PDA

View Full Version : Unusual Yet General Technique Question



Todd Caudle
24-Dec-1999, 01:49
I'm hoping some of you accomplished LFers can sort out a disagreement I have wit h a fellow photographer that goes to the basic foundation of large-format photog raphy. It's kind of hard to explain, so forgive me if this sounds a bit convolut ed and long-winded. Here goes. . .

My contention is that LF depth of field technique is basically the same as that which I have used with my Mamiya RB67 (pick a composition, select an aperture th at'll render everything I want in focus), except that with my LF I can use movem ents to "enhance" or "alter" my DOF on images that need more depth than simple a perture selection can attain. I know how the Scheimpflug principle works. My fri end, on the other hand, says that I should always use movements first, not only in terms of order of application in the field, but from the very start of the co nceptual process. Here's an example that I used on him to stress my point, and h is reply: I'm shooting two cottonwood trees, one is 75 feet away, the other is 1 00 feet away. I know that with my 210mm lens, if I focus on the closest tree and use a relatively small aperture, say f32, that everything from that tree to the horizon will be within the zone of focus. (Assuming, for argument's sake, that the tree at 75 feet isn't quite at infinity, but close.) Or I could even focus a little beyond the first tree and let some 'front-side' DOF bring it into focus. Therefore, in either case, I wouldn't even need to consider movements.

My friend contends that instead of focusing on the closer tree and stopping down , I should focus on the closer tree and then use tilt to bring the farther tree into focus. Then I can stop the lens down to increase my chances further. My pro blem with that scenario is that the tilt will not increase the DOF unto itself, rather it will only change the *plane* of focus. I would therefore have to worry not about being in focus from one tree to another, but instead, from the bottom of the tree to some point *above* my new plane of focus.

Furthermore, he argues that he can get much more DOF out of a 210mm lens on his 35mm camera than a 210mm on LF. MY argument is that both lenses give exactly the same DOF (DOF is DOF is DOF, etc.), but the LF lens covers a wider angle of vie w, so objects that aren't even in the frame of his 35mm shot will need to be rec koned with in LF. I am certain that if we shot in the exact same direction with the same focal length (him on 35mm, me on LF), I could crop a 24mm x 36mm frame out of my 4x5" shot, lay it side by side with his 35mm image, and the two would be exactly the same (discounting slight variations in lens quality and other su ch peripheral factors).

Any thoughts?

Chris Partti
24-Dec-1999, 03:49
You are definitely right on the second point. In fact, since 35mm needs a smaller circle of confusion (one of the factors in the DOF equation) than 4x5, DOF for the same length lens at the same aperture can actually be said to be greater for the larger format. As you say, if you looked at a cropped 4x5 neg, it would look the same as the 35mm. On the first point, given the specific example you use, like you I would probably stop down and forget about tilts for the reasons you give. Tilting probably doesn't gain you anything in that situation (swings might if the trees were the only thing you wanted in focus, but that's an unlikely scenario). But I don't agree that you should choose either movements or stopping down as the first option in all situations, it depends on the nature of the subject. If I were taking a picture of the ground at the Bonneville Salt Flats, I'd definitely think of tilting first.

Robert A. Zeichner
24-Dec-1999, 10:56
There's no doubt about the fact that tilting in the tree scenario would result in out of focus tree tops (or bottoms, depending on which way you tilt). Many times, in LF photography, the temptation to use more movement than is truly needed results in creating more problems than it solves. After a bit of practice there is a certain amount of intuition involved in sizing up a particular situation, but as a rule, I like to measure the depth of field I need for a scene by first focusing on the nearest and then the furthest object I want in focus. By observing the spread, in millimeters, on my focus bed, I can calculate what aperture will render the depth of field I need to bring everything into acceptable focus. I then place the front standard precisely in between the two extremes and shoot. Sometimes the spread is very little, maybe one or one and a half millimeters. When this is the case, I don't bother with movements unless improvement is blatantly obvious, such as the example one other contributor gave of shooting the salt flats. Here, there is no doubt that tilting back will reduce the spread at little or no cost in sharpness to other scene elements. That will give me more choice of aperture to use. If it's a more complicated scene with foreground objects that rise high into the frame, there's no point going through the exercise if the spread is manageable. I just calculate the needed aperture (actually I have a cheat sheet close by) and shoot. If the spread is too great, say five or six millimeters, then I have to play a bit. Sometimes I can improve the situation, other times I make it worse. It all depends on the scene. Occasionally, I have to do a compound Scheimpflug, tilt and swing. It's a little more time consuming, but sometimes worth the trouble. There is an excellent article in a past PhotoTechniques magazine on the technique I described. Forgive me but I forget the issue. It was a couple of years ago, I believe. Perhaps another adherent to this procedure has the issue number and will post it. Good luck

Carlos Co
26-Dec-1999, 17:33
Most likely, you won't be able to gain DOF by tilting unless the trees are very short.

