how to calculate the global inverse transform about gltf-tutorials

[javagl]

The global transform is explained in this section. The pseudocode for the computation of the global transform could be

Matrix computeGlobalTransformOf(Node node) {
    if (!hasParent(node)) return node.transform;
    else return computeGlobalTransformOf(node.parent) * node.transform;
}

(Note: The "parent" is not stored directly in the JSON. This pseudocode should just illustrate the idea about how it is computed in theory).

The inverse of that transform is ... just the inverse of this matrix :-) Depending on the programming language/environment that you are using, you'll almost certainly have some sort of Matrix class that already offers some invert function.

(Here is one implementation of such a function (in Java, but so low-level that it only needs arrays, and thus, could trivially be ported to any other language), but usually, you do not have to write this sort of code on your own - every 3D vector/matrix library already offers that).

[icEngineer-tech]

cool, thank you.
if my node doesn't contain any scale vector, rotation vector or translation vector. the default value for rotation are all 0? the default value for scale are all 1? and the default value for translation are all 0?
my last question is: how to convert this vector [ 0.259, 0.0, 0.0, 0.966 ] to the matrix rotation given in this tutorial https://github.com/KhronosGroup/glTF-Tutorials/blob/master/gltfTutorial/gltfTutorial_004_ScenesNodes.md#global-transforms-of-nodes

[javagl]

Yes,

• the default rotation is 0
• the default scale is 1
• the default translation is 0

meaning that the resulting matrix will be an identity matrix in this case.

how to convert this vector [ 0.259, 0.0, 0.0, 0.966 ] to the matrix rotation

The value for the rotation is (in some sense not a "vector", but) a quaternion that describes the rotation. Again, most libraries for 3D vector math should already have some function to convert a quaternion into a 4x4 rotation matrix. (In doubt, another very low-level implementation is shown here , but one should preferably use an existing function for this if it is available...)

[icEngineer-tech]

thanks for these precious info