Posts for Warp


Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
I have been having all kinds of weird problems with the forum. I used to read the forum from two different computers and there seemed to be problems in their synchronization (which is not odd since seemingly the state is stored into cookies...). However, I happened to delete the cookies of the forum in one of the computers by accident and since that happened it has been completely chaotic: New messages may or may not show as new messages, seemingly completely at random. I would say less than 10% of new messages really show as new messages. Sometimes messages I have already read reappear as unread even though there are no new posts to it, etc etc. Seems like this phpBB thing is not yet perfect...
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Makou wrote:
I imagine that there are some people, like myself, who play a game on an emulator as if it were the real thing. That may have been the case with that video.
Playing is a completely different thing than making a timeattack (or even a speedrun). We (or at least I) are not interested in seeing someone play the game.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
How about making a one-player timeattack of battletoads? After all, we are seeking for the fastest time here... By the way, battletoads has the best parallax-scrolling I have ever seen in a NES game. Technically very difficult to implement in NES. Cool.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
qFox wrote:
if the playback file only records keystrokes, then how random is kirby?
There's no randomness in NES games. I don't think there's any way for a NES game to create purely random events (a system clock which runs independently of the CPU may allow doing this, but I don't think a NES has that, and even it it has, the emulator will probably always start it from exactly where the game started anyways). What you see may look random, but isn't. Everything is consequence of your keystrokes (either their timing, their order, or both). If a NES game wants to make something look random, it will have to use the only thing that resembles randomness as its seed: Your keystrokes. Since the emulator will replay your keystrokes with 100% fidelity from the fmv, all the "randomness" in the game will be replayed exactly the same way as well.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
The Quake Done Quick speedruns are still considered speedruns even though they were made level by level (you can think about it as if they were made by making a quicksave at the beginning of each level). In certain games, such as Quake, a certain amount of "quicksaving" (in this case between levels) are allowed and still not considered "cheating" even by these purists who don't tolerate any emulator tricks. The most important thing about QdQ speedruns is that each level has been played "honestly", that is, realtime real playing without any aids not normally available in the game. That is, what you see is exactly what the player actually played. (Of course the player played the level hundreds if not thousands of times to achieve what you see, but the run you are seeing is still one contiguous real-time achievement). This raises a question: If for example SMB was done in the same way, that is, no "cheats" (if you like to call them that way) are used except for quicksaves between levels, would that still be called a "speedrun with no cheats"? If no, then why not?
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Phil wrote:
Solstice is now re-encoded is it ok?
Yes, it's much better now.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Phil wrote:
I don't think it will be better than 256 colors.It's illogical.
I'm talking about file sizes here...
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Bisqwit wrote:
Btw, when you make screenshots from a NES emulator, better save them as PNG or GIF because they are much better suitable for repetitive low color images than JPG.
And remember to convert the image to 256 colors before saving to PNG. If you save it at full color (ie. 24 bits per pixel) the file size will usually be like 5-10 times larger than an equivalent 256-color version (I don't know exactly why, but that's what I have experimented). Since NES uses only 64 colors, you shouldn't be losing anything in the conversion (in fact, converting the image to 64 colors instead of 256 could theoretically be even better, but perhaps it's a good idea to use 256 instead in case the emulator adds more colors than 64 to the image for some reason). Just be sure to use a good image manipulation program to do the conversion (such as the Gimp). Some crappy programs will make a pretty poor job.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Bisqwit wrote:
I'm not using divx to encode the videos, but libavcodec
That may be so, but libavcodec is generating an AVI which claims to be a DivX-encoded one (meaning that when you play it in Windows the player will try to decode it using the DivX codec by default). Anyways, I would suggest that you sample those two videos yourself. I would like to see them with the same high quality as the other videos.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
I was able to view the videos by changing the FourCC codes from "XVID" to "DX50" with a hex editor (which makes the player to play the video using the DivX codec instead of the XviD one). Worked fine this way, but I wonder if I'm the only person downloading these videos who would need to use this kind of hacking to view them. The videos themselves seemed a bit odd. First I thought it could perhaps be a problem with my system (eg. trying to play a XviD-encoded video with the DivX decoder), but I consulted with Bisqwit and it seems that the problem is in the original video and not a problem with my codecs/players. The problem is that the video seems a bit jerky. This is almost not noticeable, you can barely see it, but it still is annoying. It looks as if the sampler used to capture the videos was dropping frames. As a rough estimate it looks to me a bit like the sampler was told to capture the video at 30fps but it could only get about 25 frames per second (meaning that each second lacks about 5 frames or so). This causes a slightly noticeable jerkiness (the sprites seem to very slightly jump forward a bit several times per second). This effect is certainly not present in any of the other videos (including the previous solstice videos).
