Stage Hitboxes

[standardtoaster]

I found PTAD stage hitboxes. I had NTSC-J up at the time, so this code is for Japan only at this time.

Port Town Aero Drive DMG Mod (NTSC-J) [Standardtoaster]
206136EC 00000016
05504408 000003E7
055049A8 000003E7
05504F48 000003E7
05506028 000003E7
055065C8 000003E7
05506B68 000003E7
05507108 000003E7
055076A8 000003E7
05507C48 000003E7
055081E8 000003E7
05508788 000003E7
055092C8 000003E7
05509868 000003E7
05509E08 000003E7
0550A3A8 000003E7
0550A948 000003E7
0550AEE8 000003E7
0550B488 000003E7
0550BA28 000003E7
0550BFC8 000003E7
0550C568 000003E7
0550CB08 000003E7
0550D0A8 000003E7
0550D648 000003E7
0550DBE8 000003E7

All cars do 999% as a test. I will record a video of this later today. There are too many cars on PTAD to test one by one. Each line past the first line is a different hitbox. I can alter KB, angle, and pretty much anything else that a hitbox uses. Have fun! :D

 

[GHNeko]

inb4MuteCityRequests

 

[MonkUnit]

Wow this is incredible!

 

[GHNeko]

inb4BalancedPTAD

 

[GP&B]

Dude, legit. If you can do this on NTSC-U that would be sick. Seconding Mute City for sure, heh.

 

[hotdogturtle]

Hey, sweet.

 

[Eternal Yoshi]

Memory?!? NOT in the .rel file??

Which file??

 

[standardtoaster]

<standardtoaster> i can edit stage hitboxes now
<standardtoaster> i'm gdlk
<standardtoaster> people can't do it themselves
<standardtoaster> because the hitboxes actually are not in the rel
<standardtoaster> i think whatever tells the stage there is a hitbox actually points to some place in memory
<standardtoaster> and that's why using rels to port stages still keeps hitboxes

EDIT: Video of this. http://www.youtube.com/watch?v=nuJ_0g5VQ0s

 

[Thinkaman]

Memory address on my wii for first PTAD hitbox damage value was 81505968 on a random load. 81503D48 on the next.

Edit: Confirmed. These are stage hitboxes. they are routinely reloaded, but if you edit them very quickly live, this is it.

 

[Mr.Chuy]

I love you
This is MEXICELLENT! :D

 

[Thinkaman]

Changed some hitboxes on Norfair, now trying to track down the data in files.

 

[standardtoaster]

Thinkaman, did you have a texture or something on PTAD? In NTSC-J they aren't changed randomly like that, at least when I looked.

 

[Thinkaman]

Thinkaman, did you have a texture or something on PTAD? In NTSC-J they aren't changed randomly like that, at least when I looked.

Nope, I just changed characters from Bowser to Ness. I can double check later.

 

[standardtoaster]

Strange. Are you sure you didn't mistake the hitbox address you originally got with the hitbox address of one of the other cars?

 

[Phantom Wings]

standardtoaster, could you provide me with the all the default values (including angle, etc.) for those collisions? I quickly want to test something.

While I'll agree that the collisions aren't stored inside the module files, I believe that the module is still responsible for creating those collisions.

Looky here.



So far, everything I've looked at that's a hazard has these collisionAttackModuleBuilderConfig objects. Now I could just be grasping at straws here, but you never know until you try.

 

[standardtoaster]

Phantom Wings, absolutely! Sorry I don't have the values for the cars atm. I happen to have some of the default values for pirate ship, though.

Pirate Ship

Spike hitbox (bow of ship)

Damage: 20%
X offset: -50
Y offset: -25
Z offset: 0
Size: 10
Trajectory: 10E

 

[Amazing Ampharos]

Values are stored in modules. Thinkaman and I found some in Norfair. Go to FFD8 and look nearby. 38a0XXXX is damage, 38e0XXXX angle, 3940XXXX base, 3900XXXX growth. This changes the hitboxes on the lava floor. E310 uses the same mapping (but they are listed in a different order!) and affects the lava walls. Trying to figure these files out further is proving very hard for me.

Default values for PTAD cars are 14h damage, 5ah angle, and some known kbg and bkb but Norfair is the only case I have so far of editing modules succssfully and changing in-game damage behavior.

 

[standardtoaster]

Oh snap! That's an amazing find, AA!

 

[Thinkaman]

Onett car is by far the easiest collision to test, as it is the only collision on it's stage.

