First of all, the calibration wizard is a bit confusing because the voxel viewer user manual is required to understand what has to been done for cross talk calibration.
We placed a reflective object (5x5mm square) at a distance on a plane (cardboard box), it looked like a single pixel (leaking on adjacent pixels).
We did a calibration and then compared a cloud with cross talk and another one without cross talk. They both seems the same to me.
We want to minimise the cross talk effect because it tends to emboss surfaces adjacent to a surface at a different depth.
Is the threshold in the wizard impacting the result?
Does the plane need to cover the whole frame?
Why isn't the calibration code open source?
