1 2 3
13 14
Player (206)
Joined: 5/29/2004
Posts: 5712
Could it be related to what you said in the snapshot topic about VBA recording lag?
put yourself in my rocketpack if that poochie is one outrageous dude
Emulator Coder, Skilled player (1310)
Joined: 12/21/2004
Posts: 2687
Apparently... I think I was forgetting something from the change. Hopefully not too much has been done with the changed settings since 6b was released.
Player (206)
Joined: 5/29/2004
Posts: 5712
Haha Let's all wait for version 20 you guys
put yourself in my rocketpack if that poochie is one outrageous dude
Emulator Coder, Skilled player (1310)
Joined: 12/21/2004
Posts: 2687
Well, try version 6c now. It takes into account the other half of the table (oops) and also saves that stuff in the save states. It's always possible there's something else I forgot, but I'm able to at least get past the boss in MZM and past 10000 frames without desyncs now. EDIT: err, I mean, try version 7 now. (Maybe you're right, at this rate it'll hit version 20 pretty soon.)
Joined: 8/9/2004
Posts: 139
Location: Washington State
About your latest update... after you kill kraid to save some time I do believe you can bomb jump over the top of his exploding corpse and enter the next room while he is still exploding.
Former player
Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada
Ersatz wrote:
About your latest update... after you kill kraid to save some time I do believe you can bomb jump over the top of his exploding corpse and enter the next room while he is still exploding.
Heh, where did you hear that? Either way, even as it's not true I decided to bomb jump anyways. As I know Walker Boh would want it somewhere.
Former player
Joined: 3/9/2004
Posts: 484
Location: ­­
Here's a very useless fact: Running against Kraid's dying body results in a full charge for the speed booster, assuming you have it. Just like what can be done to Mother Brain.
Former player
Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada
Well another update, which is up to the first metroid near Mother Brain. It seems that shinesparking is just not working out for me. At first I tried saving a shinespark all the way up the shaft between Kraid and the Kraid/Ridley statue and using it in the room between the K/R statue and the Chozo Statue. Unfortunately it didn't end up saving any time, so I had to cut it. The second try would be shinesparking the metroid (didn't know the result as I had never tried it), but well you can see the result. This is where a Kraid first and a Ridley first run would meet up and join to be the same. Either way, at this moment i'm facing a hard desicion. I might stop this run right now and not complete it. I don't know if it's just worth it with all the new versions of VBA that are coming out (Not knocking it, keep up the good work nitsuja), I have no idea what is going to happen to the timing in the future, and am not even sure anymore if the laggy version is even acceptable. I suppose in a way I did this to myself, as I was the one who started speaking out about the lag. Now, I like the run and think that I have done a very good job on it so far. I also think there are a bunch of people who agree with me on this. Although, I honestly don't know if it will fit into the submission criteria now or later this weekend when I would probably get it finished. There are probably still atleast 10 hours of work to be done on the video, and I don't know if I want to throw that much more time away when I could be working on something else. Although, most of what is left is just running around, and it's not very much of it. This however DOES NOT MEAN I WOULD BE DROPPING IT ALL TOGETHER. I would just be dropping THIS VERSION, and start work on a Kraid first run. On the laggy version again, as even though the lag doesn't seem to effect the frame count, and it could still be compared to a nonlaggy version in terms of speed anyways. This would just be to see which route is actually faster to take. I would obvoiusly not release the Kraid version either. What exactly does this mean, assuming the laggy version isn't acceptable? Well you won't see any progress for a long time; as I will not be bothered to post updates, and I won't be pushing myself to get it done as fast as possible. After I have both versions complete up to mother brain I can get a picture on which is faster, and will sit on the results. The amount of time I sit on them all depends on how long it takes for the timing to get 100% finalised on VBA. This might be already done when the Kraid run gets finished, or it might be months after. Either way, I cannot see it being too soon. I guess that in the end, the main desicion relies with the judges and nitsuja. I asked Bisqwit last night, and I beleive his response was along the lines of 'if nitsuja thinks it's suitable then it's fine with me'. So I guess we will see the response from everyone else. Although I suppose if enough people annoyed me about it I would probably end up completing it for you jerks anyways...
