How do you focus your camera?
This particular sub-forum doesn't get the most traffic but I'm posting this here because although this has primarily to do with view camera adjustments/depth of field, the questions and principles apply to fixed geometry cameras as well, so I'd like to hear from shooters of all formats.
Note: I'm asking for feedback from people who want sharp negatives, and work with landscape/urban landscape, architecture...these kinds of subjects. This thread doesn't concern studio work, portraiture, table-top/still life and/or selective focus styles.
Some context. After a break from large format, I'm going back into 4x5 again, and I'm admitting here and now that as much as I read and practiced, I never got good at focusing the view camera. When it comes to film, darkroom work and printing, I can do whatever I want with few limitations. But when it comes to setting the camera up for the shot, for me that has always been the "drudgery" part of photography. This is probably the opposite situation of that of most photographers.
I have never been confident using tilts and swings. I understand what to do, but I'm never sure if it's right, and usually end up stopping down further than I should probably have to. Often don’t bother even trying tilts when I should. In the end I've usually settled for more diffraction rather than risk a totally screwed up picture. I'd also like to point out the uncertainty regarding depth of field made its way into my 35mm work.
There are people like Adams etc who seem to have done everything by eye and by feel. None of that ever worked for me. Then there’s all the Merklinger stuff. I sort of get it, but it is only straight forward when you have a diagram of the scene viewed from the side. When you’re actually in the field, how are you actually supposed to determine the distances and angles involved in making the decisions about tilt and aperture? It’s not like you can draw lines all over the place.
So here are some questions for people:
1.Situations where there is a lot of depth and multiple planes, and tilts and swings cannot be used:
·How do you decide where to focus?
·Do you use “conventional” depth of field rules?
·Do you focus on the near objects?
·Do you focus at a hyperfocal distance (ie focus on nothing)?
·Do you use Merklinger’s “object plane” method (usually resulting in an infinity bias)?
·Do you focus at infinity?
·How do you determine depth of field?
·Do you simply use the near-far focus method?
2.Situations where tilt can be used:
·How do you decide where to focus?
·Do you determine the tilt angle?
·If you use Merklinger’s methods and the hinge rule, how do you determine where the plane of sharp focus should be? How far below the lens should the hinge line be? How do you estimate all the distance measurements involved? How do you decide on the angle the plane of sharp focus should take from the foreground to the background? Etc etc (there are so many variables in this).
·How do you determine what lies within the depth of field? The math might work nicely on a diagram, but in the field how do you really figure out where things are in the space in front of you?
If I had to think about any of that I would never get around to taking a picture.
“The contemplation of things as they are, without error or confusion, without substitution or imposture, is in itself a nobler thing than a whole harvest of invention”
You have a plane to work with, you can put it "out there" wherever you want. 8x10 is easy because you can actually see what you are doing. Smaller formats require some experience and use of empiric formulas because you can't always see what is going on.
Personally, when using 8x10, I tend to avoid subjects where the important visual information does not all fit on a plane somewhere in the scene.
IC, how do you know where to put the plane? What about depth of field?
Cliveh, that's the point. Does anyone actually use any of this stuff? And if so, I'd really like to know how they do it with any kind of certainty in a reasonable amount of time when actually making a picture. I know I can't.
Hi Michael R 1974,
Of course LF photographers use the stuff all the time. But I get where cliveh is coming from. I selected a rangefinder 4x5 because I like the "syntax" of 35mm photography. If there are converging verticals, people have grown accustomed to it and accept it as a photograph.
I rarely used movements anyway when I dabbled in 4x5 before. So I don't have a lot of personal advice to offer here...
But my technique was simple when I had movements at my disposal... Square everthing up at the start (set all the settings at their detent). Then while composing anything vertical, like a tree that needs more, I'd keep the camera level and just raise the front. I'd keep it simple.
Something like that is not very hard to do at all.
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
I've noted your comments. If you do not employ tilt, shift, rise, fall and swing and whatnot in the larger format, then perhaps you should not be using that format. You said, "I understand what to do", but in the next breath, "but I'm never sure if it's right,". Well truth be told, there is no right or wrong. Nobody is creeping up behind you with a cat o' nine tails to whip you onto the straight and narrow of compliance to some obscure, unpublished rule that is right and that must be followed. You are participating in an experimental and quite variable branch of specialist photography. What there is to learn is done by actively using and employing the functions available to you in the larger format. There are very few published scientific texts to propel you along, and Merklinger's is one of the best, but you do not need to take all that on board to come away with a beautiful image. It does require refined skill and precision that you learn through experience, and much of the adjustment is by visually following the plane(s) as movement is introduced. That's the way it is. It is not relevant to dissect Merklinger's voluminous writing on the Scheimpflug Principles. Though informative, all that text has the capacity to dent the confidence of people who are not extrapolating writing into active practice, or feel intimidated. Depth of field, focus plane and subject alignment are all learnt behind the camera.