Chris Wray
26-Dec-1999, 19:28
What many people don't realize about swings and tilts is that the world is rarely that two dimensional. I find I use a tilt or swing about 10% of the time, even with many close images in my portfolio. I commonly use f32 or f45, however. There always seems to be something sticking out the top or bottom, or sides of a tilt or swing image. Sometimes, you can use about half the movement, and a small stop to boot. This gives a nice combination for depth, without sending the oddball parts out of focus. I find I rarely use a movement for focus control unless the obect is within a few feet of the lens.

James Chow
27-Dec-1999, 00:02
On the first point, I agree w/ Robert. You'll end up w/ out of focus tree tops. If you use tilt, I think you have to put the J point either deep underground or high in the air using a small amount of tilt. You can play tricks w/ the standards to move the J point behind the camera and in the air or below ground, but it certainly complicates the shot.

On point two, theoretically, both 35mm and 4x5 lenses w/ the same focal length and CoC should have the same DOF, BUT in a Dec 20,1999 post (http://photo.net/bboard/q-and-a-fetch-msg.tcl?msg_id=000hMf) on the medium format digest by Kornelius Fleischer of Carl Zeiss says, "In general the Biogon is less prone to resolution drop with defocussing compared to typical SLR lenses of similar focal length. This means: it has more depth of field than other lenses of different optical design. (It is not widely known amongst photographers, that lenses with the same focal length can have different depths of field. Cinematographers have been aware of this fact for more than half a century. The realties of optics are far more complex than the simple equations used for calculating depth of field scales and tables. Those are just funny toys or rough "guesstimates" with little accuracy and meaning in the real world.)." So I guess the conclusion on the second point is 'maybe.' :-)

Alan Gibson
28-Dec-1999, 12:38
As a general rule with LF, it is worth considering what plane you want entirely in focus, and then consider how much in front and behind you want 'reasonably' in focus.

When my subjects are trees, I generally want a whole tree to be in focus, so I usually have a vertical back.

For a 210mm lens used on 35mm or 5x4, at a given aperture, the DOF will be identical on the film. However, for a given print size, the 5x4 will need less enlargement, so will have more DOF.

However, with modern sharp lenses and film, and lack of grain in large formats, my senses are much more critical and I don't generally like to rely on DOF in LF.

Todd Caudle
28-Dec-1999, 19:10
from above post: "For a 210mm lens used on 35mm or 5x4, at a given aperture, the DOF will be identical on the film. However, for a given print size, the 5x4 will need less enlargement, so will have more DOF. "

Actually, if both 35mm and 4x5 are using the same focal length (210mm in this example), then to duplicate the same 24x36mm image area in the 35mm would require the exact same enlargement of that portion of the 4x5.

Bill C
28-Dec-1999, 22:06
Hi Todd. I may be missing something here, but are you sure your friend didn't mean to use a swing (not tilt) to bring both trees (trunks, anyway) into sharp focus? That is, the plane of focus would remain vertical but go through both tree trunks.

Todd Caudle
29-Dec-1999, 21:03
Hi Bill. Actually, to illustrate my difference of opinion with my friend even further, he says he uses movements in almost every shot he takes. For instance, one shot of his that I like very much is from a pretty good distance from a lake, with mountain peaks in the background and not much of anything real close in the foreground. When I see the shot, I see a shot that's at or near infinity all the way. So I asked him what he would do in that situation and he said he used tilt. Sure, if I was shooting at the salt flats in Death Valley (which I will be when I attend Ron Wisner's workshop in late- January), I would know right from the start that I want to use either front forward tilt or rear backward tilt. But in his example with the lake and mountains, I'd simply focus at some point well into the scene and stop the lens down to f22 or f32. I am nearly convinced that there's something very elemental to LF photography that is going right past him.

Blake
30-Dec-1999, 12:57
well the latest example you've given might prove overkill on both counts...you might have enough dof at f8 even with the 210 if the scene begins and ends near infinity depending on a lot of other things that i can't see and haven't been told.

timparkin
25-Nov-2007, 04:44
If you use a very small circle of confusion because you want to enlarge your photos to maximum and allow people to view up close, then you don't get much depth of field at all.. In fact, the only part of a shot that is totally in focus (if you ignore coc) is the plane of focus. So if you want to avoid diffraction and want the maximum sharpness on parts of your picture and you CAN tilt to get focua, then you probably should.. Then you can decide what parts of your picture you want to get CLOSE to critical focus..

I personally am with your friend on this one..

Tim

N Dhananjay
25-Nov-2007, 08:43
1) A lens is agnostic with regards to what size of film is there behind it - that is, it provides the same DOF at any f stop, regardless of format.

2) With regards to focus first, tilt second versus tilt first, focus second - at one level, the answer is that it really does not matter as long as you eventually end up with the (or near the) optimum combination of movements and f stop. Which one works for you might be a function of individual idiosyncracies. As pointed out, it eventually becomes fluent, something that you perform with little conscious thought much of the time.

