Help me to understand a focus technique problemgreenspun.com : LUSENET : Large format photography : One Thread
I’m looking for some advice on a focusing situation that I continue to wrestle with. I mostly photograph in the mountains where there is often an implied need for forward tilt to bring the close foreground and mountain background to the closest difference in focus (thinking of the tutorial on the LF homepage). The tenant and problem being that, as the lens is tilted the focal plane becomes parallel to a horizontal landscape. As you draw a horizontal line between closest and farthest objects, only the objects or parts of objects that touch the plane will be in focus. Everything above or below that plane will not be in perfect focus.
My dilemma being how to be more precise with getting these objects below the plane in focus while still using the close/far difference in focus to determine optimum f-stop. It appears that Fielder (very generally stated) brings close and far objects into focus using tilt and then compensates for the mid-ground using f-stop. Doing it this way requires me to go back to my millimeter ruler on my focus rail and focus again on the mid-ground objects in order to determine my optimum f- stop (by using the technique where f-stop is determined by the difference in focus between the two extremes of close and far). Others have suggested that you use a focus point for far focus as something, say, only a third of the way up the mountain when adjusting tilt. Again, this seems to require a refocus on the top of the mountain to determine optimum f-stop?
Assuming I’ve bumbled through this enough to make sense, am I approaching this problem correctly, and can you offer suggestions to help me get better at this? Thanks for the help.
-- Roger Rouch (email@example.com), May 28, 2001
(Might add that Ive been through Merklinger's web articles a couple of time as well.)
-- Roger Rouch (firstname.lastname@example.org), May 28, 2001.
One question for you Roger. How many degrees of tilt are you giving your lens?
-- Jeff White (email@example.com), May 28, 2001.
Hi Jeff, With far mountains and no real close foreground maybe 5 degrees or less. In this situation with no valley or depressions in the mid-ground the standard technique of "focus far tilt near", use the difference for optimum f-stop, and split the difference for focus works fine.
In more extremes of close tall peak and/or close foreground maybe 10 or 15 degrees (estimate). It is in these situations or where there are mid-ground valleys or depressions where I seem to get complications.
-- Roger Rouch (firstname.lastname@example.org), May 28, 2001.
Well sometimes in situations like this it is best to just forget using swings or tilts and just go to a deeper f/stop, or to rethink your image.
-- Ellis Vener Photography (email@example.com), May 28, 2001.
Well, in my experience with 4x5, most lens won't need more than a 5 degree tilt. 1 or 2 degrees is more the norm. I tend to take a guess at the tilt needed and then use the technique outlined on this large format site for focusing on the far and looking at the scale mounted to my camera and then focus on the near and checking the scale again. If the focus spread is too much then I rethink my tilt. I will try to guess where the error is, too much or too little tilt and then try again an see if there is an improvement. When in doubt I go with Ellis' advice and use less tilt and stop down more. I remember at a John Sexton workshop in the southwest, many people including John were photographing an unusual situation. It was an object very close and below us and a distant dramatic landscape. I hoped to engage John in a discussion about tilts to which he replied, "no tilts will help this, all you can do is stop down".
-- Jeff White (firstname.lastname@example.org), May 28, 2001.
Hi Roger, I'm not a tilt focusing expert. I think photographing in situations similar to your's, the solution for me was to stop down. Maybe if I understood tilt more, it would have helped me. But stopping down did the trick for sure. The optimum f-stop was the smallest I had avalible. Best, David
-- david clark (email@example.com), May 28, 2001.
The technique you're using is pretty much what I do with 8x10", and it works well for me. I'd rather tilt and stop down than just stop down, so the horizon will be sharper, recognizing that I may be sacrificing the base of the mountain a bit. The additional sharpness that draws the viewer's attention to the foreground with the tilt is part of the "large format look" for me.
-- David Goldfarb (firstname.lastname@example.org), May 28, 2001.
Roger - given that there are no moving objects like flowers or sorts in the foreground you surely can afford longer exposures. From my experience then, if you choose between tilting and stopping or stopping down only, tilting and stopping down does the better job. This way you make sure to get a tack-sharp fore- and background and stopping down will push the sharpness deep into the valleys. I'm sure you do it right the way you describe it.
-- Tom Castelberg (email@example.com), May 29, 2001.
Roger, it looks like everyone missed your point of confusion, in the last sentence of your third paragraph. I was having the same issue until someone walked me though it at a pace my short-legged brain could handle. You've obviously already figured out a lot more than I had.