Damage: 30%
X Offset: 0.0
Y Offset: 0.0
Z Offset: 0.0
Size: 5.0
Angle: 35
KBG: 80
WKB: 150
BKB: 0
float 1.0
Hitlag Mult: 1.0
SDI Mult: 1.0
flags: 007FFDE0

next word: 00200380
not sure what this is... changes haven't produced a result?

next work is 003C000
60 frames is the refresh rate of the hitbox! confirmed!



These values have been confirmed. I have them up on my Gecko right now.

 

[standardtoaster]

:O EPIC! I didn't know there were floats before the flags...

 

[Amazing Ampharos]

I would like to add that the three floats are almost certainly trip rate, hitlag multiplier, and SDI multiplier since similar values are used on character hitboxes.

Interestingly, this implies that stage hitboxes can have all the properties normal offensive collisions can have except shield damage. Thinkaman was also telling me that some of the words before damage look like they might be id and bone, but I dunno.

He also found the refresh rate on the hitbox on Onett's car is 60 frames. We are getting close to fully understanding the values in RAM.

 

[standardtoaster]

I have ntsc-u up now so I can look at this a bit more in depth now.

 

[Phantom Wings]

I've confirmed the hitbox function you mentioned on Norfair AA. It's the 119th function contained in the Magma object.



From the looks of things, those values get sent into the stack and then a function is called to create a hitbox accordingly.

Now... to find out exactly what that function is.

 

[Thinkaman]

For those playing the home game, these parameters seem to be at 41FC in the onett module file?

38A0001E 38E0002D D0010078 39000050 39200096 are the first 5 words--check the last bytes.

 

[Dantarion]

Phantom wings give me dat module viewer lol

 

[standardtoaster]

Dantarion, he already posted it. Check the main page of the workshop.

 

[Phantom Wings]

I've confirmed your car function Tinkerman inside the 117th method of the grDXOnettAttack object.

Hold up Dant, I need to update the link - I reuploaded the file for a small bugfix.

edit: done here's a link to it

Hmm... it seems that all functions are called with some sort of utility belt pointer stored in r3. If I could trace that... .

 

[Supreme Dirt]

PW just can't stay away from Brawl, can he? I might take a look at some of these when I get home. I'm particularly interested in finding the information on Brinstar, so that's what I'll work on.

 

[[TSON]]

this is erotic

EDIT: Wait, Onett's car has a float for its damage in the STDT though... Why would they do that if the damage is already set in the ASM?

 

[standardtoaster]

this is erotic

EDIT: Wait, Onett's car has a float for its damage in the STDT though... Why would they do that if the damage is already set in the ASM?

Wait, what? I thought you told me you didn't find that in the STDT when you were looking through it a while ago.

 

[Eternal Yoshi]

I THINK I found the hitbox of one of the cars at offset 7F18 in st_fzero.rel.


38A000014, DMG
390000064, KBG
394000046, BKB
392000000, Weight Knockback??

I only found 38A00014 once in the .rel.

Could this mean that all the cars use a form of go to command to this hitbox???

I cannot test that theory now.......

How much damage does the road do again? 10? I can't remember...

 

[Amazing Ampharos]

This old research of mine may prove helpful to you:

Mario Circuit cars (15%)
Rumble Falls spike 1 (15%)
Rumble Falls spike 2 (10%)
Rumble Falls spike 3 (15%)
Bridge of Eldin King Bulbin (15%)
Bridge of Eldin Bomb (30%)
Pirate Ship catapult (0%)
Pirate Ship bombs (55%) [pretty sure this has a second hitbox]
Pirate Ship boat bow in general (20%)
Pirate Ship boat bow beaching on rock (10%)
Norfair lava floor (14%)
Norfair lava walls (14%)
Norfair lava background (18%)
Norfair lava pillars (10%)
Halberd laser (42%, multihit)
Halberd bombs (20%)
Halberd claw (10%)
Spear Pillar laser (27%, multihit)
Spear Pillar Dialga attack (25%)
Spear Pillar Cresselia background beams (15%)
Spear Pillar Cresselia spinning beams (20%)
Port Town Aero Dive cars (20%)
Port Town Aero Dive track (15%)
Port Town Aero Dive wall (10%)
WarioWare speeding car (13%)
WarioWare acid rain (1% continuous, at most about 15%)
WarioWare raining arrows (30%)
WarioWare stomping foot (20%)
WarioWare background earthquake (20%)
Distant Planet monster (no damage, grab -> death)
New Pork City ultimate chimera (100% or 200%, varies: I seem to recall seeing 300% before even) [now I think this is always 100% but sometimes hits twice]
Summit icicles (20%)
Summit front of glacier (18%)
Summit fish tail (15%)
Summit fish bite (no damage, grab -> death)
75m fires (10%)
75m Donkey Kong (20%)
75m springs (20%)
Mario Bros. red crabs (13%)
Mario Bros. blue crabs (15%)
Mario Bros. purple crabs (17%)
Mario Bros. green turtles (11%)
Mario Bros. purple turtles (13%)
Mario Bros. red turtles (15%)
Mario Bros. fireballs (20%)
Flat Zone 2 chef form food (8%)
Flat Zone 2 Oil Panic form people (13%)
Flat zone 2 Lion form tamers (12%)
PictoChat arrows (10%)
PictoChat side spikes (20%)
PictoChat fire (1% per hit)
PictoChat bombs (25%)
PictoChat cart (10%)
PictoChat pirahna plant (10%)
Green Hill Zone checkpoint (10%)
Jungle Japes klaptrap (30%)
Onett cars (30%)
Corneria arwings (12%)
Corneria main cannon (30%? hard to tell)
Corneria main cannon recoil (30%)
Green Greens Bomb Blocks (20%)
Green Greens exploding apples (20%)
Brinstar acid (14%)

