Unfortunately, it’s already been tried. Even tied the axis to the same vector, which is used to build the laser, there are changes, but not the expected result, just now the discrepancy between the axes is smaller. I could not find the value manually, it seems that the dependence is not linear.