The Smash Lab: What is it?

[Mmac]

Dang, too late for explanation of Bucket Cancelling

Oh well, I'll still explain:

When we look at similar stalls such as Marth's SideB, Fox's Shine, or Yoshi's Egg's, They're not really stalling, but rather just a momentum modifiers. All they do is just counter vertical Momentum with a slight boost (Or even a vertical stall), but they don't effect Horizontal Movement at all.

Now lets look at the Knockback properties on some of the Attacks, As you know, there's a property that the programmers they put in to counter Stall and Falls (Which is why moves like Toon Link's Dair and the such don't have the stalling properties when moved up). However the Momentum of attacks are still added to slow the vertical momentum which is why moves such as Toon Link's/Dedede's/G&W/Lucario/ect. Dair's still slow your momentum down.

I am not sure if this is a cause for some specials such as the Shine to have no effect on Counter Momentum because I think the properties return after an attack, but it might. Can someone test if after a high knockback attack on someone like Lucario, after he does a regular attack, does he get his regular Dair Properties back? Anyways what it looks like for now is that because there's no stalling properties in Fox's shine, the Vertical Momentum has nothing to stall his momentum boost from the shine, which is why he shines to his death.

Anyways, my theory is that because those 3 moves has no Momentum values on all the Axis's, the Knockback properties have no way to multiply with the Momentum values, therefore causing a complete stall no matter what situation. It's only my guess though. It also might be because they're "Free Control" moves (Where Momentum isn't factor in). It does explains Yoshi's Egg Roll/Spinning Kong, but unless G&W's Bucket does allow him to move slightly in the air, it is an oversight from the developers

If they were complete stalls though, then they wouldn't be effected the same by something similar like Diddy's UbB. I've even seen this on moves like Yoshi's DownB where his Stalling flip doesn't stall and he just moves graciously to the left until he lands.

 

[infomon]

Mmac, I'm pretty sure that momentum-changing aerials like Lucario's Dair, TLink's Dair, etc. have no momentum-changing properties when they're used during knockback, even after you've escaped hitstun with an aerial. Hmm that's prolly an irrelevant detail to your post, iunno lol I'll think about the rest more when I actually have time with my Wii to test stuff

 

[Mmac]

Hmm that's prolly an irrelevant detail to your post

True, but it does explain why the other "Stalls" don't work. It also proves that those 3 moves are not a standard stall...

I'm still going with because the momentum values on startup are Zero.

 

[ph00tbag]

There is no debug menu in Brawl. But there are some codes I've actually considered to ask such as being able to color-coat hitboxes so we can get a better idea of what they look like.

I've wanted something like this for a while. Also a code that slows the playing framerate, and doesn't just make the game slower.

 

[Pr0phetic]

What would be beyond AR/Oracina codes is to code a list of our own, one that can read values such as frames.

I think its hard though, its not a computer game, so ehh... A very interesting project to invest in.

 

[ph00tbag]

What would be beyond AR/Oracina codes is to code a list of our own, one that can read values such as frames.

I think its hard though, its not a computer game, so ehh... A very interesting project to invest in.

Reading frames could possibly be done with GeckoOS. Personally, I don't have the money to get the whole deal set up, but I think it would be neat to see what someone with that equipment could come up with.

 

[SCOTU]

All you really need anyways is a mode that slows gameplay down w/o frame interpolation, and some ingenious people.

Well, you don't need that slow mode hack, but it'd help.

 

[M.K]

Instead of color-coding, why don't we assign specific "elements" to a certain hitbox via hacks?
Weak hitboxes could use electricity, moderate hitboxes use flames, and strong hitboxes use Shadow or Aura....
Just an idea, I have no idea if this is actually possible or not.

 

[leafgreen386]

Instead of color-coding, why don't we assign specific "elements" to a certain hitbox via hacks?
Weak hitboxes could use electricity, moderate hitboxes use flames, and strong hitboxes use Shadow or Aura....
Just an idea, I have no idea if this is actually possible or not.

The point was to show the hitbox exists - its size and shape and how long it lasts - and not how "strong" or "weak" it is.

 

[Pr0phetic]

Yeah but the color-coding would work fine aswell for other things. I think it may be coded with some time.

 

[M.K]

The point was to show the hitbox exists - its size and shape and how long it lasts - and not how "strong" or "weak" it is.

