Split image focusing aid on ground glass

[redu]

I have never seen such a thing so far on view cameras regardless of their size? Is there any particular reason for the absence of split image focusing aid on view camera GGs? Or... is there any such GG that i don't know?

[Gem Singer]

I've never seen one on the ground glass of a view camera.

I'm guessing that it would be difficult to make a split image range finder prism that would function properly with the wide variety of types and focal lengths of lenses that are used on LF cameras.

Also, I would think that it would be difficult to obtain accurate near-far focusing, using front or rear tilt, with a split image prism

Most LF photogs use a loupe for fine focusing.

[dazedgonebye]

Split image prisms perform poorly at smaller apertures. Even f5.6 can cause one of the halves of the prism to go dark.

[redu]

Split image prisms perform poorly at smaller apertures. Even f5.6 can cause one of the halves of the prism to go dark.

Very good point... After reading your interesting reply i've gone through a little bit more on this topic and came to know that the dual prisms used to split the unfocussed image only divert the light beams refracted from the edges of the lens to the viewer while the light beams arriving from the central parts of the lens gets refracted with such an angle that falls beyond the vision of the viewer. Hence with decreasing aperture the prisms black out. Here (http://doug.kerr.home.att.net/pumpkin/Split_Prism.pdf) is a very good paper for those who may be interested. And here (http://www.dyxum.com/columns/other/HaodaScreen/gubbe.asp) is a review of a split image focusing screen apparently designed for manual focusing on DSLRs. The guy says it's OK for f4 and may be for f5.6.

[aduncanson]

And consider how the problem gets aggravated if the lens can be shifted or raised, or the back tilted or swung, so that the lens is no longer directly in front of the center of the ground glass.

[redu]

And consider how the problem gets aggravated if the lens can be shifted or raised, or the back tilted or swung, so that the lens is no longer directly in front of the center of the ground glass.

Actually... it's not a matter of the amount of light how these split image prisms perform at their best. The light beams have to enter to these prisms with an angle in order to get refracted to the viewer and that's the reason why we need light rays arriving from the edges of the lens at it's widest aperture. A lens with much smaller aperture but shifted, would probably make the split image GG perform beautifully as far as what i understand from that paper.

[aduncanson]

Actually... it's not a matter of the amount of light how these split image prisms perform at their best. The light beams have to enter to these prisms with an angle in order to get refracted to the viewer and that's the reason why we need light rays arriving from the edges of the lens at it's widest aperture. A lens with much smaller aperture but shifted, would probably make the split image GG perform beautifully as far as what i understand from that paper.

Don't you think that moving the lens off center to the left or right would make one side perform better and the other side worse? And so would make the the split image more likely to black out on one side and be useless. Moving it vertically would probably hinder the performance of both sides.

[redu]

Don't you think that moving the lens off center to the left or right would make one side perform better and the other side worse? And so would make the the split image more likely to black out on one side and be useless. Moving it vertically would probably hinder the performance of both sides.

I guess, "but just guess", provided that the lens has big enough image circle; and image circle gets bigger with narrowing aperture anyway, shifting the lens would result the light rays passing from the center of the lens to arrive the split image prism in an oblique fashion which is what really needed for split image optics to perform well. However yes... you are right in saying that the horizontal or vertical movement of the lens would differ on the performance of the split image optics depending on how the prisms are arranged. Horizontal or vertical. However at this point i believe arranging the prisms in a diagonal fashion would result a workable split image focusing regardless of the direction of the lens shifts or rise/fall... But then again... it's just my limited optical knowledge talking here. Sorry if i am going too far and would be more than happy if somebody here could enlighten us.

[Vlad Soare]

Even if it worked, how would you use it? Place the subject in the center of the frame, focus, then recompose? OK, if your standards are perfectly parallel, and you use no tilt/swing at all, then I suppose you could do it if you were determined enough, though I don't see myself doing it with my bulky and heavy monorail. You'd also have to use a small aperture to compensate for the focus shift caused by recomposing, which in LF is more noticeable than in 35mm due to the limited depth of field (this is sometimes a problem with 35mm too, when you use long lenses).
If you start tilting or swinging, then focusing and recomposing will be out of the question. :)
I doubt that any focusing aid placed at a fixed location on the ground glass would be of any help. I personally wouldn't want one.

[aduncanson]

Redu,

You caused me to review my assumptions and I decided that I was wrong.

I had long ago convinced myself that one reason split image aids are unknown in large format is because the designer of the focusing aid cannot, in general, rely on the lens being in the center of the ground glass. (or anywhere else in particular.) It turns out that in the typical large format set up there is a way around this.

The images in a split image focusing aid are "aerial images". That is, they are not diffused by a ground glass. They must be aerial so that they can form ahead of or behind the focal plane and be shifted left or right by the prisms. You can only see an aerial image if you are looking directly into the "exit pupil" of the lens. (If the terms aerial image and exit pupil are not familiar then you might set up your camera focus an image and them remove the ground glass and look at the aerial image. Also observe how the exit pupil changes size as you stop down. You will see nothing unexpected in doing this, but it might help you to understand the meaning of these terms and the nature of an aerial image.)

As you recognize, in a split image aid, one side relies on light passing through the left edge of the lens and the opposite side relies on light passing through the right side of the lens. Stopping down the lens does increase the coverage, but reduces the exit pupil. This is why stopping down often causes one or both sides of the split image to black out. (It has nothing to do with coverage, but everything to do with exit pupil size.) Clearly, if the lens is not where the designer expects it to be, then the split image is more likely to have at least one side black out.

With a 35mm SLR you view the split image from a pretty much fixed eye point determined by the pentaprism and eyepiece optics. If the lens could be shifted with such a viewing arrangement, a split image aid would probably be a disaster for my reasons stated above. In a typical large format camera on the other hand the photographer usually will have the freedom to pick his eye point so that the eye, the split image aid and the lens exit pupil are collinear, restoring to a significant degree, the centered arrangement of an SLR with a non-shifting lens and a pentaprism. It is this ability to pick ones eye point in LF that I had overlooked. Using a reflex viewer on a large format camera would again lock in your eye point and cause black-out problems.

You might be interested in looking into parallax focusing, which also relies on the depth of an aerial image, and is occasionally discussed here.

Good Luck - Alan

[redu]

Sorry for late reply. I was on a business trip and just got back. Meanwhile I have found time to think over this issue on the plane. Yes shifting the lens to either directions would definitely blind one of the prisms hence render it useless. Unless one would need to use the camera without any shifts but exclusively by either tilting or swinging. In such a case the lens' aperture would only reduce on the axis perpendicular to the applied turn resulting an elliptic exit pupil. Hence a diagonally arranged prism system might still work if the lenses maximum aperture is sufficient. Well i now understand better why it can not be applied in the first place. Thanks to you aduncanson.