Post subject: Can't watch XviD (arkanoid, solstice v8)
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
If I try to watch arkanoid-timeattack-warpless-genisto.avi or solstice-timeattackv8-exim.avi, my system fails horribly. Winamp crashes with a segmentation fault, divxplayer is able to play the sound but the video locks up windows (win98) so badly that I have to reboot, VirtualDub can open the avis but can't process the images because of an unknown error (-100) and it guesses the avi is corrupt. There's a difference between these two avis and all the other timeattack avis: While all the other videos have been encoded with DivX (their FourCC code is "DIVX"), these two videos have been encoded (by someone else than Bisqwit) using XviD (their FourCC code is "XVID"). I tried to update my XviD codec to the latest version, for no avail. Unless someone can give me some ideas on how to get these videos working, it would be really nice if they were encoded in the same way as all the others. Granted, DivX is more or less proprietary and partly commercial (even though free third-party versions exist), but it has one advantage over XviD: It has a quite long history and is more supported, and in my personal experience is more stable (even though not perfectly). OTOH the XviD developers seem to have some odd principles: "No, we don't distribute binaries, only sources. If you want binaries, hope for someone to compile and test them for you, but don't expect us to help you if something goes wrong". "No, we don't know how to watch videos encoded with our codec, go somewhere else." They don't have official binaries, they don't have official support, they don't have anything. Seems like they simply don't care. A really odd principle. DivX might not be perfect, but at least they offer you something to use.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Bisqwit wrote:
So if you use a videocamera in a movie theater to record the movie, do you have all rights to distribute your tape? That's what these AVI files are.
In that case it's the author(s) of the game who owns the copyright to the AVI files? After all, you are distributing video footage of the game. If this is the case, if you want to go the legal way, you'd have to ask the author(s) of the game for permission to use the AVIs.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
I have been wondering about the copyright of these videos in the legal sense. If the definition of copyrighted art extends to these videos, who exactly owns it? The original game is copyrighted by its creators (admittedly we are on shaking grounds in this matter, but that's another matter). One could argue that the contents of an fmv file have automatically a copyright by its creator because, even though it's not an usual form of art, it's still a piece of original work. However, an AVI video made from the fmv does not contain any of the information contained in the fmv. Does the author of the fmv have copyrights on the AVI video created by someone else? Think about someone taking a photograph of a statue: The author of the statue has copyrights on the statue itself, but he certainly doesn't have copyrights on this photograph made by someone else. Can this same analogy be used with the AVI files? An AVI is just video footage of what the fmv file causes the emulator program to do, but in itself does not contain even a single byte of the fmv file itself. So, who owns the copyright to the AVI file? My best bet is the creator of the AVI file.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Freaksh0w wrote:
Thanks for just assuming that I'm an idiot however.
He was not assuming you are an idiot. He was assuming that there were no seeds, as you claimed. He and many others have a script running which look each 10 minutes if there are any downloaders for any files without seeds and if so, they automatically start seeding it. Thus if there really aren't any seeds, you just have to wait at most 10 mins for one to appear.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
ZtanZ wrote:
should ask someone with BBB (swe) to host it instead-- maybe could be faster and you wouldn't have to pay for the bandwidth :x
I don't think he is paying anything extra because of the high traffic. That is, even if his traffic dropped to 1/100 of what it's currently, it would cost him exactly as much as now...
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
I want to be honest with this timeattack: In no way I want to despise the efforts of the author of the bubble bobble timeattack, but I have to say that this is the only timeattack so far where I got bored. I watched the first about 50 levels and then started to just skip forward. I have never done this with any timeattack until now. This is not really because the timeattack itself would be boring. It's because the game itself is much more boring to watch than to play. The levels are way too similar to each other and there's nothing really new happening in the levels. They are too much alike. I could watch a 30-level run of this without losing interest, but a 100+ level run is way too much.
Post subject: Re: DVD player
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
blip wrote:
New patch. Supersedes the previous 240-scanlines patch.
