[Noesis] Plugin to export to a new format

[howfie]

it's no problem lol, you're not bothering me.
so, did you check out the data in gr_bone i asked you to check out? don't even worry about noesis yet, figure out the data and then worry about noesis.

[howfie]

lol? really? you filled it with zeroes and it ran hahahahahaha? ok, tell u what... lemme check out the data right now...

[Alsair]

lol? really? you filled it with zeroes and it ran hahahahahaha?

Yep, I was pretty amazed too.

Edit:
If you haven't already, you should take a look at the structure of the Skeleton I've analyzed so far in the main post.

edit2:

In that, the four bytes (before the matrices begins) is unknown, not sure what they are for...?

[howfie]

oh ok, i'll take a look at it... but first i'm going to analyze the first part, before all the animation data.

[Alsair]

oh ok, i'll take a look at it... but first i'm going to analyze the first part, before all the animation data.

I will start filling other regions with 00's to see what changes I would get. (Most of the time I would crash though.)

I'll post my findings so I can save you some time.

Edit:

Changed these two areas to 00's didn't crash (didn't screw up the model either)

The first one messed with the damage hitbox, I can hit further and do damage on monsters without even pressing the attack key.
The second one handles the the distance jumped, or the amount I slide after running, etc. I made this all into 00 which made me jump nowhere (It just showed me the jump animation)

So basically these two are unrelated to bones.


I wrote sfxData because it changed the SFX based on some bytes within that area. I will change around some more values to see what would happen.

etc. (Goes on until the count of sizeTwo)
Seems interesting, any ideas what it could be?

[howfie]

you wanna know something, i don't even think this game uses skeletal animation lol.
even the meshes are named Bone13 lol.
oh in gr_bone there always appears to be 0x0B55 matrices, which is part of the header.

lol, a quick test you can take one of those sections of 0x0B55 matrices, and replace the translation component with like double the values. the model should animate as though it is going to split apart, but should be the same animation. funny shit hahhahahaha. i think that first 4 bytes before the actual matrices is like an animation ID or something.

[Alsair]

you wanna know something, i don't even think this game uses skeletal animation lol.
even the meshes are named Bone13 lol.
oh in gr_bone there always appears to be 0x0B55 matrices, which is part of the header.

lol, a quick test you can take one of those sections of 0x0B55 matrices, and replace the translation component with like double the values. the model should animate as though it is going to split apart, but should be the same animation. funny shit hahhahahaha. i think that first 4 bytes before the actual matrices is like an animation ID or something.

Yeah I'm pretty sure it doesn't use skeletal animation either, since it goes through each frame and reads the matrices for that frame and updates the model based on that.

[howfie]

you want me to modify those translation values for you? i can do it pretty quickly using C++. i will take a whole set of those matrices and multiply the translation values by 10 hahahahaha. then u play the game and let me know what you see.

[Alsair]

you want me to modify those translation values for you? i can do it pretty quickly using C++. i will take a whole set of those matrices and multiply the translation values by 10 hahahahaha. then u play the game and let me know what you see.

Sure,

also I've modified bone04's animation data and replaced some of the matrices with 00's

[howfie]

yeah, then all this talk about weights and so on is kind of moot then. i'm no expert on noesis, but if you have the model format already figured out you should be able to dump some of this matrix data into a collada file or something and test out the animations.

[Alsair]

yeah, then all this talk about weights and so on is kind of moot then. i'm no expert on noesis, but if you have the model format already figured out you should be able to dump some of this matrix data into a collada file or something and test out the animations.

How would I map my current model import script to one keyframe's matrices for each bone in Noesis? Pretty sure that's what has to be done to make my model appear to look right.

Also, the four bytes before the actual matrices do nothing, I've changed them to 00's nothing happened.

[howfie]

yeah, it appears to be just an identifier of some sort.

you'll have to ask rich or look in his documentation... i'm just not sure. but you said you're a programmer, so try a format like collada hahahaha. have some balls; don't use Noesis or else you'll be like Rich, without any balls LOL (see some other thread to find out why hahahahahahahaha).

[Alsair]

So yeah Rich/MrAdults (That is you right?), I think the only data I can work with to make my models have a proper "stance" is to use this data:

The matrices from one of the animation key frame. However, based on your posts, you implied there was a skeleton with weights, parent-child indexes etc, that this lacks. So I was wondering if there's a solution that can map just those matrices with my model?

Each model also have a list of bonenames (as string) that references the one in the skeleton file (I've tested this by changing the name on either file, resulting in a crash).

So how would I associate the bonename found in the model file, with that very same bone's first keyframe matrices data found in the skeleton file?

This is pretty much all I need, I can deal with animations in the future.

[howfie]

for gr_bone.bon, at offset 0x2034E, you have 0x40 bytes of pointer data. it consists of a null terminated mesh name, Bone07, followed by a bunch of 0xCDCDCDCD values (invalid pointer).

the next 4 bytes is the identifier

the next 0x2D540 bytes is 0xB55 matrices, which are used to transform the Bone07 mesh in some way.

then at offset 0x4D8D2 is the next item in the list, the matrices to modify Bone01.

[Alsair]

for gr_bone.bon, at offset 0x2034E, you have 0x40 bytes of pointer data. it consists of a null terminated mesh name, Bone07, followed by a bunch of 0xCDCDCDCD values (invalid pointer).

the next 4 bytes is the identifier

the next 0x2D540 bytes is 0xB55 matrices, which are used to transform the Bone07 mesh in some way.

then at offset 0x4D8D2 is the next item in the list, the matrices to modify Bone01.

I know I am already parsing it in my script (check first page).


Unless this is for MrAdults?