I personally hate the "wobbling". It's distracting and annoying and doesn't add anything to the video. I wish people would stop doing that.
If you must wait, then honestly wait. Don't start flickering the sprite for no reason.
Say, does anyone know of any "frame-by-frame" mode for making movies? A lot of people here talk about doing something "at the earliest possible frame", but I'm not quite sure how to find an exact frame while I'm playing because I can only slow a game down so much and I can't see the frame count during play.
I don't think they are talking about making the run frame by frame when they say that.
Making a movie "frame by frame" was actually an urban legend widespread with all the lies told about Morimoto's SMB3 run. I don't think it's possible to make a timeattack "frame by frame". You can slow down the emulation quite a lot (I think even a speed of 5% is still feasible using the 60fps patch for famtasia, but afaik lower than that makes it just too difficult), but that's not the same thing as making it "frame by frame".
When they are optimizing runs like SMB1, the already-existing runs are already so optimal that they are near perfect, and almost any further optimization will be so small that it's not feasible to measure it in seconds anymore. Since famtasia tells how many frames exactly the movie is long, that's a much more accurate measurement. If you can make the run with 2 fewer frames than previously, you will know that it has been improved. You may save a frame or two with extremely precise movements, like jumping just a fraction of second earlier (possible when slowing down the emulator) etc. I think that "you can save a couple of frames here" refers to this.
IMO a tool-assisted speedrun (can we finally use this term instead of "timeattack"?) of a modified ROM is ok only if the modifications do not make speedrunning easier compared to the original game. For example modifying a ROM so that you get immunity to enemy hits or that you can walk through walls (in such a way that is completely impossible in the original game) is absolutely not ok.
However, a modified ROM which is basically a new game based on the original game, for example by changing the original game graphics and texts, should be ok.
One sad thing about this is that there are often situations which are very complicated and as you say need hundreds of rerecords and takes lots of time until you finally get it right, but in the final video that part just looks easy (because that's often what results when making a perfect run: it looks very easy) and you can't really appreciate the difficulty of that part and how much effort the timeattacker spent in it.
For example in my Rygar run at the end of the spider cave there's a section of the cave where there are lots of flying monsters and gaps and you have to jump a lot while not getting hit by the monsters. It took me tons of time and rerecords to get it like you see in the video, but it doesn't really look all that difficult.
Those who oppose tool-assisted speedruns often say that emulator slowdown and free rerecording makes it easy to perform such a run. Well, they are free to try. It's only when you try when you really realize how difficult and laborious it can be.
xbox is hardly a 128-bit console because it uses a Pentium-III as its processor, which is a 32-bit processor. The same is probably true for many of the other consoles (even though some of them may use 64-bit processors).
Don't confuse the memory bus width with the internal register width of the processor. The memory bus can be 128 bits wide even though the processor is a 32-bit processor (meaning that its registers are 32 bits big).
As for the original question, the bit "size" of a processor usually refers to the size (in bits) of its internal registers. Internal registers are used for integer arithmetics (additions, substractions, etc) and thus the bit "size" describes the maximum limits of integer operations of the processor.
For example, the intel 80286 was a 16-bit processor: Its registers were 16 bits wide and it could perform only 16-bit integer arithmetic (with a single machine opcode, that is; naturally you can perform arithmetic with larger numbers by making operations on several registers with several machine opcodes). That is, you can for example calculate 200+200 (=400) in one single machine instruction in a 80286, but not 40000+40000 (=80000) (you would need to use two registers and two machine opcodes to calculate the correct result of that addition).
The 80386 is a 32-bit processor (which can be run in 16-bit mode for compatibility). When in 32-bit mode its registers are 32 bits wide and you can perform 32-bit arithmetic (for instance you can calculate 40000+40000 in one single opcode). (Actually you can use the 32-bit registers in 16-bit mode as well, but with a speed penalty.)
"8-bit" computers had, thus, 8 bits wide registers and you could only perform integer arithmetic with that limit (ie. values from -128 to 128, or from 0 to 255).
However, I think the definition is fuzzier with these computers than with current computers. For instance, the Spectrum is considered a 8-bit computer even though it had 16-bit registers and you can perform 16-bit arithmetic with them.
In my opinion adding a overly lengthy explanation at the beginning of each video is needless and in fact annoying. What I mostly feel when I would see that would be "yeah, yeah, I already know that, I don't need to be reminded every time, don't insult my intelligence".
After all, the suggestion of adding lengthy explanations to the videos is caused by stupid people out there. Are we going to add an annoying disclaimer just because all the stupidity?
A short disclaimer is ok, like "This is a tool-assisted run; for more details see http://..." (similar to the current one, but shorter). If this is not enough for the stupid people, then that's their problem, not ours. We gave them enough info for any person with a minimum of intelligence to handle. That's enough.
