Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
As you know, I can't take my hands of this game. It's just that it's so complex to TAS, each time you think you got it perfect you eventually find something you didn't consider in the previous attempts. This makes the game very challenging and fun to TAS. At the same time, the game is very edit friendly, so if you find a 1 frame improvement at the start of the TAS, you can be sure that you can edit this in without desynch.
The improvements for Deja Vu keep getting smaller and smaller. Perhaps this shows that I'm getting a lack of new ideas, or that it's getting closer and closer to perfection. I'd like to think it's the latter.
I have revised the published run once again. I've tried out some new ideas that I had - doing things in different order, picking up stuff in different order, new cursor movements, etc., but with little success. Non the less, I have improved the published run by 5 frames. Here are the improvements:
*I pick up Key2 before Key3. The reason for this is because the inventory order is a little nicer when I unlock the car door and the bar door with these keys. Although this is a noticable change, it only saved 1 frame.
*I use a slightly different cursor movement in the elevator after visiting the casino. This saved 3 frames.
*I corrected a mistake in the doctor's office, saving 1 frame.
Looking for mistakes in a run like this is very tedious. If you do it manually, you have to watch the run frame by frame, trying to spot if you stand still for 1 frame too much somewhere or clicked 1 frame too late on a text message. I solved this by writing a lua script that detects these kind of mistakes. It turned out that there were no text message mistakes in the published run, so no frames saved here. It was thanks to this script that I found the mistake in the doctor's office. Also, according to this script, there are no mistakes left in the run, so unless the lua script is wrong, this run should now only be improvable by new strategies and/or cursor movements.
I will look some more into improving this before I submit it. Perhaps when the submission queue is a little more empty and I'm fed up with looking for improvements, I'll submit this.
If someone wants to watch this, here it is.
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
I have a question about a script I made, namely this:
while true do
mf = movie.framecount()
if mf<500 then FCEU.speedmode("turbo") elseif mf==500 then FCEU.speedmode("normal") end
FCEU.frameadvance()
end
As you can see, the script is simple. When it's running, and you're playing a movie, FCEU will run at turbo speed up to frame 500, where it will change and start running at normal speed. I tried this script on FCEU 0.98.28, and it works just fine.
However, when running the script on FCEUX 2.1.1, it just keeps running at turbo speed even after frame 500. Somehow, FCEUX doesn't recognize the FCEU.speedmode("normal") command. I have tried changing FCEU to FCEUX in the code, but it didn't work.
Does anyone have any idea why this script works in FCEU but not FCEUX?
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
Sticky wrote:
The scenery looks really photo shopped, Randil :P
The sky reminds me of the opening to the Simpsons.
Huh, now that you mention it, it does look shopped... How strange. But I assure you it's not, I don't know how to use Photoshop, and I wouldn't go through all that trouble anyway. :) It's just a beautiful summer day in Sweden, I guess.
Shinryuu: <3
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
It's been a long time since I posted a picture of me here, so I thought I'd post a new one, taken today. I look a little weird in this photo, but only a little.
Picture.
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
I too have been thinking along the same lines. Perhaps it could be a good idea to have two different pages explaining how we make these videos and how the site works - one page for those with good technical knowledge, where terms like FPS, save state, RAM addresses etc. are used. The other page could be for those who don't know anything about emulators, where all these emulator terms are explained in a way that anyone can understand, and where the goal with our movies are explained in an easy way. Think of it as one page for computer geeks, and one page for those who barely touch a computer.
I guess it's inevitable to mention the "Beginner's TASing guide" project when talking about a beginner's section - perhaps we could fit one in that section. Something like one guide for each emulator that goes through all the basic features of that emulator, step by step, and how they are used in the making of TASes.
I also think that it's extremely important that newcomers see this section as soon as they come to the site. If they have to look around to find it, it doesn't really serve its purpose. Perhaps a large, red, flashing banner at the top of the main page as well as the forum? ;)
Overall, what I want to say is that I support this idea. :)
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
TaoTao wrote:
FYI: I have watched another run of this game (J version, "Kakefu-kun no jump tengoku speed jigoku") played by narimasa:
http://dic.nicovideo.jp/v/sm6972518 (fcm/fm2 seems not available)
Your run seems to beat that run successfully in most stage, but in a stage, that run seems to be a little faster than yours (watch from 10:00 of the movie above). But I have never played this game, so I don't know whether your run is improvable (I don't know the differences between J and U).
Nice, now I have something to compare to!
