Hey everyone,
Since this is kind of an unusual Question that doesn't fit in any category I'm not sure if this is the right place to post it, but I figured that the SDK Developer Community is the one that can come up with an answer for my problem most probably.
I'm a student of Computer Vision and currently working on a Project where I need some tracking data. Since our University has some licenses of Mocha for AE I want to try to import the data from the exported .shape4ae files.
I have currently a hard time figuring out how After Effects handles Transformations. In the .shape4ae File, there are basically two things to be found.
First of all the keyframed data of the shape of the object (e.g. the points for the xspline shape) and secondly and what is more interesting to me the information about how this shape is transformed for every frame.
There are 8 categories of data for every frame:
Translation X, Translation Y, Scale, Rotation, Shear X, Shear Y, Perspective X, Perspective Y
I wonder now how all this data for a single frame can be put in a 3x3 homogenous transformation matrix as can be found here
http://doc.trolltech.com/4.6/qtransform.html
It doesn't seem to be straight forward, especially the shearing. For the first frame in the sequence the values are set to perform no transform at all, so:
Translation X = 0
Translation Y = 0
Scale = 1
Rotation = 0
Shear X = 0
Shear Y = 1
Perspective X = 0
Perspective Y = 0
The fact that Shear Y is set to 1 makes no sense to me right now. The value should be set to 0 if this would be the value responsible for vertical shearing in a transformation matrix.
I hope that someone here can explain to me how these values are applied to create a transformation frome one point to another in after effects.
If it helps I can post a .shape4ae file in here.
Thanks a lot in advance!