visual scene

Other Standards COLLADA
1

Hy.
I have a collada file with 200meshes , they ara positioned like the visual scene of the file.
Is possible to set for each mesh a the visual node transform matrix , that is multiplied by the it’s graph scene parents like the collada ref?
At the end of the visual scene parser i have a list of meshes with the precalculated transformation matrix for each mesh.
Now for render the scene i have to iterate the list and transform for the precalculated matrix for each mesh.
The question now is :How collada export the shene in a graph? and not a list with the multiplied transform???

THanks.

2

Do you have 200 different meshes? or 200 instances of the same meshes, each with different transforms?

You can make a conditioner that compresses your transforms, into one <matrix> transforms for each <instance_geometry>.

That would make your dae more optimized for loading and rendering.

Actually, can you elaborate more on your question?

Herbert

3

You can make a conditioner that compresses your transforms, into one <matrix> transforms for each <instance_geometry>.

How i do this with collada refinery?
What conditioner i must use?
Thanks.

4

Hy.
I have a problem with colada refinery , i use axisconverter to transform from z_Up axis to Y_UP the problem is that some meshes transform(in particular those with matrix transform)are wrong , i try to import it in 3dsmax 9 with feeling max importer and the result is the same of mine in directx.
I see that axisconverter change the matrix , but i can’t know if the result is correct.
may be a bug?
There is a new version of axisconverter?
can post me the link?
I try to use the update.exe file for update refinery ,but I catch an error(“type mismatch”) when it try to connect with an url(http://collada.org/refineryupdates/….).

Thansks.

5

Hi giugio:

A new version of axis converter will be uploaded to source forge soon. It fixed multiple bugs from previous converter and hope that it will help you. :smiley:

I will post the update when it is done.

Best regards,

Wei