
There is actually a use, scrimpeh found technique, then I found the place where to use for our purpose, for those interrested its scIVzip.smv on the link he provided, it work, and it avoid to wait for a full cycle of the blocks, thats to say, this is likely the biggest improvement that there will be in this movie (this trick is nuts really, megaman style)
