Instead of measuring the distance from the wall to the node (wherever that may be), I measure the size of the projected image of the window. You will need a few more calibration points for this, but it has the great advantage that it works equally well for telephoto lenses and others where you don't know where the node is.
My "calibration set" is a casket set with 7 cells from 150mm to 750mm focal lengths. That gives me a sheet of cardboard with eight lines - left edge of the window, and one line for each 100mm step. Projecting the image of the window with the left edge lined up with the base line, it's easy to read off the focal length directly to within a +/- 10mm tolerance.
Without a casket set some calculation may be needed to make FL lines on the cardboard, but three or four known lenses should be enough.
Tortellini are too large for this, but spagetti-widths should be a useable unit of measure.
My "calibration set" is a casket set with 7 cells from 150mm to 750mm focal lengths. That gives me a sheet of cardboard with eight lines - left edge of the window, and one line for each 100mm step. Projecting the image of the window with the left edge lined up with the base line, it's easy to read off the focal length directly to within a +/- 10mm tolerance.
Without a casket set some calculation may be needed to make FL lines on the cardboard, but three or four known lenses should be enough.
Tortellini are too large for this, but spagetti-widths should be a useable unit of measure.