Posts for Heisanevilgenius


1 2
22 23 24
28 29
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
pirate_sephiroth wrote:
Are you boycotting his pens too?
Damn straight.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Just want you guys to know that I've organized a boycott in my city on Phil's Fresh Eggs for besmurching Phil's name. I have currently about eighty-six mothers, sixty-eight Hispanic men who don't understand English very well, and one hermaphrodite hamburglar impersonater all ready to pull their support. Good luck selling your eggs to your target market, NAMESTEALERS!
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
:D http://hyena.kiwibonga.com/M-I_NEW.fcm Certain parts of Level 1 are more optimal, level 2 is more entertaining now, the pass in level 3 is still spawned in the good place. Yay!
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
After a lot of trial and error, I now have it back up to normal, with one small difference. In level 2, I now take damage from the helicopter shootout (where I narrowly miss it before all the new improvements) otherwise the boat race is exactly the same. I'll probably have to edit it (was thinking about redoing the helicopter bit anyway since half of the time I unintentionally manipulated the helicopter into not firing instead of dodging the attacks, which would be more entertaining.) I'll back it up anyway. The way it is right now, the pass is still spawned in the good place in level 3. Hopefully redoing the helicopter part won't alter that somehow. Overall, thanks to Randil's tip, I've saved 181 frames. My rerecord count has shot up, though, from 1943 to 2445. This is because of three rooms I redid (a couple weren't quite optimal anyway) and because I manually altered the FCM files to add 1 to the rerecord every time I hex-edited. I figured hex-editing is like a re-record anyway, and that would keep the number more accurate. I'm a dork that way.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
First, N64 is strange about how it renders frames. Although NTSC is 60 frames per second, N64 seems to render only half of that (maybe it differs from game to game) so every other frame seems to be wasted. My facts aren't clear about this. You'd have to ask someone with more experience. Second, many games have a delay like the one you're talking about. This actually makes it more realistic and easier for the player. When you squeeze the trigger, the gun takes several frames before it actually goes off. This is hard to time for a TAS. You'll need to test to see how many frames each gun takes before the shots are fired. Side note: I know that pistols work differently than most other games in Perfect Dark because their firing rate depends on the player. Potentially (especially for a TAS) a pistol has the fastest firing rate because you can theoretically squeeze the trigger more times per minute than the fastest machine gun would automatically fire. If you play with slow motion or combat boosts you can really see the difference. It's actually really fun and cool to play against sims in slow motion and see them slowly turn a corner with their high powered machine guns sputtering out a bullet every couple of seconds, and then you empty your full pistol clip into them in a half second and watch them very slowly fall to the ground.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
From my limited experience, I can say that N64 above all others is a pain in the ass to TAS. As a general rule for me: I only use one save state when doing a TAS. Maybe it's just some sort of superstition, but with some of my initial experimentation I had at one point had a strange desync when I used two save states. (I saved one at the beginning of the level so if I made a mistake I didn't have to skip ahead as far, then used the other for more immediate stuff.) Since I started using only one save state for every TAS, desyncing has never been an issue (Although I've only done this on NES, SNES, and briefly on GBA.) Doing this on an N64 emulator would be frustrating at best. If you make a mistake, fast-forwarding from the beginning would take a very long time. Guanobowl had many, many desync issues with his Ocarina of Time run before it was finished. I experimented with Perfect Dark for a few minutes and the aiming is a very big pain. I have a controller with analog and I thought that would make it easier. It didn't. I gave up on the idea after I couldn't kill the first enemy without careful aim. (I wanted to drop off a ledge and headshot him in midair without losing speed. If you can't manage that, a TAS just isn't going to be fun to watch.)
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Ah. Didn't see that. If a mod wants to append this topic to the end of that one, I have no problem with that. I see you never actually got around to it. I have a few projects I'm working on, but this one will be on my list fairly soon I hope.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
After ten levels or so I just stopped watching. I'm sure you did a good job, but this is a terrible game for a TAS and the fact that you did it more-or-less optimally with only 50 rerecords is a testament to that. Speaking of which, 100 levels with 50 rerecords.... On average that's 1 rerecord for 2 levels. That's nuts. Edit: By the way, a couple of rules might have also been broken. 1. Like someone mentioned, this game is homebrew and might fall under the hacks category. 2. You didn't include your real full name in your submission.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Well, I'm sure the whole server will come crumbling down in dust and small pieces if we have two Monopoly movies.