Joined: 6/4/2004
Posts: 284
Well, it all depends; are you dead-set on submitting the first publish-worthy run, or would you be content to finish this timing-uncertain run, let someone else watch it and submit their run on the completed emulator, and then submit a faster run then they?
Joined: 8/9/2004
Posts: 139
Location: Washington State
my demand for updates on this video is too high if not boltr then someone else should finish it off
Emulator Coder, Skilled player (1310)
Joined: 12/21/2004
Posts: 2687
Oh... well first of all, I believe the non-laggy timing is now acceptably finalized to me, as it is no longer an arbitrary hack and I really do not want to spend any more time changing it. Secondly, I think the lag does affect the frame count considerably when it happens, although it does not happen all the time so I'm not sure how much it would add up to. I do not think a run done with laggy timing should not be accepted just because of that, but I would consider it to be like making an NES run with Famtasia when FCEU is now available, sort of. Honestly, your first set of updates came so quickly that I didn't think you would have much trouble with just re-tracing the steps you'd done so far. Either that, or finishing the run with the previous timing, and then going back to redo it with improvements and the added bonus of better timing. The former sounds like less work to me. But I didn't realize you were also unsure of which of 2 routes are better. Is it not possible to do some tests to get a rough idea of which one is faster? If not, I guess I'd prefer to see a run with the route you originally intended and leave open the possibility of it being improved on later, than to wait a long time for a lengthy comparison of two runs with no progress updates. Also, I know it's not much, but you get to keep the first 5000 frames or so of your run with either timing you use since there is no lag at all in the game until then.
Former player
Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada
The reason I say it doesn't effect the frames at all is because when i'm playing the laggy sections I do not know they are going to be laggy until I play it back. Samus moves the EXACT same amount of distance in those sections. The only exception to this is the room before bomb, where the sprite glitches a bit and the controls aren't very reseptive. During this time Samus still moves the same amount/speed during frame advance though, so it wasn't too terribly difficult to deal with the lag. I just delt with it like I would with lag in any other game, by throwing in an extra frame of delay so it accept my movement properly. The reason the first set of updates came so quickly is because it's a game I like to play, and have a large amount of experience with. For the most part I have spent ~6-8 hours a day on it so far. Right when I get back from class I start working on the run, up until I goto sleep (I'm like mini Teri, I suppose). On weekends and days off i've worked on it probably ~10+hours each. That is when i'm not with friends, although I didn't to much recently with them though. Alot of it has been testing paths through rooms and counting frames. My original plan was to do a Ridley run as best I could, and submit it. Then work on a Kraid run to offset against the Ridley run. If that was faster, submit that. If the Ridley run was still faster, I would then redo it with what i've learned from the Kraid run. I can continue to do this, but have the third version on the non-laggy version because that just makes sense. That would be more than fine by me. As for the tests: Not really. The whole idea is a slower route through Kraid, and slower Kraid kill for a earlier speed booster. How much time is wasted doing Kraid first compared to how much speed booster saves in Ridley? Hell if I know. As for the first 5000 frames, it's good to know. Although if I start a run on the non laggy version I will not be using my laggy versions beginnings. From what i've learned now I would redo it. Mostly for cosmetic purposes, but I think I could squeeze a small handful of frames out it aswell. Not worth restarting the whole laggy version for 4 frames off very beginning.
Emulator Coder, Skilled player (1310)
Joined: 12/21/2004
Posts: 2687
Hmm, well I know it is slower on playback, although not always as much as twice as slow, and I can see how when using frame advance (or slow motion) it is much harder to notice that sort of slowdown, even though it seems like it should be easier to notice it then. Once I recorded falling some distance in lag and moving after landing, and upon playing back without lag, Samus fell the same distance but got to the bottom faster, then stood there for a while, then started moving again causing it to temporarily re-synchronize. Anyway, if that's what you want to do, fine. The only thing I'm not sure of is whether a 2nd version you do is worth also doing with lag just for the sake of making it more comparable - I think it will be clear or at least possible to determine which is faster without needing to do that.