How about "optimal run"?
By the way, I think the term "timeattack" is not that good. Instead, the term "tool-assisted speedrun" is much better and has actually been in use in the speedrunners' world way before these nes videos. For example, there exists a tool-assisted speedrun of Doom (made many years ago) which used freely savestates to get an optimal run (but didn't actually always get it optimal, as a later genuine speedrun proved... :P )
Since this will require adding additional code to the program, not modifying existing code, I think it's quite unlikely that this will ever happen.
I'm not saying it's impossible (it has been done, eg. with many (s)nes rpg translations), but it's very difficult.
In that case he should really explain why he wants the videos in a closed proprietary format instead of a more open and widespread one.
Being closed and proprietary causes its problems. For instance some programs don't support it because of licensing issues (VirtualDub is a good example of this).
And besides, what would be the advantage of using WMV over AVI in this context? It's not like Windows Media Player was unable to show AVIs.
-Emulator time attack videos do undermine legit runs despite their intention to entertain.
You can't please everyone. Many people enjoy watching these "perfect" runs which are theoretically possible but in practice impossible for a human to do. Some of these people might find it boring to watch a "legit" video with tons of imperfections.
I'm not saying speedruns are not enjoyable. In fact, I enjoy for example the QdQ speedruns a lot and would actually find them more boring if they were tool-assisted.
However, IMO most nes games are different from Quake: Most nes games simply don't offer the same possibilities to show pure skill as Quake does. For example, imagine a "legit" speedrun of Castlevania 2: Since you need to collect a quite vast amount of hearts and in a speedrun it's practically impossible to control the randomness of the dropped hearts, the vast mayority of the game would be watching the player killing zombies for the hearts (which would take probably 10 times longer in a speedrun than in the timeattack). This would be extremely boring and certainly uninteresting. However, a timeattack can make an interesting video.
-People are misinformed of how these videos are made, so they are accepted as legit runs. Emulators can be used to create cheated runs that can pass themselves off as world records.
The problem is not so much that people are misinformed (which is something we can't avoid, no matter how many disclaimers we include to every possible place) but the reaction of people when they find out: Instead of thinking "ah, I understand now, silly me" they will often start a holy war against the "cheaters".
When I first saw the SMB3 timeattack by Morimoto, I did not know anything about timeattacks, but I immediately realized it was not a speedrun, but a tool-assisted run (it's quite obvious IMO). I was not disappointed at all, but I enjoyed it. Too bad most of the people are not intelligent enough to realize the same thing.
-Emulators allow poorly skilled gamers to produce videos that are faster, better, and far more entertaining than pure speedrun attempts of highly skilled gamers.
Is this a bad thing?
Besides, timeattacks is not gaming. It is its own skill in itself. You may not need gaming skills to make a timeattack, but you need lots of other skills, and making a good timeattack requires its work. If making a perfect run would be easy there would not be lots of newer and newer versions of timeattacks.
An unskilled person making a timeattack of almost any game will most probably get a much much worse result than the currently existing timeattacks in the nesvideos page. It's not thateasy.
-Using savestates and slow-motion is in fact cheating.
I disagree. It's cheating only in a certain context.
If people were competing on who can make the fastest speedrun, then a timeattack would indeed be cheating. In a speedrun there are written and/or unwritten rules, and a timeattack would blatantly break many of them and thus would be cheating.
However, why would a timeattack be cheating in the context of timeattacks? Of course it's not cheating: It's doing exactly what is promised: A tool-assisted run as fast as possible through the game. There are no cheats, there's no faking.
It would be possible to cheat in a timeattack as well: For example, you could use a patched version of the ROM which gives you eg. immunity and would allow you to make things no-one else could do with the unmodified ROM. In this case this would not be a legit timeattack and would be cheating.
The term "legit timeattack" is not used too often, and IMO it should be.
You cannot do the preceding on a regular NES and it gives you an unintended and unfair advantage. An advantage that legit players will never have.
Advantage on what? We are not making speedruns.
-Time attack players have too much pride to honestly mention in their videos that the games have used slow-motion, savestates, and are illegitamate records produced with virtually no skillful play. Being clever is not a skill unique to time attacks.
This is bending the truth way too much.
Not having a 10-seconds initial screen in each video explaining every little detail on how the video was made is simply a question of practicality: It's not practical to have the same thing on each video, and once you have seen it, having to watch the same thing each time is annoying.
Besides, going too far in making sure the viewer really really really knows what a timeattack is can insult the viewer's intelligence. If the viewer is so stupid to believe it's a speedrun and when he finds out it's not and starts a flamewar, then it's simply that: Stupidity. We can't shepherd everyone like little children.
can these files be converted to windows media player?