If you pick a “far” focus point that is only part-way up the mountain, then tilt forward to focus on a “near” point, you’ve put the top of the mountain above the focal plane, in the “near” part of the wedge of focus. This is confusing, because it’s counter- intuitive, but you can confirm it by noting that it will take slight movement of the lens plane AWAY from the film plane to bring the mountaintop into focus.
So, using the "far" spot only part-way up the mountain is part of the solution, but you don’t usually want to use the top as the refocus point to calculate f-stop. Because the mountaintop is physically farther away, out in the wide part of the focus wedge, it will come into focus with too little lens movement, causing you to calculate an aperture that is not small enough. You’ll bring the mountaintop into focus, but still have a soft spot in the mid-ground.
The “Fielder method” described first in your post avoids that by forcing you to refocus on the point that will require more lens movement, and thus a smaller calculated aperture. The mountaintop will be in focus, but so will the soft spot that would have been missed by using it as the refocussing point.
-- Lyle Aldridge (firstname.lastname@example.org), May 29, 2001.
Thank you all for the answers. I need to get under the darkcloth and try a few new things. It does at least look like I'm on the right track.
-- Roger Rouch (email@example.com), May 29, 2001.
Roger, you wrote.... As you draw a horizontal line between closest and farthest objects, only the objects or parts of objects that touch the plane will be in focus. Everything above or below that plane will not be in perfect focus.
Keep in mind, this is true even when you do not tilt... there is only one plane of sharp focus regardless of tilt....how much of an area in front of or behind the plane of sharp focus is acceptably in focus is then determined by f stop. The only difference being that when not using tilt, all the planes are paralell, while using tilt the planes cuased by the f stop are a wedge, or cone shaped where the intersecting point is below the lens...
Now, to better address your question... you need to first decide if tilt is benefical to your scene... that means there must be a single plane in the foreground that extends outward awhile before it becomes 3 dimensional. For example, if you had a tall tree 15 ft away, titl is useless.... However, if you are shooting up a hill of flowers, tilt is ideal since there is one single plane near you for quite some time...
If you deem tilt is required, then your next task, as per Merklinger, is to determine how many ft under the lens the plane of sharp focus will be. This can be done by visualizing where you want the plane of sharp focus to be. It also can be done on a graph, but most people find this too confusing for the field... lets say it will intersect 10ft under the lens. Simple rule of thumb, fl/10ft x 5.5. So if you are using a 150 mm lens, you would need 3 deg tilt. That is an exclellent starting point and will should solve your near condition...then to align the far, just rack the back up and back until the plane of sharp focus is centered in the middle of the far subject. Then stop down as as required to acheive the remaining DOF by opening the wedge... to be safe, stop down to f32 if you have no polaroids or are unsure how to best determine this. As a general rule, if things look great in the gg, just stop down to the desired f stop, such as the sweet spot, say f16, if the far subject is very tall, and you can see the top and bottom blury, you have a lot of stopping down to do... setting the f stop does not have a simple rule of thumb like determining the tilt angle, unless you want to get out pencil and paper... Hope this helps...
-- Bill Glickman (firstname.lastname@example.org), May 30, 2001.
I'll stick my neck out here and oversimplify. Combine all available movements that will lessen the bed travel between your "nearest" item focused and your "farthest" item focused. When there's nothing else you can do with moves to lessen that travel, hold a fingernail on the rail at one "end" and travel the bed to the other "end" for a measurement. Best focus compromise is dead center of that travel and stop down. If you've still got more than 1/4" or so, you'll be up in the f45 world to equalize and you've got to do some soul searching about whether to take the trade-off in diffraction loss or go find a better layout. Sometimes the solution is to put the 50mm on the Mamiya where everything is sharp at f8!
-- Jim Galli (email@example.com), May 30, 2001.
Bill has the most precise answer. It's basically a summary of Merklinger's articles. The DOF is a cone/wedge-shaped region about the plane of sharp focus. It opens up to infinity at infinity, as has a size of +/-J at the hyperfocal distance (measured along a plane parallel to the film plane), where J is the distance from the lens to vertex of the wedge of DOF (below the lens). Once you understand the theory of DOF with tilts, the application in the field becomes elementary. Two things I might add. One is to inspect the stuff in the foreground w/ a loupe, around 4x (say, maybe your subject is a group of rocks) plus anything tall like mountain/tree- tops in the foreground/background. Nothing is more annoying than having the top of a mountain or tree go soft. Second, sometimes, you need to set the vertex of the wedge of DOF underground so that it intersects the ground in front of the camera at the foreground subject. This becomes important if the foreground subject matter is very close to the camera and you still want to use tilt to ensure infinity is sharp.
-- James Chow (firstname.lastname@example.org), June 01, 2001.