Former player
Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada
nitsuja wrote:
Anyway, if that's what you want to do, fine. The only thing I'm not sure of is whether a 2nd version you do is worth also doing with lag just for the sake of making it more comparable - I think it will be clear or at least possible to determine which is faster without needing to do that.
The reason I was going to do this is because I could use more of the first movie. I want the bit up until the paths diverge to be the exact same. That and I don't like doing that spike worm boss at the very beginning. I guess I could just take down both times when I enter the hive room and subtract taht difference from the final.
Player (206)
Joined: 5/29/2004
Posts: 5712
Could you do a "from savestate" movie so you wouldn't have to redo the beginning stuff to compare? That's how I like to do it.
put yourself in my rocketpack if that poochie is one outrageous dude
Former player
Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada
I guess that would be an option too.
Player (206)
Joined: 5/29/2004
Posts: 5712
Hooray! I'm saving you time. :) You just have to be able to do a little frame arithmetic
put yourself in my rocketpack if that poochie is one outrageous dude
Former player
Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada
Ok well it's updated again, as i've hit a major point where I want the people's input. Do I get seen when i'm running around as Samus or not? The VERY first spacepirate takes FOREVER to turn around so you can shoot in in the back, and it is ~125 frames slower to wait for him. I only end up in "alarm mode" for 1 room really, as I hide away in the second one. This should be the only time where getting spotted is faster than being stealthly. So: 100% Stealth, or get seen once at the very first Space Pirate to go ~125 frames faster?
Former player
Joined: 7/15/2004
Posts: 124
I think it would be better to set off the alarm. Going an entire 2 seconds slower will likely have people asking, "Why?" If there HAS to be a video that pulls off the 100% stealth stunt, it probably shouldn't be the any% item collection run. I have a question: Is it impossible to manipulate luck so that the pirate turns around sooner?
Emulator Coder, Skilled player (1310)
Joined: 12/21/2004
Posts: 2687
BoltR: Hmm, it seems a flag somehow got set in your movie that makes VBAs above a certain version think it isn't valid. I uploaded a quick fix for this so your movie now works in 8c. Backward compatibility remains unbroken even for recording, so please update to this version because it probably has less bugs than whichever version you're using. (Now watching, it's still looking great...) yet another EDIT: It is definitely possible to manipulate luck here. Just delay for a frame or so while crawling just before entering the room with the first space pirate - you can get it so it turns around at the exact moment you fall down to shoot him with no waiting time at all.
Former player
Joined: 3/13/2004
Posts: 1118
Location: Kansai, JAPAN
Speed > Stealth
Do Not Talk About Feitclub http://www.feitclub.com
Emulator Coder, Skilled player (1310)
Joined: 12/21/2004
Posts: 2687
feitclub wrote:
Speed > Stealth
Speed + Stealth > Speed (i.e. I say, do best speed AND stealth) (Well, at least test again to see which is faster, the difference is nowhere near as large as 125 frames.)
Joined: 6/20/2004
Posts: 292
Location: United Kingdom
I've seen a run of complete stealth, and it sure looks impressive but this is a run for fastest completion, so I'd say set that alarm off. Oh, and I'd never usually say something like that so you'd better set that alarm off!
Former player
Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada
Well i'm at Mecha Ridley in just under 34 minutes real time. Only about 3 minutes left of game. I don't think i'll finish it tonight though, as i'm pretty tired. I'll have the final thing uploaded probably tommorow afternoon sometime.
Joined: 6/20/2004
Posts: 292
Location: United Kingdom
Well from what I've seen, you're doing a great job. There was one thing I wasn't too sure about, I tried to rewind but I couldn't for some reason... It was one of the missile expansions you got... I dunno I can't really remember and I don't think I'll watch it until the final thing. Good Luck! It looks very impressive!
1 2 3
13 14