Does this add the 60fps support as well? If not, could someone make a final single patch which adds all the fixes solved so far? It's getting quite confusing with tons of patches coming and going.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
I believe that trying to level up by killing smaller monsters takes more time than just killing the boss. In my 2nd timeattack I killed more at the beginning (ie. from the start to the forest boss) but it didn't really help at all. In this 3rd timeattack I made this beginning 15 seconds faster and it didn't take any longer to kill the boss. By the way, I tried to make the trick where the forest boss walks into the wall and gets stuck, but I didn't succeed. It just didn't walk into the wall. Does anyone have more specific instructions?
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
WalkerBoh wrote:
When you visit the gods, like getting the rope, if you jump down there ain´t u being transported back to the beginning? Dying of course but this could maybe save time somewhere?
Even if this saved time (which I doubt), I would be opposed to use dying in this way to save a bit of time. In fact, I don't like dying to speedup the timeattack regardless of whether it saves a few seconds or 10 minutes. If it saves just a few seconds then it's just not cool; it makes the play look imperfect (even if the viewer knows that the death is intentional). If it saves 10 minutes then it's not much different from using a cheat code to jump right to the last level of the game, ie. not very much fun to watch. After all, the goal of these timeattacks is perfection. Dying does not feel perfect.
And that flying glitch. I have just read about it here and I didn´t get the .fmv file to work with that. Anyways, is it possible to perform such flying in the "robot land"? If so it would be a great way to boost Rygar to kick other bosses more easy.
The flying glitch is too slow for almost any place. There aren't many places where you could skip significant parts of the level by using it. One place, though, where you could use it is in the spider cave level. There are places where you could skip about half the level if the flying glitch works there.
And finally, this is a rumor from some childhood friend. I don´t know if it works or if it could be useful but I can mention aswell as not. Anyways, my childhood friend claimed he had tricked the first boss out the door to the left. Don´t know if that was "killing" it or if it was easier to fight outside the actuall bossroom.
I haven't tried it, but as far as I can understand, you need to just jump on the monster and wait for it to walk into the wall. This sounds way too slow (specially because you still need to kill it after it's stuck, and that takes time too, even if you don't have to dodge the fire blasts anymore).
Just saw another thing. After defeating the spiderboss you head up for the next castle. Once over the water there are rocks you have to walk around. Is it possible to jump around them?
Yes. This is my bad. I didn't notice that shortcut when I made the timeattack. It would save several seconds.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Bisqwit wrote:
WalkerBoh wrote:
It was ok in the speedrun of Solomons Key
I'm not sure whether it was ok there either.
Personally I would prefer a Solomon's key run without the death, but that's just my opinion. :)
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Truncated wrote:
12) The Sky Castle has an interesting bug. Instead of walking east on the courtyard, go north a bit and then jump left. You would expect to die, but instead you land on top of the castle walls (!) due to some wraparound glitch.
That's certainly something worth studying. If it works, it's fortunately at the end of the game so it's easy to remake just the end of the timeattack, keeping the rest the same.
13) Just a thought: Does the game really check for all five items to give entrance to the sky castle, or does it assume that you have the rest if you have the flute?
I read a walkthrough of the game and it said that you must have all the objects in order for the rope to appear (ie. the flute alone does not suffice). Even though I didn't actually check, I trusted that this is so.
14) One source (http://www.gamefaqs.com/console/nes/code/8096.html) claims that you can fly by holding autofire on A and B as you drop into a hole. I don't have autofire so I can't check this, and I don't have a specific idea for where it can be used, but anyway. Being able to fly can't hurt.
Rygar can indeed start "floating" with the autofire trick when walking off a cliff. However, this floating is too slow for it to be useful anywhere, as far as I can imagine. Interestingly, I found a quite funny glitch in the game using this floating trick. I don't think it's usable in the timeattack, though, but it's funny. You can watch the glitch if you have famtasia with this fmv: http://www.cs.tut.fi/~warp/rygarglitch.zip Perhaps Bisqwit would be so kind to make an AVI from it for those who can't view it directly?
15) The same source also claims that if you stand next to the door at Belzar and face left as you are hit, you will die and then be teleported to Ligar when you press start. This sounds absurd, but you never know. You could skip three castles if it does...
I'm not comfortable with the idea of dying in order to speedup the timeattack. What do other people think about it?
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Truncated wrote:
2) I'm fairly certain the thump signals damage dealt. Isn't the enemy flashing between the thumps? In any case, there is one thump when you get a hit in on the way down and two thumps while standing. Sometimes you missed the one on the way down.
