I agree with the others that
- The math isn't that important.
- The best way to include more of the foreground is to use falling front.
I would add one thing: unless your subject is planar (and it sounds as if yours isn't), you usually want the rotation axis slightly below ground (and if you check most of Merklinger's examples, that's where he puts it), so you'd use less tilt than you calculated. I hardly ever calculate tilt, but it's possible to use the formula as you have done to calculate the maximum tilt you'd normally use. This really just another way of saying what's been said here time and time again: a little tilt goes a long way.
The effect of pointing the camera down? Recall that J is measured in a direction parallel to the image plane, so if you tilt the camera forward by 5 deg, you theoretically would divide the vertical distance by the cosine of 5 deg. But the cosine of 5 deg is 0.996, and the difference in the calculated tilt is far less than I can accurately read, so unless you point the camera waaay down, it's usually not something to worry about.
I think "look at the groundglass" may be a bit of an oversimplification for setting tilt. I'd suggest reading QT's How to focus the view camera for arriving at the optimal tilt in a systematic way.
Bookmarks