League of legends model problems....

[Uli]

Edit: Fixed it forgot to remove the null read which is in v4 but not v5

Maya 2015 64 bit plugin attached

Note: bunch-o debug code still attached.

[Uli]

Also with the introduction of Azir there is a new type of animation format which is "r3d2canm" where all the others have been "r3d2anmd"

Attached Azir's Run which has the new format.

I haven't got further than the header which is
8 Byte - "r3d2canm"
4 byte - version? - 1
4 byte - Length of File - 0xC
4 Byte - "canm"
4 Byte - Sub-version? - 3

Any help would be great! Unfortunately models, skeletons and animations are my weakest area.

[Wobble]

[out]

[Uli]

Also with the introduction of Azir there is a new type of animation format which is "r3d2canm" where all the others have been "r3d2anmd"

It looks like it contains the same information from the last .anm format. But, the data appears to be re-organized in new sections. It would help if you could find and post the same .anm file, but in an older format. And maybe the .skn file too.

Attached Azir Solider as it had animations which went from v5 to this new format.

[Wobble]

[out]

[Uli]

Awesome, thanks for that!

Time to try work out the position compression then, probably 11-11-10 format.

[Wobble]

[out]

[Wobble]

[out]

[Uli]

Anytime Riot do a file format it always comes out screwy, tried digging through the assembly but its a nightmare to read.

In other Riot formats 0x40 has been Vector3<Byte> and 0x80 has been Vector3<Float> but there has never been a section for key 0, was me being hopeful for a second but its probably an entire new key list for the file. I'm struggling to think/find anything at present

[Wobble]

[out]

[Uli]

Code: Select all

[ 4] w: (0000) (  0, 40), (7E04, E229, 0000)

Code: Select all

[ 1] pos: 0.252120, 91.229507, -20.523630

if those are meant to be the same position (Hope they are) then its literally taking the short and divide by 65535.

Going to test on a larger scale

decompressed gives:
0.252728, 91.229667, -20.523630.

Code: Select all

32260 / 65535
57897 / 65535
0 / 65535

0.49225604638742656595712214847028          = 0.252728163552
0.88345159075303273060196841382467           = 91.229667763103685053788052185854
0                                           = -20.523630

Edit:
Altered it to scale between -1 and 1 instead of whats posted exactly above, obvious what I changed to do that.

Edit 2:
Scale between 0 and 1:
http://i.imgur.com/x6jWA4p.png

Scale between -1 and 1:
http://i.imgur.com/QIsbvCj.png


The initial scale between 0 and 1 feels more accurate in terms of numbers instead of scaling it over -1 and 1.
Just need to work out how it all comes together now in terms of what bones are related to what frame.

I have a slight concern about how some seem to have multiple of the same values, unsure if its just for the first frame or not.

[Wobble]

[out]

[krolik]

Hey, Uli!

Could you please share the Maya plugin code including last changes (better with source code)?
or may anyone explain the order of the steps during reading this new animation format? v5 anm were more clear for me
to program it.

I have been trying to understand what you were discussing, but it seems like you are talking on unknown language =)

[Uli]

Hey, Uli!

Could you please share the Maya plugin code including last changes (better with source code)?
or may anyone explain the order of the steps during reading this new animation format? v5 anm were more clear for me
to program it.

I have been trying to understand what you were discussing, but it seems like you are talking on unknown language =)

The new animation format which we were just discussing needs a better layout mainly because I haven't implemented it yet so unsure on how indexing is done and it be nice to have a explanation just so there isn't any gaps but for v5 its perfectly documented.

[Wobble]

[out]