Youtube encode is up now. :)
http://www.youtube.com/watch?v=L1RrLvJ4qsA
Aktan are you going to make an encode for publication, too? I think that is the only thing holding this back now, isn't it?
Well you are playing on hard now. So chances are there will be enough enemies to do that. Nice side effect. :)
Btw. great WIPs so far. With Bass this gam is even more fun to watch.
Ofcourse it backfired. MM doesn't sync from states. :D
Btw. there is still the youtube video in the first post for you to use. Dön't kno why you won't do that...
Aktan is still capturing this. MM is kind of a bitch to capture.
I said it once and I'll say it again: We should just go for what looks best for the particular game.
Those encodes are for casual visitors of the site who don't want or can't set up an emulator. Correct?
Well those people aren't as major nerds as we are. They won't remember how a 20 year old game looked back in the day on an outdated piece of technology (the 4:3 TV). They will however notice if an encode looks crappy and if you are going for a true-to-console look it often will look crappy.
We are TASvideos, we are not SDA. We make movies on emulators, everybody knows that. Trying to visually simulate a console and sacrificing quality in the process is just silly.
Yeah Grunz has an old crappy computer. So do I (Intel Core Duo [without the 2], 1,66Ghz and windows XP) and it syncs without a problem.
Make sure that you delete the majoras mask .fla file in mupens save folder, since mupen won't do that itself and check if your plugins are the same. MM is very desync prone and won't sync with anything changed.
Btw. Aktan did you solve the encode problems?
You don't have to. Grunz is wrong in saying that this will only play back on anti-ds. I played this back on normal mupen from the beginning.
I told him a thousand times, but he won't listen. Btw. can somebody edit the submission text to that effect?
Actually I also said this in this tread already. So for everybody:
This movie plays back without a problem in normal mupen rerecording!
If it won't play back for you anti-ds mupen won't help. The main thing changed on anti-ds mupen is how it handles savestates. That is an important factor in making this TAS, but has nothing to do with playing back the finished product!
Yeah Hopper is right. Our encodes look way too good. There have to be some unsharpening filters applied and the contrast reduced. Otherwise it won't look like it did on console! [/sarcasm]
1. Once and for all, SGB is not a console. It's a GameBoy-to-TV adapter
Well it does have the complete Gameboyhardware inside. So it is kind of a console. But that console is a gameboy. ;D
Anywho I don't get this discussion. The images stretched to 4:3 look like crap. Even the borders do. We should aim for what looks best.
Nobody is going to remember that the super gameboy used to stretch the image and that that looked crappy. I sure don't.
So the argument that it should be TV aspect ratio really has no merrit because nobody is going to remember how it looked anyway.
Btw. in the case of Pokemon there was not only the SGB on TV, but also the gameboy player in Pokemon Stadium and I'm pretty sure that one had correct aspect ration. So maybe people are going to remember that one and get weirded out by this wrong looking ratio?
Also soon (or already) there will be VC versions that will also have the correct aspect ratio. The only odd one out is the SGB and people are going to think we are incompetent in encoding stuff, not praising us for keeping it real.
Btw. I also think you could just crop out these borders. In Pokeon they don't do anything, so you might as well remove them and get a bigger game picture.
PS: I think there hasn't been much life in this thread because people who are not encoders usually don't come into this subforum. I only saw this thread through a link in the "gotta catch em all" submission thread.
But I doubt anybody will agree with sgrunt, to be blunt.
It only applies to chu hovers against a wall, not all hovers against walls. I really can't think of any besides the GB fence and the first block in STT. That's basically nothing.
But like I said, those differences are probably not enough to warrent a switch.
Nah, english MM draws two characters each frame, japanese only draws one. So the time difference is negligible. It's probably 1-2 minutes the english version loses.
Actually, it's only 30 seconds. I timed this myself a couple months ago using sony vegas.
Yeah but your timing can't be trusted. :P
Wasn't first cycle +beginning of 2nd cycle like 30 seconds already?
Anywho. With MM other version differences are more interesting than the text.
Jap MM has power crouch stab, which would mean you could only get one deku stick and still dish out 4 damage to every opponent.
Furthermore it has a little shortcut in PF.
On the other hand you lose flying zora and you will have to slash your sword before chu hovering against a wall. But these two shouldn't be much of a problem for a TAS.
So yeah I'd say overall Jap is somewhat better for TASing, but probably not enough to warrent a switch.
Yeah I guess it should be a normal Joe run form a clean file.
Ultra-V doesn't seem interesting enough on it's own to warrent starting from a save.
However if somebody wants to run any of the other characters I think Ultra-V should be used.
Ofcourse then the question becomes if wit upgrades or without...
V-Rated and Ultra V-Rated have all of the same enemies as Kids and Adults, but the tougher enemies appear much sooner. For example, there are already Cromartys in "Joe the Hero." There are Gelbys in "Some Like It Red Hot" and Metal Leos in "The Viewtiful Escape." There are much more obstacles in V/UV rated. More lasers, dripping lava, swarms of bats, frogs and moles, etc. Most of the regular Biankys have been replaced by Bianky Boxers and Cromartys. There are more enemies overall. The enemies have more health and you dish out less damage. Gelbys give less warning and attack faster. The bosses, especially have more health. King Blue has a stupid amount of health bars on Ultra-V. There are NO skull icons in Ultra-V mode and you take quad damage.
So there are changes but enemies also have more health and you deal less damage. So that could get boring. So hard to tell what would be better.
Where a game has multiple difficulty levels, it is preferred to play on the hardest difficulty level (for more interesting gameplay) unless the only difference between difficulty levels is enemy/boss hit points, in which case the easiest difficulty levels are preferred in the interest of speed.
These rules are not strict, but are motivated by the same concept as the guideline that says you should play on the hardest difficulty. As such, you can use a code to unlock the hardest difficulty, although it's better to first ask on the forums if this is a good idea.
Cheat is interchangeable with save here. So it is generally accepted that you start from a save to unlock the hardest difficulty. If it doesn't give you other advantages.
I think Ultra-V adds more than just more hitpoints to the enemies, if I remember correctly, right?
And since VJ is hard on easy, it would just be so much fun to see an Ultra-V TAS.
Sadly my computer still isn't working yet, so no screenshots, but I guess I'm going to make a few commentents anyway, if you like it or not ;p
Just ammending a few things Grunz already said:
before CT:
There really is nothing special now, however from Grunzs comment that swordless had to redo this and first cycle 3-4 times you can see that there were a few changes.
For example getting 2 (here 5) rupees before ever entering CT seems obvious now, but it took us quite a little time to get there. I think I finally came up with the idea. This way minimize transitions scenes were game time doesn't move and still get the adault wallet.
1st cycle:
I really wish mupen wasn't such a bad emulator. Here 2 resets would have really helped to cut down on the cutscenes.
Preparing For The Journey:
Since we now do the rest of the game in one cycle, we can actually withdraw all of our money and don't have to visit the banker ever again.
Furthermore we can fully stock up on explosives, since we won't use them through timetravel.
next:
Going "Clock Town -> Great Bay -> Ikana -> Swamp" was actually my idea and yeah I'm proud of that.
In the swamp we sadly have one minor improvement that was discovered way too late, which is down-A sliding past the Big Octos. But it doesn't cost much time.
Next is entering the deku palace through the backdoor to make the travel distance a little shorter and skip the cutscene. I think it saves like 10 seconds. Would have been more if there wasn't so much lag due to the dragonflies.
Great Bay Temple:
Oh that Boss Key skip was really fun to find. Grunz and I decided one evening in IRC to find a clip in that room an hey we did. We pretty much both came up with using through, then I looked thing up with Utility of time to see where there is actually ground to stand and if there is a back wall to it and Grunz actually got it to work in the end with his idea to die.
Ofcourse that took a few days all in all, but still really fun.
Stone Tower Temple:
Ah yeah the "real" chu hover. I nagged Grunz A LOT during the making of the run to try crazy shit to make this more optimal, but I think nowhere as much as here and he probably didn't ever get so frustrated. It took a long time to manipulate this thing, sadly basically nobody will notice, but I think it was really worth it, since the saved chu really helps.
Majoras Mask:
After killing phase 1 there is the Broken Deku Stick again. You have seen them a few times before, but here they are most important. That is actually the only glitch in MM I found alone. Yeah Broken Stick was known before, but I actually found 2 new ways to activate them and the damage way is the only thing that makes them usefull. So yay! :D
Deku sticks do 2 base and 4 jumpslash damage, compared to the Kokiri swords 1 base and 2 jumpslash damage.
This actually cuts the time it takes to kill phase 2 and 3 considerebly. Without it sub 1:30 probably wouldn't have been possible.
Fire arrows also do 4 damage, but they have a way too long recharge time to be usefull.
I think this were the most important influences on the TAS I had, but there were quite a few others. But I don't want to bore you. ;)
Argh my harddrive just broke, So no screenshots from me in the near future, as I have to try to salvage some data from that thing. I really hope there wasn't anything I absolutely needed.
I should have taken the weird computer problems I had more serously. Damn time crunch.
Great timing as always. XD
@Aktan
Just fyi mupen screws the audio for MM. There will be cracks in there, so you'll have to capture audio directly from the audio plugin. Maybe you know that already since OoT has the same problem, but just to be sure...
Oh and any changes in the video plugin will desynch this TAS. So don't even try. Antialiasing has to be done on the encode.
On the other hand, argument over how the Japanese rom should've been used as it would be faster in 3...2...1...
Nah, english MM draws two characters each frame, japanese only draws one. So the time difference is negligible. It's probably 1-2 minutes the english version loses.
klmz wrote:
Am I the only person here who has never heard of Anti-DS Mupen before?
It's a version of mupen that okaygo made, but sadly never published the source to. Afaik the main thing it does differently is how it handles savestates.
savestates didn't work right with MM, which caused quite a few of the desynchs. I guess it's because of the RAM expansion.
But you can play back this movie with normal mupen rerecording. I have been doing that since the beginning.
BrainStormer wrote:
I'm glad I never watched Abeshi's run and don't even feel like it after watching your run.
Yeah for the first part it really is a much less optimized version of this run. For the second it's just really bad. Because there was no WIP he could copy input from anymore and we did not publish route changes after a certain point. So since he doesn't have any creativity of his own, he was stuck was a sub optimal explosives route.
MrGrunz wrote:
Is actually anyone encoding the video? Swordless told me he gave the m64 to Aktan a few days ago and according to Swordless Aktan would do a HD encode of the run.
[...] The youtube people have been waiting even longer for the run now so a good encode is very important to me within the next few days.
As far as I understood it, you would also like your youtube channel to be the official channel, for the final video. Maybe you should clarify that.
Anywho great TAS, I think that goes without saying. Sad that after 3 years at the time of publication I don't have much time myself.
But I think I will find a little time tomorrow, to make a few comments on the publication text, the TAS and post a few screenshots I'd suggest. :)
Oh and this deserves a star without a doubt.
Am I in court all of a sudden? Since when do I have to proof things on the internet?
Fact is he did that in "his" Majoras Mask TAS. He hexed in most of Grunzs WIP. The play quality diminished drastically after he was out of WIP to hex.
Fact is he doesn't use the TAS input plugin but a controller, so he can't pull of some of the things in this TAS.
Fact is it looks like it was hexed or directly copied and people, who know OoT TASing better than I do, confirmed that.
Ofcourse I can't proof it without the m64. But then we are not in court here so I don't have to. If we were in court and I was the prosecutor I could force him to hand the m64 over. But we are not.
Yeah, I watched just the first 5 minutes of both. The camera angles are totally different, the town entrance uses an extra swordslash... So this whole "OMG HE HEXED AND STOLE BLOOBIEBLAS INPUT" is a total bunch of made-up bullshit by SlowKing. He has no proof, other than they take a similar route.
You really don't listen, do you? He couldn't use bloobs original rupee route, since that manipulation won't hex. So he did a suboptimal one himself.
And he probably didn't get the water HISS to hex either, since it's very desynch prone and to my knowledge Bloob is the only one who ever got that slide to work to begin with. So he used a suboptimal method here as well.
Just because he simply can't hex or directly replicate some small parts doesn't mean he didn't do it where he could.
Hmm yeah he could at least time correctly, though. Since he knows how it's done.
Btw. reading back I can see how people might have thought I was mad. "fraud" maybe is a little bit too strong a word. Sometimes it's hard to find the correct vocabulary for what you want to say without sounding to harsh if english is your second language. ;)
Well only improvement I noticed is that he pulled out the odd mushroom twice, thus saving some time. But that was already known.
It seems he didn't read the submission thread, since the improvement I noticed isn't in the movie.
"He didn't choose to use our timing convention, and even though he clearly stated how it was timed, we should ridicule him because our arbitrary way of timing things is better and his posted time will make us look bad because it isn't represented in the same fashion."
How reasonable.
If he had written "Thanks to Bloobiebla for certain sections" would it all be okay, or would you still be mad because the OoT team's glory is stolen? The language barrier seems to make it difficult for you to have any understanding of his motivations, so instead you just continue to bash him.
I'm not mad. I'm quite indifferent. Just saying how it is. People here seem to become really really mad though.
Thing is TASes were always and always will be timed by the length of the movie file. It's not completely arbitrary it's actually the most clear cut timing for speedruns out there. Even abeshi knows how to time TASes as he timed his first ones correctly. If he did time this way from the beginning I would cut him some slack, but he didn't.
And yeah writing "Thanks to Bloobiebla for certain sections" would imo be the least he could do. Though he shold probably write "Thanks to Bloobiebla for 90% of this TAS".
Soulrivers wrote:
In only the first five minutes, he takes a different path for the rupees, landing at 49 (which I guess isn't optimal), skips buying the deku stick and therefore listens to Kaepora Gaebora when leaving Kokiri Forest. I have no idea if any of this saves time though. In any case, I'm sure he was taking notes from Bloob's run, but to say he copy-pasted the input is just being ignorant.
So he couldn't get the luck manipulation since that won't hex. You are kind of making my point for me here.
I should say that I myself have no shame in completely replicating parts of a previous movie's input if I can't find any improvement when trying to obsolete it. I've done it for all movies I've obsoleted and I don't think there is anything wrong with doing so. So even if there are some sections, small or big, that he completely copied/replicated, I don't see anything wrong with it as long as he added some content of his own, which he apparently did.
Well I for one think you should at least give credit when you do that. However there is a big difference between taking small parts here and there and copying 80-90% of the movie. Just saying...