About that improvement, I'm feeling really, really stupid now. Someone should hit me with a stick or something, because I did a really silly mistake in this run. I didn't see that I could jump over the pole in 5-3, saving a whole lot of time.
In the light of this, and that I have a feeling that some of the boss fights might be improved further, I will cancel this run and start working on an improvement. Sorry for the inconvenience, but the next submission will at least be faster! (perhaps even under 14 min? who knows)
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
Lord Tom wrote:
Also, just found that I can, in fact, drop a frame in track 2, letting me start track 3 3 frames earlier, so it should be possible to achieve a (tiny) net benefit from the annoying counter behavior.
I think this brings up an important question about entertainment/speed trade offs for this game - personally, I'd rather see track 2 beaten 1 frame faster and wait 3 more frames between levels - I really don't mind those 3 extra frames between the levels. I think it's more important to optimize the actual play time. That's just my opinion, although now that we're talking about these very tiny time differences, I don't care much either way. :)
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
Wow, that's a lot of points brought up. Here are the answers:
Did the throw at the intro to 2-3 cost time at all?
Nope, throwing your furry friend never wastes any time. I did it here because when he zooms back to me, he kills a few enemies at the same time.
For the pole-shot-thing in 2-3, would you have been able to shoot at a lower angle and hit max speed faster, or does it launch at full speed as is?
I'm not sure I see what you mean here, but I'll try to explain anyway. Your acceleration after being launched by the pole works like this: The higher up your grab the pole, the faster acceleration you'll have. You'll see that I grab the pole at the very top, thereby gaining maximum acceleration. How high the pole launches you has no effect on your acceleration.
Why did you hop at the end of the second boss fight?
As stated in the author's comments, this is done to manipulate the herb to move from left to right, rather than moving from right to left. If I don't jump here in order to kill the boss slightly later, the herb would be moving in the othre direction, and it would waste more time to pick it up and then run to the end of the level.
If you maintain speed when falling from a platform, why didn't you start running from the ledge at the end of 2-3?
I'm not sure I know which place you're talking about here, so you might want to provide a around what frame in the movie this occurs. If you mean the jump around frame 11385, running down the ledge would make me miss the small platform that I land on after the jump, so in this case, running down the ledge isn't possible. But again, perhaps you're talking about a completely different part of the level. :)
Did running backwards in 3-1 serve a speed purpose?
The reason why I run back a bit at the start of 3-1 is because I need to have a large enough speed to make the jump from the first spring to the second spring. Had I just started the level by running forward, I would make the jump to the first spring, but not to the second one.
Why the pause in killing the third boss?
Same reason as the second boss, see above.
In 4-2, could you have hopped off the pole-shot-thing earlier and simply ran back to top speed?
This is yet another quirk with the poles - after you've released the pole, you actually accelerate just as fast you would have done when running on the ground. You can't release the pole earlier, you are stuck to it on a fixed time.
Why the pause at the group of enemies in the middle of 4-3?
As yes, this. The reason why I hesitate is that if I jump on the first enemy, I would scroll the screen up, so I want to avoid jumping on him. At the same time, I cannot hit him with Furry, because he is thrown slightly upwards when you throw him. I can't move straight forward either, because then I would run straight into him, and take damage. The only option left is to avoid touching him, which is why I have to wait until he has landed, and then jump over him.
Were the two quick screen transitions avoidable in 5-3?
Now here is one thing I did not consider. There is one more pole after that first one, so I'll have to grab one extra pole this way, but it's still faster. I redid this level from this point, and succesfully edited it in, and this trick saved about 80 frames. I also managed to kill the boss slightly faster, so a total of 106 frames were saved! Thanks!
Couldn't you manipulate the fourth and fifth bosses to move away from the pump by jumping behind them while the pump reoriented itself?
I've tried using this strategy on all bosses, but it turned out slower in all cases.
Why the slowdown in the beginning of 6-3?
I don't know what you mean here, I couldn't find any slowdown at the start of 6-3, so a frame count and a more thorough explanation would be nice. :) It's true that I don't reach full speed before the first jump, but that's simply becuse I run out of ground to run on, and running back a bit at the start of the level wouldn't save any time here.
Phew... I think that's all. I've sent a PM to adelikat, asking him to replace the movie file with this 106 frames improved version. For all those who have watched this run, the only difference is that level 5-3 has been improved by 106 frames.