It would not surprise me if every car had the same data in the module. I know the lava walls and floors on Norfair are a lot of different hitboxes in RAM but only have one entry each in the module (and for that matter, both the left wall and the right wall reference the same data).

EDIT: 38A0XXXX may not always be used. I pretty well failed to find the hitbox data for Norfair's other hazards unless it actually is in a very different region of the module than the two I found for sure (and even the ones I found for sure didn't contain all hitbox data that I was able to find, only damage, base knockback, knockback growth, and angle). I think letting PW do a litle of his magic with the module editor has a high probability of significantly easing this process, on a related note. If we understood a little more about how this works and how modules are structured, finding stuff would doubtless be far easier.

 

[pikazz]

this is awesome! :D
but there is one thing I am curious about and it is about to swap the hitbox to somewhere else like from the car in onett to the upper tree in onett so each time you are near/on top of the tree, you get attacked by the tree.

 

[[TSON]]

just btw but mario bros' turtle & such are coded as items in PSA language

 

[standardtoaster]

To add on to what TSON said, this is only because you can pick them up and are therefore treated as 'items'.

 

[Eternal Yoshi]

Thanks for the info.
Just a crazy thought, but would it be possible to change the bone that the hitbox is attached to??
Like say, to different models like a collision in MiscData[2]?

Edit:I forgot to tell you guys, after changing 1 car hitbox, all the cars followed suit.
It's like they all are using a go to command.

 

[Amazing Ampharos]

They count as items even just roaming around? I figured hitting them resulted in them "creating" items but that the roaming hitboxes were treated as stage hazards. I suppose it makes sense though; they must be programmed like Bob-Ombs as independent item actors that are capable of having hitboxes. Oh well though; it's not important.

I know Thinkaman found what were maybe ID and bone in the RAM before all the other data (the previous two words). As far as he could tell, ID was always 0 on stage hazards but bone could be any number of values (I don't remember which hazard he found as having a bone value of 56 or even if that's 56d or 56h). Changing that value would doubtless have an effect if the guess as per function is correct.

 

[Eternal Yoshi]

A pity that they don't follow that consistent pattern.

All I got when I tried to find a pattern is this..
.Damage
??Flag??
Knockback Growth
??Flag??
Angle
??Flag??
Weight Knockback??
Base Knockback

Sometimes the flags in between the damage and knockback growth aren't there....

Anyway, using the Sakurai angle to look for hitboxes seems to be easier than looking for damage.
I found a hitbox in Pirate Ship's .rel @ E800.

It does 35% damage @ default. Maybe one of the bomb hitboxes???

Edit: Yes, it's the main hitbox of the bomb.

Edit2: Is 13200 where the bottom of ship's hitbox located??
5FE0 seems to have identical entries in KBG,KBB, and damage.

Could B8C0 be the catapult????

Edit3: Yes that's the catapult

 

[standardtoaster]

Great finds, EY! I can look at this later and get the full params for them.

 

[Eternal Yoshi]

OK. To clarify about the bombs, they have 2 hitboxes. The primary one does 35% damage.
The secondary one does 20%.

I wonder which one is the size anyway.
If we could find the XYZ offsets and the size, we could resize the ship bottom's hitbox to cover the rudder as well.