[Bf-committers] COLLADA Node Transform

Gaia Clary gaia.clary at machinimatrix.org
Thu Nov 3 10:09:25 CET 2016


Hi;
thanks for this information.
Would you mind to create a report for this?

Here is the link:

https://developer.blender.org/maniphest/task/edit/form/1/

You also can assign me to the report.
thanks
gaia

On 03.11.2016 08:59, Recep Aslantas wrote:
> I've another issue about node transform, since the subject is same I'm writing here
>
> I think Blender exports COLLADA matrix incorrectly in some cases
> maybe it is related OpenCOLLADA directly I don't know yet
>
> I created a scene with a sphere and cube, then I scaled cube (ScaleY) a bit
> then I added sphere as child of cube, everythings looks normal in Blender
> So cube is scaled and sphere is not and sphere model matrix must be ID Matrix in World (I think)
>
> Then I exported as COLLADA, It seems that  Blender exported sphere matrix incorrectly
> I was not sure and test in other tools the exported file
>
> The exported Node looks like:
>
> Cube Matrix:
>
> -0.8466222 -0.507823  0.4833452 0
> -0.4927458  1.530348 -0.5537914 0
> -0.2010783 -1.612002 -0.6780063 0
> 0                  0                0                1
>
>
> Sphere Matrix:
>
> -0.8466221  -0.4927458 -0.2010783 0
> -0.09768837 0.2943883 -0.3100958 0
> 0.4833452   -0.5537914 -0.6780063 0
> 0                   0                  0                1
>
> If the sphere must be Identiy (in World) then Sphere matrix must be inverse of Cube (in Local) in scene-graph
> am I right?
>
> After imported the COLLADA file, Blender renders same as I did and different from original,
> but Apple Quicklook still render correctly, weird! modo, FBX Review, tree.js renders like my result
>
> I'm sharing .blend, .dae and some screenshots via Dropbox:
>
> https://www.dropbox.com/sh/9nwr0t9j6e7rj1o/AACPnQ3m-OEq9amNemxZ6TcHa?dl=0
>
> Recep
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> https://lists.blender.org/mailman/listinfo/bf-committers



More information about the Bf-committers mailing list