Post subject: Double Dragon 2 - 2-Player Mode B
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
I'm still working on a few other TAS projects, but it occurred to me today that there's some potential here for something cool so I thought I'd start a discussion. There's already a 2-Player Double Dragon 2 run in Mode A. (If I remember correctly, it was done using slow motion instead of frame advance so can be improved in any case.) Mode B, for those who aren't familiar, allow the 2 players to hurt each other. I thought of a few possible time-saving advantages to this: 1. At the beginning of the second level, two players need to climb down a ladder at the same time. Dropping off the ledge would be faster, but doing so causes you to fall in a pit. Here's what you do. Say Billy begins climbing down. Jimmy kicks him, causing him to fall straight down. Jimmy jumps down off the ledge at the same time so Billy doesn't fall offscreen. Jimmy dies, but Billy saves a lot of time getting to the next area, and Jimmy respawns near him. 2. In certain long rooms, one player sacrifices one life by kneeing the other all the way across the room, and landing on spikes or in a pit in the process. They respawn closer to the other player. More possibilities might be explored that I haven't thought of yet.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Okay, this is crazy! The time-saving feature Randil brought up saved be 190 frames exactly. Because a car was spawned differently, I may have wasted one frame to avoid it. So it's a difference of 189 frames. One scene transition required me to add a frame so we're now at 188 frames saved. Then later, I find a transition is actually four frames faster. So now it's 192 frames. Apart from the one wasted frame dodging the car, this was all done with hex-editing, so no new stuff was added. I went through that very short laser grid room and re-recorded it. The room takes a couple of seconds to shoot through. It's possible it may be a few frames faster than before, but I may have also been a frame or two slower because I had to reposition myself a little bit to go through the exact middle between laser grids (which I didn't have to do as much before because the run spawned me at the right moment that the laser switched just before I walked through). The exact frame I exit that room in the original run is 15261. I checked twice. In the new run, I exit the room at frame 14083. Again I checked twice. I subtracted 14083 from 15261 and I got 1178. I checked this many times. How could I possibly have saved 1178 frames!?!? I'm baffled. Edit: Somehow after checking this thing TWICE I read it wrong. It was 14261. I saved 178 frames, which is MUCH more believable.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Okay, I've figured it out. It's actually pretty easy. Strangely, the key seems to be walking through them at exactly the edge of each "square" that the laser goes through. This means that hugging the wall will allow you to pass right through them, but also walking at the exact middle between two squares will do it. You can do this diagonally as well. I wasted a lot of time and re-records when the answer was very simple. This seems to somehow be related to the sludge damage principle, but I'm not sure exactly how that works either.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
I actually think this game could make for a good TAS. There were some things I liked (like turning all of the bells yellow and collecting them all at once for lots of points.) It just needs improvement. I don't want the author to feel discouraged, but just try harder.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Here's the funny bit. Everything is now syncing properly up until the room with the trip wire lasers, which now sound the alarm when I step on them. Since I have no idea what caused the walk-right-through-them glitch in the first place, I'm now going to need to find out. I'm hoping it's some sort of case of luck manipulation. Once I have that figured out, the rest of it seems like it'll be very easy to reproduce.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Okay, crisis averted.
JXQ wrote:
You should hex-edit anything at the beginning as early as possible because it may desync later, and that would mean less overall work to redo.