Good point, but we can also use it to find the true "sweetspot" hitbox for moves like Tipper FSmash, Zelda Sex Kick, and the Falcon Knee of Justice.

 

[crewster]

That would be useful.

 

[Pr0phetic]

Anybody know the language Brawl is coded in?

Wow, look at what was started... genius.

 

[infomon]

Anybody know the language Brawl is coded in?

I dunno, I'm sure the Brawl+/etc. hackers could tell you, but it's not all that important since noone has the Brawl sourcecode. The hacks are done by just manipulating the raw binary. Language-dependent issues still matter for that, but not all that much, since I'm sure the game's not done with a virtual machine language like Java or Smalltalk.... it's prolly C/C++/probably some raw assembly for whatever the Wii's architecture is.

 

[Pr0phetic]

I dunno, I'm sure the Brawl+/etc. hackers could tell you, but it's not all that important since noone has the Brawl sourcecode. The hacks are done by just manipulating the raw binary. Language-dependent issues still matter for that, but not all that much, since I'm sure the game's not done with a virtual machine language like Java or Smalltalk.... it's prolly C/C++/probably some raw assembly for whatever the Wii's architecture is.

Yeah it's probably C++, if you can burn data from the disk to the computer it may make things more easier. it not hard to make a value reader in C++. Hitboxes, thats another thign ebcause it wire frames and stuff. Bleh.

 

[ph00tbag]

The point was to show the hitbox exists - its size and shape and how long it lasts - and not how "strong" or "weak" it is.

Precisely. Hitbox strength is known, and its rough area is known as well. What is not known, at least not empirically, is the actual space covered by hitboxes, hurtboxes, grabboxes, super armored hitboxes and invincible hitboxes. A color-coded system would make this a great deal easier.

Anybody know the language Brawl is coded in?

It wouldn't really be useful or necessary knowledge. No matter the programming language, that's solely used as a means for the engineer to have an easier time telling the computer what to do. Once the code is compiled, it's been translated to a much more raw language, which is probably proprietary to Nintendo. The original programming language would likely be discernable based on relics in the machine code, but really, using C++ to modify it wouldn't be all that productive, IMO, especially since the hex-hacking done by our lovely hackers is working well enough.

They could probably write a program to compile some kind of code into hex commands to make it more accessible, though. However, that feels like it would be more work than it's worth.

 

[Pr0phetic]

Precisely. Hitbox strength is known, and its rough area is known as well. What is not known, at least not empirically, is the actual space covered by hitboxes, hurtboxes, grabboxes, super armored hitboxes and invincible hitboxes. A color-coded system would make this a great deal easier.


It wouldn't really be useful or necessary knowledge. No matter the programming language, that's solely used as a means for the engineer to have an easier time telling the computer what to do. Once the code is compiled, it's been translated to a much more raw language, which is probably proprietary to Nintendo. The original programming language would likely be discernable based on relics in the machine code, but really, using C++ to modify it wouldn't be all that productive, IMO, especially since the hex-hacking done by our lovely hackers is working well enough.

They could probably write a program to compile some kind of code into hex commands to make it more accessible, though. However, that feels like it would be more work than it's worth.

Excellent point, and editing hex is normally done by ASM, I have to look into Brawl hacking wit Gecko and stuff, it seems there a lot that may be done.

 

[phi1ny3]

Instead of color-coding, why don't we assign specific "elements" to a certain hitbox via hacks?
Weak hitboxes could use electricity, moderate hitboxes use flames, and strong hitboxes use Shadow or Aura....
Just an idea, I have no idea if this is actually possible or not.

This is an interesting idea, but color coding is only really meant to show the hitbox, which we need, and this already does the job. Further still, this would affect "some" physics in the game (flame and slash have different kill properties, different damage properties, etc., if I can recall).

 

[SuperSmashTopaz]

Yeah it's probably C++, if you can burn data from the disk to the computer it may make things more easier. it not hard to make a value reader in C++. Hitboxes, thats another thign ebcause it wire frames and stuff. Bleh.

Totally agree.

 

[Pr0phetic]

Totally agree.

But did you see the errors in my post man??? Gosh i can't spell, i type to fast on my sidekick.

Lol on a serious note, we may be able to make wire frames out of the characters because its a coding feature, but a debug menu is very unlikley.