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
I recognize this, I've TASed several games where the in-game time depends on a timer that's constantly running in the background, even between the levels. For example, it could be a timer (let's call it A) that runs from 0-255 constantly, and the in-game timer increases/decreases whenever A is divisible by 16. In this case, you want the level to start when A is as far away from a multiple of 16.
I haven't looked at it in this game, but there usually isn't a way to manipulate this timer, usually you'll have to start the level slightly sooner/later in order to achieve the best possible in-game time.
If you're having trouble finding the necessary RAM address(es), I can try and find it for you, I've dug up these kind of addresses before. Just throw me a PM with the movie file if you want to keep the movie from the public for the time being. :)
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
adelikat wrote:
Handy! Can you mod it to work with Eggerland? :)
Done! See my previous post for the updated script. Put game=4 for eggerland. Note that resetting the steps counter by pressing start does not work in Lolo 3 or Eggerland, because pressing start has an effect in those two games. I'm not sure how to solve this, but at the moment, the only way to reset the steps counter in these games is to reload the script, or get to the next level.
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
It's been some time since anyone last posted a lua script here, so I thought I might share one with you now. It's a script for the Adventures of Lolo games and eggerland. Just change the variable "game" to the game you're using the script on (so if you're using it on Lolo 2, set game=2, eggerland is game=4)
game=2
poslist = {[0] = 0x06D, 0x091, 0x0CB, 0x083}
steps=0
while true do
input = memory.readbyte(0x008)
if game<=3 then levelover1=0x007 else levelover1=(0x017) end
levelover = memory.readbyte(levelover1)
xpos = memory.readbyte(poslist[game-1])
ypos = memory.readbyte(poslist[game-1]+2)
if game<=3 then bonus=5 else bonus=-3 end
direction = memory.readbyte(poslist[game-1]+bonus)
if game<=3 then dirval1=2 dirval2=3 else dirval1=3 dirval2=2 end
if direction==dirval1 or direction==1 then changeval=1 end
if direction==0 or direction==dirval2 then changeval=7 end
movement = math.mod(xpos+ypos,8)
if levelover>0 or (game<=2 and AND(input,16)==16) then steps=0 end
if movement==changeval then steps=steps+0.5 end
if math.mod(steps,1)==0 then c1=".0" else c1="" end
gui.text(10,10,"Steps: " .. steps .. c1)
FCEU.frameadvance()
end
You can reset the steps counter in Lolo 1 and 2 by pressing start. The steps couner is reset when you get to the next level or reload the script in all four games.
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
Nice improvement. This kept me entertained the whole way through. Here's a yes vote from me.
And I too know "Murphy's Law of TAS" far too well... *shrug*
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
I suggest you try to TAS a NES game, preferably a platformer. There are quite a few simple platformers that you can try TASing, to get a feeling of all the tools. Here are a few tips:
*Kid niki 2
*The smurfs
*Hudson's adventure Island
*Amagon
*Flintstones
I suggest that you just focus on getting to know all the tools and the method of TASing before you decide to try and make a run that you'll submit.
You need a FCEUX to TAS a NES game.
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
I know that adelikat and Lord Tom are currently planning on doing a warpless run on SMB 3, so you might want to get in touch with those two. Given the amount of work he and Mitjitsu put in to to the warped run, I'm guessing that Lord Tom also has good knowledge about the RNG.
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
Hello again, I have a forum bug to report:
Whenever you quote a forum post by TASVideoAgent or TASvideoGrue, the name used in the quote is NesVideoAgent.
Experienced Forum User, Published Author, Skilled player
(1890)
Joined: 4/20/2005
Posts: 2160
Location: Norrköping, Sweden
TASVideos Grue wrote:
om, nom, nom... want more!
Haha, seems TASVideos Grue wants more of this game. :) I didn't know he had the right to be this biased.
Well, it doesn't come as a big surprise that this was rejected, seeing as Ragowit's run was rejected for pretty much the same reasons. Personally, as stated in the author's comments, I found it fun to TAS this game, and I'm happy just having finished this long project, so I certainly don't feel like this was time wasted or anything. I'm also happy to see that a lot more people enjoyed watching this than I thought. :) At the same time, I do see the points from those who voted no - the game is indeed pretty slow and uneventful. Perhaps this is one of those games that are more fun to TAS than to watch.
Anyway, it didn't really concern me if this got rejected or published, I can rest happily that I've given this game another chance. On to the next project!
Perhaps this could even end up in Gruefood delight? (yes, I'm looking at you, alden)
