garageCube

I am planning to sum the lumen of several projectors by stacking them. If I calibrate them over a flat surface, their intersection can be calibrated seamlessly. However, is there a way of overlapping signals in a 3D environment so MadMapper understands the different positions and sends a slightly different output to each projector?

So far, the only option in my mind is to create an exact 3D replica of the environment as a 3D surface for each projector independently. So 3 projectors would equal 3 3D surfaces, one for each, using the same media. Am I over complicating it?

The 3D object thing should work, that's tricky but interesting. If you manage it I'm curious to see pictures
Except this solution, you'd have to do many quads manually, I see no other option. Not sure which one would be the fastest, it depends on your experience creating 3D objects and the complexity of the environment...

This reminds me on my experiments with the 3d calibration. It would be nice to have direct access to the transformation matrix after calibration. Would also be nice to have parameters for the lensshift.

It would be nice to have direct access to the transformation matrix after calibration.

What would you do with it ?

Would also be nice to have parameters for the lensshift.

Can you be more precise ?

As far as I remember there should be a command line option to log the calibrated camera matrix.
You'd get 16 floats, end of the story.
Pretty raw.