The problem with those bosses is that they don't only shoot, they also advance from time to time. This means that I must retreat in order to not to get hit (or start jumping on the monster). Retreating with the fire blasts floating around is quite tricky manouvering and it's practically impossible to shoot at the monster at the same time as you are retreating and trying to not to touch those fire blasts. You are right in that the boss fights could probably be optimized a little more, but I am quite certain that not very much.
Also about this boss, I read in a faq that you can trick him into the right-hand end of the screen where the door later will be, so that he becomes stuck and cannot fire. Then you can pound on him standing still. I don't know if this is quicker or not - have you tried it?
I have read about this trick as well, but I don't believe it helps much because the monster advances so slowly. You would just have to be jumping on it and wait until it walks into the wall. You can't hit it while you are jumping on it. I haven't tried it in practice, though, but I believe that it would take way too much time. But perhaps I should try it, just for the sake of it. :)
8) The snails are hell. I remember always dieing there. But as long as you get the jumping right, I think the only thing you need to do is to hold down autofire early enough, so you hit them on the way down too. You could hold down autofire all the time really, it looks a bit sloppy but it works.
The snails have the same problem as the forest boss: They advance, forcing you to retreat. It could perhaps be optimized a bit further, but probably not too much.
9) Alright, I understand the thing about the flying moster now. And yes, I had something concrete in mind: the bosses. A lot of time is spent on jumping or otherwise avoiding shots, and not doing damage.
The bosses also inflict more damage than the smaller monsters. Usually a hit takes two energy units away. Another problem is that when you are hit, you are thrown back (this doesn't happen when you are hanging from a rope, though).
11) Another thing is that you can use the flute on the last boss to remove its arms for a couple of seconds. I don't know if this makes him shoot less or if it's just a graphics bug. Also the menues have a delay so I don't know if it's worth the effort.
Didn't know about that. If it makes the boss to shoot less, then it may actually help.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Even though I have made (three versions of) the rygar timeattack, I have to admit that I might not know 100% of the little details of the game and there perhaps may be some things I have missed. But about your suggestions: 1) Unlike in many other NES games in Rygar the items are not dropped by the monsters randomly. It seems that each 6th-10th (or something similar) monster drops an item and there doesn't seems to be any way of influencing this frequency. You can influence which item they drop (star or bottle) with a small probability, but that's it. In the timeattack I had to sometimes make rather long parts again because the monsters were dropping a bottle when I really needed a star. No simple variation in movement seemed to affect this, but when I played a longer part again, it changed to a star. 2) Honestly I don't know if the thumping sound is directly related to the monster getting damage. It probably is, but the first time I tried killing the forest boss I did it without autofire. The second time I did it with autofire, and it died almost twice faster. It may be that the second time the autofire helped make more hits in the same time (even when you can only make damage at certain intervals). Anyways, it's very difficult to optimize those triple-fire-thingie-throwing monsters because you need to constantly predict when will the next one be shot to jump over it. If you jump too late, you get hit. I'm not saying it's not possible to optimize those boss battes, I'm just saying that if it's possible it's not trivial... 3) If this was a smb-type timeattack where you are optimizing even single frames away (which IIRC are about 1/20 of a second in famtasia) that would certainly be true. However, since this is such a long game and the random nature of the game may cause several seconds slower or faster runs each time anyways, those fractions of seconds probably don't affect the overall time in any drastic way. It's certainly true that I might have saved fractions of seconds here and there, but they all probably sum up to a few seconds in total (which could then be lost with an unlucky random monster placement later or whatever). Rygar is quite a laborious game to make a timeattack for... :) 4) Rygar is too close to the wall at the start of the spider boss screen, so he must run some distance to the right for the ducking tactic to work. Also, you can't hit the spider right away, when it's coming towards you because it will change its direction and start going away from you, thus you have to wait for it to get to your left. I don't believe running a bit to the right, ducking and waiting for the spider to get to your left saves any time compared to the jumping (or if it saves, it's some fractions of seconds). 5) This might actually be true. I have never tried it. Could save a couple of seconds if it works. 6) The Dorago boss is difficult as hell. In the first timeattack I made it took me (IIRC) more than an hour to defeat it, and it resulted in a 3min 20secs battle. In the second timeattack I optimized the jumping better and was able to drop it to 1:40. In this third timeattack I was able to optimize it even further to 1:20 by not jumping so much to the other side. It may be possible to optimize it a bit more, perhaps even 5-10 seconds, but it's a very difficult one. Those projectiles make it quite difficult. 