What do you mean by "Situations where there is a lot of depth and multiple planes, and tilts and swings cannot be used"? Why can't the movement(s) be used? If you are using an ultra-wide angle lens you will generally not have to introduce tilt to extend depth of field. Shift/rise/swing can be employed for perspective correction and elevating/reducing the scene.
When I photographed a waterfall in 2003, previous research of the subject alerted me that it was not facing the camera in an ideal plane, nor could I move the camera closer to it. Using tilt, I "turned around" the waterfall, in doing so, creating a 'peg' of near and far focus (extensible depth of field) while choosing to leave the periphery (of unsharp/out of focus) surroundings 'as is'. Depth of field and by association, focus, was completed visually.
The depth of your enquiry is troubling because it introduces a huge and potentially complex layer into the photographic practice that is not only distracting, but unwarranted. Will you ever get the picture done if you are considering so many principles, functions, calculations, permutations, measurements, conversions etc? To learn all about the how, where and wherefore, load the film holders up and go out and shoot, applying movements as you desire (and keeping notes of what you are doing as a valuable learning aid). Then come back and show us the beautiful pictures you created. Stuff the mathematics and diagrams.
.::Gary Rowan Higgins
One beautiful image is worth
a thousand hours of therapy.
"It is horrifying that we have to fight our own government
to save the environment."
I found this YouTube video helpful:
Scheimpflug/Plane of focus
The fix is in!
I like simple processes, Occam's razor if you will. I simply pull the lens stage out to rough focus, work out the Shimmyflugey thingy (aim the lens stage at the ground in line with the back) then do fine focus. Simple! I think most people that use view cameras overcomplicate the hell out of it. I am happy for you if you can twist your camera into a pretzel but you never need to in the real world. Maybe in brochures....
My 2 cents on when tilt is used:
I first estimate the tilt angle by estimating how far below the lens my desired plane of sharp focus will intersect the hinge, then I simply calculate it by Merklinger's equation for tilt angles less than 15 degrees---the largest angle I've ever used is about 8 or 10 degrees: (focal length in mm) / (5*J), where J is the distance in feet from the lens to the hinge. It's very simple. I always rack the film plane forward or backward when focusing, so, and this is key, once I understood that the plane of sharp focus rotates, at the hinge, away from the lens with forward movement of the film plane and toward the lens with backward movement, then placing the plane of sharp focus became easier for me to understand.
Try it by applying some degree of tilt to the lens, then just focus at a vertical subject like a telephone pole until just the base of the pole is in focus, the rest out of focus. Then, while keeping the eye on the base of the pole, slowly rack the film plane backward, causing the plane of sharp focus to rotate up, toward the lens------now, successive portions of the pole, from the base and up, that lie in the plane of sharp focus will appear sharp as that plane rotates upward.
It's a lot of words to describe it, but once grasped, is surprisingly easy.
I don't use LF and don't practice the Scheimpflug laws. I can answer what I do regarding 135 as this is also what you seem to be interested in.
Originally Posted by Michael R 1974
First of all, I'm very wary of DoF marks for serious work. The reasons are:
- The marks are generally set according to a circle of confusion a bit too tolerant for my taste: if using them, I tend to close 1 stop more, that is, close at f/11 while using DoF marks for f/8;
- Focus is some "binary" concept to me. It's either there, or not. Something in "acceptable" focus is typically just "tolerably" out of focus.
So what is the answer?
My personal answer, and YMMV, is that even in a situation where there is the need of an ample depth in focus plane, there always is one element of the picture that must be perfectly in focus. If that one element is perfectly in focus, the focus error in the other elements are more acceptable to the viewer and the scene looks natural.
To make an example: you have the landscape where you have the near field (let's say a stone 5 metre away), the middle field (let's say a tree 15 m away) and the background (let's say some mountains).
The normal approach in this kind of photography is: which is the right focus point and f/aperture which will give me an acceptable DoF so that everything, from the stone to the mountain, is acceptably in focus? By having recourse to some DoF tables or marks, one chooses the "optimal" focus point. This approach I don't follow.
This approach will lead to some focus point whereby none of the three elements is perfectly in focus. As I repeat, close is no cigar and acceptable is often a poor compromise.
My answer would be: focus on the tree. Do some reasoning about what is the f/aperture for the rock and mountains to be reasonably in focus. Remember they will never be "perfectly" in focus. But whatever decently close f/stop you use, the tree must be perfectly in focus.
The fact that there is an element perfectly in focus in the image makes the "acceptably in focus" elements, stone and mountains, much more acceptable to the eye (as acceptably in focus) than what would happen if no element were perfectly in focus (even if the rock or the mountains were a bit "more acceptably" in focus).
Stated in alternative terms: if nothing is perfectly in focus there is something wrong in the image. If one element is perfectly in focus, the mind sees as acceptably in focus also elements which are not perfectly in focus.
On the fallacy of DoF tables and marks I suggest reading this: