I sent this sort of message to DeHackEd not long ago, but I found out he doesn't accept private messages...so I'll post in public instead.
Sometime last week, I found out that DarkMan's GPU/SPU Recorder can be used to produce TAS-like videos (I use ePSXe). This is probably what Omega was referring to when this topic was started, but I'll explain anyway.
The plugin works well enough that I was able to manage a "mock TAS" (as I call it) of Mega Man X4 by slowing the gameplay down to 20 frames per second and using savestates. The GPU Recorder plugin records to a .rec file that plays back with GPU Player (which is a buggy program). As long as you did things correctly, it should play your video smoothly, with no traces of re-recording (nothing like the video jerking back or something).
Of course, this plugin isn't perfect (the SPU Recorder lags) and likely isn't ready for this site's standards in it's current form, since it doesn't produce input files for emulator playback. I just thought that someone would like to know about this option, and maybe modify it to make it acceptable for use here. Perhaps it would be easier than modifying an emulator itself?
Also, I've seen some sort of plugin for FPSE called JoyDemo in my searches. Apparently, it can record input for emulator playback. Likely worth looking into, if not just for making the job easier.
Anyway, I might have just repeated things people already knew, but I hope this was new/interesting/of use to someone.
I'd just like to point out that I believe you could save time by using Kyle to kill that Mercenary in Chapter 5x. This way, you avoid gaining a level with Forde, which eats time. That is, unless you had a reason for powering him up.
Other than that, I'm not really any good at spotting TAS mistakes, so I can only focus on things that jump out at me. Good job so far though. :)
laughing_gas is correct. The TAS will easily take 40 hours to make, so comparing it to a 7 hour speed run isn't relevant. I am fully capable of making the run on the J version, but like I said before it will be less enjoyable for me and the vast majority of the viewers.
I'll have you know that I spent days[/b] working on each one of those 7+ hour runs. They were the result of many, many hours of saving/loading state when things didn't go well enough. Therefore, they are comparable.
Also, to not TAS a certain version just because it would take you longer seems absurd to me. As far as I know, TASing isn't about how long it takes to get a finished product, but the quality of that product.
There's a big difference between merely watching a tas and making a tas. It is annoying beyond hell trying to go through a four hours of the (J) version decoding all the attacks and whatnot.
Not to drive this off-topic, but do you realize who you are speaking to here (no offense)? If I can manage 7+ hour playthroughs of Chrono Cross (multiple times) in Japanese, a TASer should be capable of around 4 hours worth. Honestly, you get used to it and learn to recognize certain things, even without knowing the language. It's not that difficult.
First of all, I'd like to put a disclaimer on this. The following comments are the result of things building up and these are all my opinions. I realize some of them sound rude, so I apologize in advance. It's just reached the point where it needs to be said. Thank you.
JXQ wrote:
But they're not the same game, so that comparison is flawed, which is what several people have been trying to tell you from the jump.
hero of the day wrote:
It seems like I will get some no votes either way, because some uneducated people want to compare two completely different games.
Saturn wrote:
Some uneducated people still compare 2 completely different games.
This doesn't make sense. In both games, you travel through time and stop Lavos, doing whatever tasks along the way. Both have the same characters, same story (barring version changes), same attacks, same enemies, same music, etc. So, how exactly are they "completely different?"
Because a game is in a different language, it's instantly a completely different game? To be blunt, I find that idea completely absurd. If you wanted to use that excuse, why not call Mega Man and Rockman completely different games? If they have any text at all, you can see that they are just completely different!
hero of the day wrote:
Playing on the J version is a sure fire way of getting about 50% no votes.
Now, that I can agree with. Sad that a lot of people will fail to realize that the game you remember (assuming you played it before watching the TAS, like you always should) is still there, only with a different language. This leads me into the whole "following" the TAS thing, which is ridiculous.
If I see Crono spin around a guy 4 times and hit him, it's Confuse...regardless of what the text says. If I see the cook in 600 A.D. give Crono's gang an important item, it's still the Jerky for the knights. If I see a big ugly freak with two arms and Doors of Doom, it's still Lavos. The list goes on and on, but you get the idea.
Now, I'll apologize again for saying things in a rather blunt manner. But please, somebody needs to see the error in the "it's a different game" logic.
Though I've stated my opinion before (more than once, I believe), I'll repeat it here. I'm of the same opinion as Stanski (option 2), in that if it's faster, use it. If this means using a Japanese version of something that is text heavy in order to save time, so be it.
The less time spent doing something anybody can do (even without tools), such as skipping through walls of text, the better. Less time spent skipping through text means that a higher percentage of the TAS' run time is spent doing other things, which would probably mean more action. I also believe that action is far more interesting than skipping through text, but that might just be me.
About the whole "knowing what's going on" argument, I believe that you should easily know what the idea is, if you've played the game before (which in an RPG TAS case, should be kind of important). If you ever get confused about what just happened, I think the author should write detailed comments explaining the actions taken in the run. That way, you can read about anything that confused you.
Also something that may just be me, but any "I can't read the text" problems outside of battle (though I covered that issue in the above paragraph) shouldn't be an issue. If you need to read the text, play the game or read a script FAQ.
On the other hand, if the game has advantages for a specific version (like a new trick or something, not "I can understand the text"), I believe that version should (likely) be used. When this is not the case, and they are all alike in tricks, use the fastest version.
Anyway, this is all my opinion on the matter.
But you speed through the menus so fast anyway! Can't you watch what the characters are doing to figure it out?
I definitely agree with this statement. I believe that by watching the animations, you'll easily know what attack was used. Same goes for items, just watch how much they heal (200 on one person is obviously a Mid Tonic, for example)...or what stat is raised (in the case of Tabs).
This becomes slightly more difficult in the area of equipment, but you should be able to tell some of it from the area they pick stuff up in. But to make things easier, I believe the author should write detailed comments about what is picked up and when it is equipped. That way, if you get confused, just look at the comments and it will all make sense again.
Personally, I hope anybody who does an improvement TAS uses the Japanese version. I feel that if you want to read the text or know the plot, you can play the game yourself. Less time spent mashing through dialogue means more time in battle and doing things that can actually be interesting, and it even lowers your final time as an added bonus.
About endings, I'd say three categories for this game make sense: fastest New Game completion (regardless of ending), fastest New Game+ completion, and fastest New Game 100% completion (sidequests, like SDA's 100% category).
Fixed. :)
Anyway, since I figured this stuff out from the forum post before the report, here are the listings for final equipment and what seems to be overall item usage (possibly just in the final battles):
Crono: Swallow, Vigil Hat, Aeon Suit, Green Dream, and 15 speed.
Ayla: Fist, Vigil Hat, Lumin Robe, Silver Earring, and 13 speed.
Marle: Lode Bow, Hide Cap, Hide Tunic, Speed Belt, and 12 speed.
Items used: 1 Potion, 1 High Potion, 4 Mid Ethers, 5 High Ethers, 2 Elixirs, 7 Revives, 2 Barriers.
It seems his Lavos 1 strategy was basically: "Give Lucca +10 magic defense with the Wall Ring and Taban <whatever>. Have her survive the first attack and revive people." Sounds like a nice way to torture yourself, doesn't it?
Also, he used three powerlevel areas: Tyrano Castle (something about slapping a button for fights), Enhasa (Nu group once, for tech points), and some other location that Google screwed up the table for.
On certain missions when you have to defeat the boss, you could use the control boss glitch to make the boss drop his items and send him towards your forces for execution... That should speed up some levels.
[Edit]: After speaking to Nitrodon, I take back what I said about the uses of the glitch. Ask him if you want to know good places, he has good judgment.
petrie911 wrote:
Also, has anyone considered a TAS of FE6?
While Fuuin (Hard) would be interesting as well, shouldn't you have said Thracia SSS rank? ;)
I'd personally love to see Seisen, but the best I could do a while ago was sub-8 (I think)...with heavy glitching. I'm getting off-topic though, so no need to mind me.
I'm glad to see you're working on this run again and it looks good as always. I'm a bit doubtful about whether or not going down the right side of Chapter 26 is a good idea, since all the enemy killing sucks up time. I doubt Jaffar will prove to be of much value, since you can manipulate critical hits anyway.
Personally, I'd test a 2 unit strategy on this chapter. One person carries Jaffar out of fighting range and the other person grabs Zephiel plus the Boots and hides in the same room (possibly with no weapons equipped).
As for future chapters, I'd highly recommend saving your Bolting uses for the Final Chapter. It's too bad you didn't grab the Hammerne, but 4 Bolting uses still makes for 4 dead bosses at the end. I know the Warp staff goes without saying, but Athos warping is really fun.
As for the Boots, I'd try Heath or Florina. The mine trick strategy for Victory of Death is absurdly cheap and with +2 movement on your flying unit of choice (that is capable of lifting a promoted Eliwood), it might end even quicker. Since you have Jaffar anyway though, you might want to give him the Boots so that he can keep up with your mounted units.
There is a second RNG in the game, which until this post I assumed only affected the items in the desert and various graphical effects. That RNG (at 03000008-0300000b) is advanced when you do your shopping, causing a desynch when the game chooses a formation.
The only way to advance that RNG before Ephraim shows up is to go into the item menu during your turn. I tested, and doing this 4 times will make Ephraim's troops arrive in the desired formation.
Man, I wish I had known that ages ago, it might have saved me from the Chapter 8 troubles I was having myself. The RNG seemed normal, but when it came time for Seth to kill the boss, things didn't go well. Thanks for the information.
Anyway, the movie looks good, but I've noticed a few things that could be changed. When manually going to Fort Rigwald on the map, you should be able to hold B, move the cursor down-left, and then press L. This should set you on Fort Rigwald, since I know a similar process works for Port Kiris with Eirika (sitting at Ide, move right, press L). I don't know if it would be any faster though.
The other thing was sending items when you have too many. Once the dialog shows up saying "Sent <item>", you can press A to skip it. Should save a few seconds, since you did it with a Knight's Crest, Chest Key, Elixir, and a Killer Axe.
I guess you're right about the Tower turn count, it's just that it doesn't show up in the final listing. Even if it wasn't counted, I doubt many people would want to watch super RNG abuse for good level ups for about 8+ characters.
The control enemy glitch is something I think would be neat to watch, mainly because it's a difficult trick to do without PWASE. Seeing someone nail the trick on the first try every chapter would at least be a bit amusing, as you laugh at the enemy and their inability to win. Myrrh with Wretched Air and an early Naglfar might be added bonuses.
As for Renais castle, I skipped it by mashing Start and A repeatedly, if I remember correctly. I doubt the A button really did anything though.
While I wouldn't call this a huge thing, I'd like to mention that the Prologue scene that displays Renais castle is skippable. That should save a few seconds, since you can't do anything while the location box is there.
I like your use of Vanessa's critical hits (flying unit strategies are fun to watch, especially in a TAS). I must admit that I'd prefer pure speed (Normal/Hard), but minimum turns should be pretty interesting. The only problem is that to really achieve minimum turns, Tower abuse is probably necessary. This is only if you want to be technical about it, since I guess there would be a different category for such a thing...if it would be allowed at all (Tower abuse would be seriously boring).
Speaking of a different category, would the control enemy glitch be possible in every chapter for a TAS? I've tried it with frame advance on Chapter 4 I think, but had no luck.
Anyway, sorry to go on a little ramble there. The improvement you made looks good to me, and I wish you luck with this project. I'll be looking forward to it. I can also attempt to help with basic strategy ideas if you want.
P.S. Nitrodon, I watched your practice run a while back and just wanted to say I found it amazing. O_o