 

[SCOTU]

The language in which brawl was written in makes no difference. We can reasonably assume they used some object oriented language, however, and the raw data is likely to reflect that. To rip the hitbox data from the game would be easiest accomplished by somehow cracking the filesystem on the disk (which may or may not have already been done), and hoping that in there somewhere you can find something that is named for the hitboxes/ attacks. Otherwise, you're looking for random bytes of data that are probably 4Byte signed floats that represent at minimum two vertices and two radii (the minimum data required to represent a capsule -- the shape used for all hitboxen in Melee). Good Luck. If you can't find it in the filesystem of the disk, i highly reccomend not wasting time trying to find it some other way.

 

[Pr0phetic]

The language in which brawl was written in makes no difference. We can reasonably assume they used some object oriented language, however, and the raw data is likely to reflect that. To rip the hitbox data from the game would be easiest accomplished by somehow cracking the filesystem on the disk (which may or may not have already been done), and hoping that in there somewhere you can find something that is named for the hitboxes/ attacks. Otherwise, you're looking for random bytes of data that are probably 4Byte signed floats that represent at minimum two vertices and two radii (the minimum data required to represent a capsule -- the shape used for all hitboxen in Melee). Good Luck. If you can't find it in the filesystem of the disk, i highly reccomend not wasting time trying to find it some other way.

After what we did to Melee, nintendo was ry careful to not include that stuff in its final release, im 90% sure.

But then again, look how balanced this lovely game is. (No sarcasm on the lovely game part.)

 

[M.K]

I have absolutely zero knowledge in computer languages, so I'm sorry that I cannot contribute to that topic, but I do have an outlying point to bring up.
If I recall correctly, part of the problem with tapping into the Brawl hardware was the fact that it was a double-layer disc, in which some material was on each side, leading to a complicated mess of procedures to get to the data.
I'm sure if we didn't find the file for hitboxes, it would be on the unhackable side of the Brawl disc.
I guess it makes a tiny bit of sense to hide it away from the public.

Spoiler

...ok, I guess not. Where is SamuraiPanda, already?

 

[SamuraiPanda]

Here is whats going down:

Currently Ankoku has been added as a moderator for the Smash Lab, and he'll be helping me direct the room once it gets started. This week I have finals. So the official opening of the Smash Lab will be early in the week of the 14th. That means one more week to go.

In the meantime, I've recently heard that stale moves have less shieldstun and pushback. Is this true? And if so, how does it work? Is the reduction substantial? Is the reduction gradual as the move stales? Does it ever get back to the original shieldstun before staling? If you're going to analyze specific attacks, then I recommend using MK's Dsmash, Dtilt, Fair, and/or Dair, as those are some of his more popular shield pressure attacks.

Remember that the admissions list isn't finalized just yet, so you still have time to make an impression.

 

[Red Arremer]

I've tried it with MK's DTilt on Wolf.

So, here are my results:
The pushback is not different when the move is stale, it seems to depend on how stale the shield is. Means that if the shield is fresh and big, it shoves the shielder a bit further than if the shield is small and near breaking. The shieldstun seems to be dependant on this, as well. The difference in between is pretty minimal, though.

Of course, I could be completely wrong, but it does make sense to me.

 

[crewster]

Final a topic I can contribute to!

I don't have acces to a wii right now but I will later today.
Spade, that makes perfect sense.

 

[Steel]

The change between stale and non-stale moves when it comes to shield stun and HIT LAG is quite noticeable when you slow things down..


Meta Knight's Fresh Dsmash done on Marth: (keep in mind these dsmashes are non-tipper)
8 frames of hit lag
12 frames of shield stun


Meta Knight's Stale Dsmash done on Marth:
5 frames of hit lag
6 frames shield stun

When it comes to shield knock back, well these things are also slightly reduced as the move gets stale. Let's take a look:

Meta Knight's Fresh Dsmash:

Meta Knight's Stale Dsmash:

I also tested this with other moves including Marth's fsmash and got very similar results.


This also will change the frame disadvantages on block if your move is stale.

Let's see here....

ADVANTAGE FORMULA: SHIELDSTUN + HIT - SHIELDHITLAG - END (or IASA)

Fresh dsmash:

12 shield stun + Hits on frame 5 - 8 frames of hit lag - 34 (when the move ends)

= -25 Frame advantage (-18 OoS advantage)

Stale dsmash:

6 shield stun + Hits on frame 5 - 5 frames of hit lag - 34 (when the move ends)

= -28 Frame advantage (-21 OoS advantage)

Small, but still notable.