7) I thought it was impossible too, in my first two timeattacks. In this third one I tried it again, just in case, and to my big surprise I found a tiny spot there under the rope where you could jump from and it would reach the other side, which saves quite some seconds compared to having to go around the river. The second time I had to cross the same river in the same direction I just didn't find the spot even though I tried dozens and dozens of times, so I just gave up and used the rope instead. (Btw, if someone doesn't know, the rope can't be used until you get the hook from the end of the forest level.) Everywhere else I have tried jumping over that kind of gap I have failed. If there are other places which such a peculiar feature, I was unable to find them. It might have something to do with the rope itself: Perhaps the game engine adds something there where you can stand on and which usually isn't there. 8) The snails shoot more often than the forest boss and they are even more difficult. It might be possible to optimize those too, but it's not trivial. 9) I never take damage on purpose. I always try to find a way to run without taking damage without slowing down more than perhaps a fraction of a second. The 15:34 situation is much trickier than it may look from the video. The flying monster is too high to be hit while standing, so the only way of hitting it would be to jump. If you jump, however, you must just jump on place, without advancing (because else you just fall to the gap and die). It's not possible to jump the gap, you must use the rope. Also, the flying monster is too far away to be hit if you jump before the rope (and try to hit while falling back to the ground). I tried tons of things there, but couldn't find any way of solving it fast, so I decided to take the damage instead of spending more than a second jumping without advancing. My principle has always been to avoid taking damage unless the only way of not taking it would be to waste too much time (I consider even 1 second to be too much). As for taking damage on purpose in other places, do you have anything concrete on mind? I'm not aware of any place where I could have saved time by taking damage. And as I said earlier, you can't control the frequency of the items. Even if taking damage would speed up something, you can't get the bottles whenever you like. 8 stars is a must in the timeattack, 3 for the "power up" and 5 for the "attack & assail". You can get a total of about 10 or 11 items in the timeattack, but you really don't have time for more. For those who don't know: Rygar can cast three types of magic: Power up, attack&assail, and recover. "Power up" takes three mind points (which you get with the stars, that is, you need to collect three stars in order to cast this spell). You can cast it only once, but it lasts for as long as you are alive. It makes your hits stronger and your shield to reach farther. For this reason it's very important to get three stars as early as possible, definitely before the first boss (the forest monster). Once you get it, you have this boostup for the rest of the run. (And as I said, you can only cast it once, so you don't get any additional boostup by casting it again later.) "Attack & assail" gives you 10 hits to every monster in the screen and it consumes 5 mind points (meaning you need to collect 5 stars). Even though this would be handy for all the bosses, you simply don't have the time to collect the required stars. In the timeattack I reserve this spell for the final boss, which is the most difficult one. 10) It's usually unnecessary, and I didn't do it in my first nor second timeattack, but I had to curse the entire family of that **** boss because no matter what I did (and I even tried to play that entire level leading to that boss again) this time it shot two of those fire thingies right at the beginning. I certainly tried everything I could to avoid them without losing time, but for no avail. What you see in the video is the *only* and *fastest* way of avoinding them which I could find. Honestly.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Bisqwit wrote:
I hope you don't use so much left-right-left-right wobbling this time. It doesn't really look cool - especially if done all the time.
I agree that it's disturbing and annoying. I would certainly prefer if the character was honestly standing still when it has to wait than do that annoying flickering.
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Jumpman wrote:
The only movie that i didn t like was morimoto's speed upgrade gradius because you could tell it was fake and it wasn t fun to watch.
I have never understood (and probably never will) this obsession in using the term "fake" in this context. If Morimoto had claimed something like "this is a genuine play which I have been practicing for years and what you see is exactly how I played it in real-time" then you can certainly call it a "fake" because that would be what it is: Morimoto would be lying through his teeth and the video he is showing you is not what he claims it to be. However, I don't think Morimoto has ever said or even implied that (well, who knows, perhaps he actually has, but I find that unlikely). If you say, however, "look at how this game can be played when a computer is told how to play with almost superhuman reflexes" then there's absolutely nothing "fake" in the video. It's showing you exactly what is promised: An almost perfect, almost superhuman play, which is theoretically possible with a real NES with the real game. The video is certainly genuine.