Nikon backfocus adjustment and infinity
 

I'm curious how everyone else is adjusting their backfocus distance between lens and ground glass. It seems that my method of measuring 46.5mm from mount to matte side of groundglass doesn't always result in good infinity focus. So I end up manually tweaking it forward or back until it appears sharp to the eye. This is with the lens set to infinity and aperture set to lowest setting (1.4, 1.8, 2.8). When infinity is "right" the distance markers on the lens vs the actual distance has a wide variance.

Am I correct that 46.5mm is measured from the register -- which I understand is the front edge of the F-mount? Most other backfocus adjustment I've read do it from the rear lens element then use a chart like the seiman star, putora 7a9, or a magazine at 45-degree angle. Maybe this is a property of the 50mm and 28mm prime lenses? Please so a sanity check on me.

I have been mounting a good 50mm f1.8 Nikon of known focal accuracy as shown by the focus numbers on the barrel, setting up a target one metre from the focal plane, lighting it for the lens to be at f1.8 with the camcorder at 0db gain.

This makes the depth-of-field as shallow as it can be made to be, therefore assuring as much accuracy as possible. I then set the groundglass so the image is sharpest when the lens is set to 1 metre.

I use the camcorder autofocus periodically as moving the GG also affects relay focus, only slightly, but enough to trick you.

I've been using a Lemac chart which has a Siemens star in each corner. This enables me to set the focal plane at exact right-angle to the optical centre axis.

I also run the disk during adjustments so that any shift due to motor end float or disk runout is accounted for.

I then test with a larger Siemens star at longer measured distances and then with a near infinity target like the sea horizon or distant landmarks.

46.5mm is the Nikon flange to focal plane.

There's some funny things happen when focusing through a disk onto a rear groundglass surface,

Maybe a slight refraction affects focal distance, more likely build errors or disk runout due to my own shoddy workmanship.

Whatever, I have never found the lens focus to be spot on accurate to the 46.5mm measurment to the rear of the GG.

Just where in the depth of a 5 micron grind, the image is actually formed I don't know either, but it may be slightly forward from the rear surface enough to affect very fine adjustment.

Thanks Bob, great post. I've been scratching my head over that measurement myself.

Lenses like the Peleng 8mm f3.5 have provision for a colour correction or ND filter in back of the lens in the form of removable flat glass disks or plane panels (not sure if that is correct term - somebody give me a metaphoric smack over the ear if if have misled).

If no filter at all is chosen, there still has to be a clear disk in there as the backfocus of this extreme wide angle lens is otherwise affected.

In the Nikon F Mount to CP16 mount adaptor, there is also a plane panel of glass. This is apparently also there for optical reasons, not just to keep dust out.

Quyen Le's Letus35 also has a plane panel in there between the SLR lens and the groundglass, maybe for dust exclusion as I have believed but maybe for optical reasons as well.

These are what sponsor my thought that the thickness of the groundglass itself would affect backfocus if the groundglass surface is to the rear.

That would be 100% correct and confirmed, Bob. I run a lot of those Russian lenses myself, and most of them have the replaceable rear filter element. Quite inexpensive to use, since you can swap out a clear for an 85 and not have to invest in (sometimes very expensive) front end correction glass, but of course, since it's so near to the focal plane, you have to keep everything spotless.

Quite definitely if you have the GG facing away from the stills camera lens the backfocus plane will necessarily be affected in some way.

I do have another question for Bob that I'll start another thread for, though...