 

[Red Arremer]

Odd, because the DTilt doesn't change really much for the pushback when stale. Maybe it's so small that it isn't really noticeable.

 

[Steel]

Yeah, the dtilt is pretty weak on shield. Smashes are much more noticeable.

 

[Red Arremer]

That explains a lot. Did you try if the state of the shield matters on the pushback, as well? Because that's what I noticed - due to the DTilt being so weak for the pushback; but the smaller the shield of the shielder was, the further he was pushed, even if it only was just a little bit.

 

[Steel]

I just tested both dsmash and dtilt, there was no knock back difference when it comes to the state of the shield, at least from what I can tell.

 

[Red Arremer]

O_o

Then I'm dumb or didn't interpret the results correctly. Durr. Well, I did recognize the change of pushback, but obviously I didn't have the right conclusion on this.

Ah well, mistakes are human. So... I guess I can live with that.

 

[ph00tbag]

The language in which brawl was written in makes no difference. We can reasonably assume they used some object oriented language, however, and the raw data is likely to reflect that. To rip the hitbox data from the game would be easiest accomplished by somehow cracking the filesystem on the disk (which may or may not have already been done), and hoping that in there somewhere you can find something that is named for the hitboxes/ attacks. Otherwise, you're looking for random bytes of data that are probably 4Byte signed floats that represent at minimum two vertices and two radii (the minimum data required to represent a capsule -- the shape used for all hitboxen in Melee). Good Luck. If you can't find it in the filesystem of the disk, i highly reccomend not wasting time trying to find it some other way.

I imagine that hitboxen are in the filesystem (which has been cracked, as I recall). To have them just floating around on the disk with no home is Bad Programming 101.

 

[SCOTU]

they could be encoded in a larger file that represents the attack, or a group of attacks, or all the attacks, or in a file for each character. I wasn't suggesting they'd be scattered around, but rather that they may be hidden inside of some larger file structure. If there was just a directory structure like <character>\attacks\<attack>\hitboxen\<frame#>.bin or something that'd be nice, but they're probably grouped together in a larger packet of data, so it'd be really hard to find.

as for Bad Programming 101, I am a Computer Science Major in my Junior College year (focusing to become a game developer), and I do make Doujin games as a hobby, I know what game archetecture is like.

 

[shadowtroop]

The hitboxes may indeed be unseeable, but, why can't we MAKE then seeable? As in on the disk...
not in the game. If I had Ocarina I could try...

Scotu: Nintendo wouldn't do that on a disk publicly released, because then the info would be easy to find. But I think every character has a pack of data all clumped together to rep a character.

 

[Pr0phetic]

The hitboxes may indeed be unseeable, but, why can't we MAKE then seeable? As in on the disk...
not in the game. If I had Ocarina I could try...

Scotu: Nintendo wouldn't do that on a disk publicly released, because then the info would be easy to find. But I think every character has a pack of data all clumped together to rep a character.

Their most likely not included on the disk, it would be attached to wire frames or blob models. It is possible they left i nsomewhere deep like the Hot Coffee mod in GTA:SA, but i highly doubt it after such a turn from Melee.

 

[M.K]

Information.

This is correct, you beat me to it.
When a move is fresh, it throws the shielder backwards alot more.
When a move is decayed, the shielder is not thrown back as far.
It all depends on the amount of knockback that the move has, which is why moves like Meta-Knight/Marth's DTilt are good shield pokes.
EDIT: Just for testing purposes, do , moves stale when they hit items like crates, soccer balls, and capsules?

 

[∫unk]

Actually useful information.

Definition of ****.

Steel just gave you guys the guidelines on how to make a perfect smash lab post for the community.

So far I've only seen a few that have actually contributed. Everyone else is squabbling.

I'm just a messenger.

 

[Red Arremer]

Steel just gave you guys the guidelines on how to make a perfect smash lab post for the community.

So far I've only seen a few that have actually contributed. Everyone else is squabbling.

I'm just a messenger.

That was quite mean y'know... I just had the wrong conclusion on this one.

Besides I'm missing the equipment to upload pictures and/or videos... I intend to get me that, though.

 

[∫unk]

The fact that you think I was referring to you is funny.

btw you're squabbling now. As am I but I don't care to get in the Lab.