I would like to know what exactly do you mean with this. What does "converting to windows media player" mean?
Windows media player is just a multimedia player. It doesn't have its own format. It can perfectly play the AVI files which you get from the nesvideos site, as any other multimedia player does (assuming you have the correct codecs installed in your system, which is something completely non-related to wmp itself).
If you mean that you are having trouble with the .torrent files, you should be more specific on what troubles you have. BitTorrent is basically just a way to download the files, it has nothing to do with playing the videos themselves. (If you are suggesting that the site would provide the AVI files directly instead of using BitTorrent to share them, then you should be prepared to download an AVI file for a couple of months instead of a couple of hours like currently.)
Maybe proggesional was not the right word, but you dont need to jump all over him because of it.
"Jumping all over him" may, in my opinion, be justified.
When he said that his video looks more professional he directly implied that the timeattack is amateurish and imperfect. Regardless of whether he did this on purpose or not, that was a direct insult to the author of the timeattack. In practice he is saying "I can make a video which looks a lot better than yours, thus your video sucks", and he did not mean that he can make the timeattack faster, but more "professional-looking", whatever that means.
By this he did not only insult the maker of the video in question, but all timeattackers. He implied that making a timeattack is amateurish and looks bad (umprofessional), and that making it "the right way" is much better. Thus he was (again, consciously or unconsciously) despising what we are doing.
He might not have done it on purpose, but he did it nevertheless. Telling him his mistake is justified, in my opinion.
Warp suggested using Antonio Vivaldi's Summer- Presto song. I liked the idea. He apparently couldn't find this thread again to come here to say it...
Why is it ok to use that song? Even though the original composer is long dead, isn't the song played and recorded by another artist? Wouldn't we have to get permission from him before using it? Plus, I'd rather use Storm by Vanessa Mae, which is a redone version of Summer.
If you for example make your own midi version of a copyrighted song and publish it on the internet you will be violating the copyright laws of most countries. However, those classical songs are not copyrighted anymore and anyone can make their own versions of them and publish them freely.
That's why it's technically illegal to use a song from a videogame (even if you make the midi yourself) but not from a 100+ years old classical piece.
I think Speedruns are Movies, where you can see all of the Game, and TimeAttacks are only, maybe, one level in a good time.
No, a speedrun is a run through the game as fast as possible without using any aids not normally available in the game (even though sometimes it's ok to eg. make the speedrun in a level-by-level basis, as most QdQ speedruns are made; you can think of it as making a quicksave between levels, but anything else being forbidden; the quicksave between levels is the only feature not provided by the game itself which is allowed).
A timeattack is also a run through the game as fast as possible, but any aid you can use (eg. provided by an emulator) is allowed to make a perfect run. These aids typically include slowing down the emulation, free quicksaving/rerecording, autofire, etc.
I may be a geek and a computer nerd (ok, I am), but personally find it cool to see how a game can be played through when a computer is told how to play it perfectly, without a flaw, as fast as possible, pushing the game physics to their limits.
And that's what these fmv files are: They are instructions for the computer to play the game as perfectly as possible. I don't care how these fmvs were created, the only thing I care is that they show how to push the game to its limits.
People often have difficulties in viewing XviD-encoded videos. If you use XviD, make sure it sets the FourCC code to DX50 (iirc XviD has an option for this) which makes the player to play the video using the DivX codec.
Another possibility, of course, is to use DivX directly for encoding.
I don't really understand what's wrong with XviD. Some XviD avis play just fine, without the slightest problems, while some of them can even cause windows to completely crash. I do not know which settings are "safe" in this regard.
I thought all you have to do is include the authors name and song title in the credits.
Copyright law does not work that way.
By the way, another suggestion for background music: Some classical music (like eg. Vivaldi or something similar).
but how could i get the movie to desync? i played level 10,11 and 12, but the movie was desync at level 9.... i don't want that to happen again...
There are many possible reasons for Famtasia to desync when creating an fmv file, some of them probably because of what the player did (eg. quickload the wrong save or something) and some of them probably because of timing problems (if there's lots of cpu or I/O traffic it may be that the internal timing of famtasia gets garbled).
Anyways, it's always a good idea to check the integrity of the fmv file from time to time (by just playing it from the beginning), for example between levels.
What I do when I want to check the integrity of the fmv is to make a normal quicksave (eg. at the change of level), wait for some tens of seconds (it doesn't really matter what you do during that time, it's just to give you enough time to quickload later), stop the recording and then play it from the beginning. If everything is ok, then after the movie gets to the place after the last quicksave, just quickload and continue. (But be careful to select the correct quicksave with alt+number; famtasia seems to always reset it to quicksave 1 when you start playing an fmv).
