I have been struggling for too long with matching chopped videos with differing non-standard aspect ratios when projected onto a set stage with two 3D Quad distort actors. I have come up with an empirical user actor (attached here) to output the somewhat required driving parameters: the aspect ratio modulation factor (aspect mod input of the 3D Quad distort), and the three spatial factors for which the stage will be covered from -100 to +100 (-xmin,+xmax, -ymin,+ymax, and -zmin,+zmax). For the z-depth, I was more concerned by the zero (initially set at -2.5 in the 3D Quad distort). The inputs of this Piece normalization actor are the matrix size of the chopped videos and the stage number (to extract its matrix size).
Fine, it works for most of the configurations I had faced so far. The attached screenshot of the main part of the actor for (128x256), (1024x768) videos and a (960x540) stage is a workinig case. However it is ad hoc and it sometimes fails. The matching is not good and the superposition of the two videos does not work.

I guess there must be equations to rule all this better than the one I came up with.

• First, I do not know why the z translation zero is for standard resolution -2.06 (not -2.5) and, for the others, an integer number of -2.06
• Second, the equations that work for aspect ratios greater than 1 differ from those for aspect ratios lower than 1
• Third, why the initial values for the 3D projectors differ from one to another (-150 for the 3D Rect Project...)

Anyway, would you - Mark, I guess you have - have a clue on how to handle the aspect ratios and ensure matching? I would really appreciate.

Thanks for your help,

Ikse

f9e091-piece-normalization.iua