Man, were you right. It's going to take quite a while to fix the desyncs. They mostly seem to be with scene transitions. Enemy behaviour and whatnot are all seemingly based on stuff that happens in each room, so that's not a problem. But here and there it looks like the black screen between scenes takes up shorter or longer periods of time. Periodically I'll have to move my input back by one frame at some places, and forward one frame in other places. It should probably even out so that it takes the same amount of time, and I'll still have saved 190 frames. I hope it's all worth it.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
WAIT! I think I fixed it! Edit: Okay, seems the problem was that the first byte of controller data was 81 in one and 82 in the other. I *think* that means the difference between reset and power on. Well, the first part ran perfectly fine. The rest of it may still be a problem. I think it's going to be a long process getting everything fixed up, but at least I know how to do it now. Thanks for the help.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Strange. To provide a few more details: I had been working on a TAS that's now 7 minutes long and realized there was something that could be approved right at the very beginning. I made two FCM files. The full game without the fix, and the beginning segment. I converted to FMV to append the rest of the full game data after the beginning, but it caused a desync during an area where it's still at the beginning. I checked frame-by-frame between the unedited version and the new edited version (converted back from FMV) and the input is exactly the same. I thought maybe it was nesmock messing with header data that was causing it. So I tried this: I took the original movie of the beginning segment and made a copy, then modified the "length of controller data" and "number of frames" entries, then just pasted the controller data from the broken video to the original. Once again, the movie desyncs (although differently this time), even though the input is exactly the same. The input is the same right up until past frame 1500, and it desyncs around frame 900 or so. What is causing this?
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Comic guy never actually said that....
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Okay, I watched some of it. I guess the "missed shots" thing is what throws me. The movie doesn't seem so bad, but I think it could be done better. I'm voting Meh.
Post subject: Hex Editing and Nesmock
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
I've had some bad luck with Hex-editing. Using nesmock to convert to FMV and back to FCM seems to have altered important parts of the FCM file. (I did make a backup so my work isn't ruined, however I still want to figure out how to fix it up.) The input seems to be exactly the same, but it causes a desync somehow. Opening the old FCM and the new one in a hex editor shows me that things such as the md5sum, rom information, and FCEU version information may have all been altered. (Maybe because of an old version of nesmock? I used 1.4.0 because it was the latest Windows version) Also, it looks like it changed the flag that says the movie starts from reset, even though they both start on the same frame. How would I fix this?
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Cool. Voting yes.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
It's been years since I played this game and I don't remember everything. First, it seems like the reason you avoid fruit is because it adds extra time to your timer, which delays you as your score is added up at the end of the level, is that correct? My only other questions: 1. Why do you avoid certain eggs at the end of each level? 2. Why did you dump that blue dude in the lava in that one level? It seemed to slow you down a lot for that small stretch at the end. 3. There are a few points where it appears you collide with walls that may have been easily avoided. Mostly it's with your dolphin-like creature, but it seems like the up/down controls for it are difficult to maneouver, so I can understand if there's no way to miss them without drastically slowing down. But a few times near the end you do so with your pterodactyl when it doesn't seem necessary. Why is this? Oh, and by the way, saying that the game is too linear is kind of a dumb comment to me. Half of the games on the "recommended for first-time viewers" page are quite linear, especially Super Mario Bros. 2.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
What the?? Okay, I did the hex-editing. All the input all the way up to frame 1724 or so is identical in the new movie. Yet, it causes a desync before that frame. (The pass spawns in a different place) I don't understand the problem. Every single frame, up until past that point, is exactly the same. Edit: Hmm. Looking at the FCM files in a hex editor shows that stuff other than just the frame count has been messed with. Maybe a problem with the old version of nesmock? By the looks of it, it no longer considers the movie to start from a reset (even though they both start on the same frame) and the md5sum of the ROM, rom name etc, are all rearranged. Lucky I made a backup, eh? So how would I fix this?
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Prince of Persia: Sands of Time.
Experienced Forum User, Published Author, Active player (405)
Joined: 3/22/2006
Posts: 708
Okay, I'll attempt to do so now. By the way, Randil, I managed to beat your startup by five frames, so I save 190 instead of 185. Let's hope everything comes out okay. I'm always very wary of doing hex editing. Making a backup just in case. Edit: Wait a minute.... You need a C++ compiler just to run nesmock? Am I missing something? Edit again: Wait, I see it. The latest version for Windows is two versions ago. Is there anything in the latest two versions that are crucial?
1 2
22 23 24
28 29