Time permitting, you can experiment with various combinations that provide you with the optimum setting. Typically, focus at one extreme and make movements while looking at the ground glass to see if the movements are making things better. The exact method depends upon the mechanism for the movement. For example, if you have base tilts, as you tilt down, the bellows extension at the bottom will change less than at the top, which means the bottom areas of the scene (and the top part of the inverted ground glass) are going to be 'sliding' into focus to a greater extent than the top. So, you would focus on the distance, tilt to bring the foreground in, refocus to bring the distance back in if it has drifted out, tilt a bit more and so on till everything is in focus. If you want to see if movements have helped, without movements, mark the focus spread (make marks on the focussing bed when focussed on the nearest object and then the furthest object). Now predict the required movement and again mark off the extremes of the focus spread - if the focus spread has reduced, movements have helped.

My suggestion is that, at least initially, it is useful to visualize the entire area as a three dimensional volume to predict the combination required. You can then see how close your guess was. Do this for a little while and soon, it just becomes part of the hand-eye-brain mechanism and becomes fluent.

Cheers, DJ

Donald Miller
25-Nov-2007, 09:50
I am with your friend if the trees are less than 25 feet tall (considering the distance from front to rear as my basis) and with the nearest ground desired to be in focus. Given this scenario, I would use front tilt and focus 1/3 of the distance down from the top of the nearest tree and also the nearest point of desired focus. My plane of focus would be the from the nearest ground to 1/3 down from the top of the tree. I would determine the tilt needed before I even thought of stopping down the lens.

I use tilt on virtually every landscape shot and I try to keep my F stop to F22 or less on 4X5.

Of course time and experience will determine what works best for you.

David_Senesac
25-Nov-2007, 13:23
I'm guessing you are describing a situation where the closer tree is lower on the frame with the further tree above. Generally since my Wisner doesn't have easy detents for perpendicularity of front/rear standards, I have to focus all quadrants regardless thus using movements adds nothing I would not have to do anyway. On the issue of using tilt I might agree with your friend as long as the trees took up about the same percentage of frame vertical and the tilt didn't throw any other elements out of focus. Although just stopping down is likely to still do an acceptable job of providing good resolution, one could expect using a small amount of tilt would squeeze out a bit more resolution by moving the critical focus plane say from the mid height of the first tree to mid height of the second tree. Now as the closer tree becomes a larger percentage of the frame's vertical, one ought to decrease any tilt accordingly and focus stronger on the closer tree. Also if the front tree is the more compelling central element, one may decide to focus critically on it in order to exact the best fine details bottom to top. Thus as in many landscapes there are considerations to be noted and compromises to be chosen depending on how one wishes to render a scene. ...David

Phong
25-Nov-2007, 14:20
Guys, this thread is like 8 years old.
I wonder if they are still friends. :-)

cheers,

riooso
15-Dec-2007, 18:07
Phong you are a maniac! You must be reviewing your technique also. I wondered the same thing.

Richard

Michael Gordon
16-Dec-2007, 09:13
The poster ditched LF for digital a few years back, and if the friend is who I think he is, he's also GD (Gone Digital).

No further responses will be necessary :D

Alan Davenport
16-Dec-2007, 09:57
Guys, this thread is like 8 years old.
I wonder if they are still friends. :-)

cheers,

Geez, and I was all set to enter the fray...

uh, never mind, I guess.

danno@cnwl.igs
10-Mar-2011, 13:39
Reading this interesting old thread raises the question: are two lenses of the same focal length likely to show the same characteristic depth of field if they are of different design?

For instance, I have a Tele-Xenar 360 mm, and a Zeiss Tessar 360 mm. One is a retrofocus design, with a nodal point of about 270 mm at infinity, while the Tessar is dead sharp at 360 mm.

Paul Fitzgerald
1-Apr-2011, 22:49
"Reading this interesting old thread raises the question: are two lenses of the same focal length likely to show the same characteristic depth of field if they are of different design? "

Yes, they will show the same DoF with the same FL, aperture and focus distance with any normal design.

There could be some esoteric designs out there that beat the system but all normally available lenses use the same DoF tables.

Ole Tjugen
2-Apr-2011, 04:10
There could be some esoteric designs out there that beat the system but all normally available lenses use the same DoF tables.

which doesn't necessarily mean that the perceived DoF will be the same in the finished picture. Try comparing otherwise identical shots with a Tessar, a Dynar and a Plasmat...

Oren Grad
2-Apr-2011, 09:01
which doesn't necessarily mean that the perceived DoF will be the same in the finished picture. Try comparing otherwise identical shots with a Tessar, a Dynar and a Plasmat...

Yes. Perceived DOF depends on how the transition to OOF is rendered. See the section "Depth of field in practice - Departures from theory" in Ray's "Applied Photographic Optics". In my copy, which is the 1988 edition, it's section 22.3.1, p. 185.