Collada animation bug

Hi All

i just wanted to put a spread of LFO’s into Mesh (EX.9.Geometry Collada). I tested with the Astroboy file and noticed that by inputting a spread of 4 LFO’s into the time input, 4 meshes per subset was created. meaning 16 subsets. but I then realized that each part of astroboy is animated the same. the strange thing is that the behavior is different if the number of time inputs is even or not.

I can think of a couple of behaviors the would be more logical than this.

see attached file

Edit: updated the file to show it more clearly. most clear when 4 values are input into the time input

AstroboyAnim.v4p (19.9 kB)

Update: I have noticed that it is the skinning transforms that are strange. new patch illustrating it once more.

Lower row with 3 astroboys behave correctly.

AstroboyAnim2.v4p (32.4 kB)

Anyone have anything to say about this?


the problem is how you distribute the spread of transformations you create.

Have a look at the attached file (is saved with 28.1, but shouldn’t do any difference, I’d guess)

Let me know, please, also if you find a better way.

AstroboyAnim2_mod.v4p (24.1 kB)

Hi h99

you have a problem with your approach when you change the spread count.

I will try to look into it a bit more

Please would you be a little more clear?
If you’re referring to the way in that patch the spreadcount feeding Stallone is generated, well, I’ve patched that way obviously because it’s the quickest hack possible in that situation.
Different situations, different approaches.
At least I can argue that it works and general logic is not wrong…

Bye :)

I am trying to wrap my head around it… I am puzzeled a bit… need some patch/think time… to recreate my understanding of it :)

Time pin is not valid for any dynamic connection…please check this

@androsyn, what do you mean by dynamic connection?

it’s prolly a bone matrices,
if you have 4 subsets it should have 4*60 Bone matrices,
and there is a limit for that somewhere.
as i tested with exporting collada biped’s meshes, it’s quite a chelange also, to have 60 bones per subseted mesh.
if you don’t have that it don’t work…
That’s might be a result if you spread the bone animation thru time pin. So might, it some lack of implementation for that stuff in there…

i must admit there is few more hardcore problems in that area