Before the check it's a good idea to make a backup of the fmv and every quicksave you have made. What I do is that every time I check the integrity, before starting playing I make a new subdirectory (eg. called "gamename timeattack backup n" where n is a number) and copy the fmv file and all the quicksave files there. If you blunder when quickloading (typically you quickload from the wrong quicksave, thus trashing your fmv) you'll be extremely thankful for the backups. If you blunder and your fmv is broken, just copy the contents of the backup subdirectory over the current files and do it again (but more carefully this time, of course).
If you want to download data from someone, you need to know his IP address. There's no way around that.
I don't understand what is it that you want. If BitTorrent didn't tell the user's IP address, you wouldn't be able to download anything. Where would you download it from if you don't have the IP?
If what you want is that everyone else shows you their IP but your client does not tell your computer IP to anyone else, that's quite selfish. Basically you will be only downloading and not uploading at all. (I don't know if the tracker has a way to limit this kind of misuse.) Also, if everyone would do the same, no-one could download from anyone.
hmm you and Bisqwit are thinking very strangely.Olympic games yooohoo it's competition.I think sports games are good for competition.Btw Bisqwit was consulted before I start Castlevania.He said to me that he doesn't care if I do one.So you and Arc should consult before writing such conclusion.
Please read my whole post. I said Castlevania did not fall into this category. I was speaking more generally to everyone, not just to you regarding to a specific event.
I have to say that all this "hey, I make a better score, why won't you publish it" attitude is a bit childish.
Different games have different goals due to their very design: Not all games are suitable for timeattacking simply because there's no time to beat. In some games it's not the playthrough time which counts, but for example the score you get (and often getting a higher score means that the complete playthrough is actually slower).
However, there's an important difference between games suitable for timeattacking and the other types of games where playthrough speed is irrelevant. In the first type of games playing through the game as fast as possible is exactly the thing which makes it awesome and entertaining: You often require so awesome planning, acrobatics, timing and precision that that in itself makes it very entertaining to watch.
However, in a game where score is the goal, making an entertaining video can be harder. If the highest score is achieved by repeating the exact same perfect achievement several times, that can get boring.
Track&field is a good example of this: In most cases you have three attempts to get a score. Making the exact same achievement in all three attempts is boring, and in this case even useless: The scores are not summed up, only the best one counts.
Thus it's only natural that you get the highest possible score once and do something funny and entertaining in the other attempts (if it's possible). If something like that gets you to laugh, that's a thousand times better than watching the exact same perfect achievement three times.
The goal of the track&field video is to show how funny things happen when you extend the physics of the game to their limits (or to show "hidden" features purposedly coded into the game, such as the falling ufo). Certainly in some of the events (such as archery) this is not possible, but whenever it is possible, that is the goal. Even though getting the highest score in each event is also a goal, the goal is not fastest playthrough nor highest sum of scores.
As for beating other people's videos, I think it's polite to consult first with the author of the original video, specially if he has put tons of efforts in it (which usually results in many versions of the timeattack), unless you can really greatly exceed his achievements with many truely innovative tricks.
If you have some idea which could save a couple of seconds in a timeattack, give the author a hint. It's not a nice feeling that you have spent so much time on some timeattack, polishing it as much as you can and making many new version of it, and then someone else submitting a 2 seconds faster version which overrides all your hard work without even having the chance of doing it yourself.
I'm not saying the castlevania video falls into this category. These are just some general thoughts about trying to beat other people's videos. Be considerate and don't just insensitively try to demolish his beloved "child" without even giving him a chance of polishing it himself.
If he does not want to implement the ideas himself, or you can clearly outperform him, specially if there have been only one or two versions of the video, or if quite many people have made new versions of the video, then it might be ok to just give it a try without asking. However, if a video has had only one author, and it's going on v8 or whatever, and you can only beat his time by a couple of seconds, then it may be considerate to ask the original author first...
I was finally forced to remove this video and the dragon warrior video from my hard drive since they caused software crashes, triggered strange windows bugs and more by their mere existence.
Try changing the FourCC code of the AVIs from "XVID" to "DX50" with a hex editor. Works for me when I have problems viewing XviD avis.
XviD viewing problems can often be solved by viewing the video with the DivX codec instead.
The only way I know of doing this (unless the DivX codec has some option to "grab" XviD streams for itself, which I doubt) is to change the FourCC code of the AVI file with a hex editor.
Open the AVI file with a hex editor. Near the beginning of the file there should be four characters which read "XVID". Change them to "DX50". Play the AVI normally. It should work now.
(The difference now is that the player is using the DivX codec to decode the stream instead of the XviD codec. Their format is compatible and they can decode each other's data.)
