William Hallett wrote:

Sorry - I should have defined the variables in the equation more carefully. The object distance is the distance from the object that you're photographing to the nodal point of the lens, while the image distance is the distance from the nodal point to the image (i.e. groundglass). To generate a scale, you calculate the image distance at a number of different object distances. Start at infinity (for which the image distance is equal to the focal length, since 1/infinity = 0), then do (say) 50 ft, 25 ft, 15 ft, 10 ft, 6 ft (these are usually the numbers on Graflex's scales). For each of these, calculate the difference between the image distance and that at infinity - this is then the distance from infinity to lay off on the scale for that particular point. Note when you are using the equation that your units must be consistent - i.e. all measurements must be in the same units, whether mm or ft or inches. The only objection I can think of to this method is that the actual focal length of a given lens may be a little different from that marked on it. However, if you calculate the effects of an error of a couple of millimeters in focal length on the scale, you will find that it is almost negligible. The final test, of course, is to check your scale against groundglass focussing.

Your last paragraph suggests what I was talking about. Measuring the distance between the lens and the subject is logical, but instead of trying to measure the distance between the lens and the groundglass (which would be prone to errors because it varies only a tiny bit for each distance) why not just focus the lens on that subject? That should move the lens forward by the correct amount, AND it would correct for tiny differences from the marked focal length.

What you are describing is marking the scale experimentally, and certainly you can do that. The advantage of calculating the scale instead is that it is much faster, and that it avoids the errors that can occur in trying to mark off a location from the camera itself. If I were going to determine offsets for a scale experimentally, I think I would fasten a depth micrometer to the camera bed and use it to measure the front standard position as the camera was focussed to different distances.

When developing https://graflex.org/helpboard/viewtopic.php?t=4653 and when making said cams I tripod mount the camera and focus on an infinity target. Most instructions state to use a target at least 5000 feet away. I find the further the better. Celestial objects will be soft if 4000 feet or less is used for the infinity target.
Next I measured from the film plane to a focusing target consisting of several copies of http://www.normankoren.com/Tutorials/MTF5.html Koren Test targets printed at the printers highest resolution taped to a 20" x 30" sheet of 1/4 foam core 100 feet away, then 50 feet, 25 feet, 15 feet, 10 feet, 8, 6, 5, and 3 feet. I used a digital caliper https://www.ebay.com/itm/Digital-Electronic-Gauge-Stainless-Steel-Vernier-150mm-6inch-Caliper-Micrometer/111575281506?epid=685006111&hash=item19fa67af62:g:PeEAAOSwYGFUuMTK and measured from the edge of the bed to the rails at infinity then at each focused distance using a loupe on the ground glass for focusing. I used the center of the glass and moved the target keeping it center of the ground glass image.
The faster the lens the better and use the lens wide open.