Using tilt when shooting landscapes

[stephen gale uk]

I have an Ebony RSW45 which I am using for landscape work. This

camera is a lightweight non-folder with just front tilt and front

rise/fall (there was no way that I could carry my Cambo monorail out

into the mountains!). Besides which, tilt and rise/fall should be

just about enough for most landscape work. I am using it with a

Schneider SA 90mm.

 

To achieve the maximum DOF, I have been either stopping the lens

down to f32 and using a DOF chart to calculate the near/far points.

Or I have been using a bit of front tilt. Using the front tilt has

the advantage of not requiring such a small aperture (which can

result in long exposures causing motion blur).

 

My question is how do I accurately predict the near/far focussing

points when using the front tilt for a given aperture?

 

My current approach is to do this purely visually. First, I focus

on the far point and then to use the tilt to get the near point in

focus. This seems to be a bit of an iterative process, but seems to

work. But it has the drawback that it requires focussing at the

required aperture (rather than full open and then stopping down).

Is there a better approach?

[davidroossien]

Have you read this?

 

http://www.largeformatphotography.info/how-to-focus.html

 

The visual approach usually means focusing with the lens wide open. Hope this helps.

[troyammons]

IMO tilt is overhyped for 4x5 and under especially for a 90mm lens.

 

There are just a lot of time when you cant even use it. If you are in a flat plane type location it is great though.

 

A 90mm lens at F32 on 4x5 is in focus from 5' to infinity, so using tilt in that case might not be even nessesary unless the camera is right on the ground.

 

All that said the easiest camera I have ever used for tilt was a Sinar. It has a built in tilt calculator to the focus knob.

 

Personally I shoot a super graphic now and use a similar method as the link above.

 

Sort of a far focus, near focus method, then back and forth mini adjustments refocusing everytime until I get it dialed in. Then I check the extreme corners and look for other objects out of the plane of sharp focus.

 

90% of the time 3-5 degrees is dead on.

 

With a Sinar you just follow the instructions and read the tilt off the knob, based on far near focus points.

[leonard_evens]

I echo the above recommendation about studying the how-to-focus web page.

 

But let me address your specific question. Tilting changes the plane of exact focus. This plane is ordinarily parallel to the film plane. It has nothing to do with the aperture. Your method of doing it works well. Just keep in mind the following rule. If you have to increase the lens to film distance when focusing on the near point, increase the tilt and if you have to decrease the lens to film distance when focusing on the nearpoint, decrease the tilt. There is another method due to Bob Wheeler, which is described in several places, including my essay www.math.northwestern.edu/~len/photos/pages/dof_essay.pdf.

But many people find this method too complicated, and in any event it is not clear it would do what you want.

 

You seem to be describing the problem backwards from the usual way to express it. Bear with me while I review the basic facts. Afterwards, I'll try to answer your question.

 

Usually, you start with a near point and far point which you locate on the rail and you set the focus half way in between. You then choose the f-stop based on the total focus spread between near and far points. When you tilt, you first choose the plane of exact focus, as above, and you then choose points above and below that plane, mark positions on the rail and then use that focus spread to determine the proper aperture. Trying to determine the tilt with the lens stopped down is going to be frustrating at best and not likely to lead to success.

 

The zone in focus has the form of a wedge centered on a line (usually) below the lens. That line is called the hinge line. The exact plane of focus is centered in that wedge in the following sense. Any plane parallel to the film plane meets the wedge in a band bounded above and below by two parallel lines. The exact plane of focus intersects that band in a line bisecting it. The top and bottom boundaries of the wedge pass through the points chosen above and below the exact plane of focus. With the tilt fixed, focusing pivots the entire wedge about the hinge line as well as modifying its shape.

 

The size of the wedge is related to the aperture in a fairly complex way. Namely, at the hyperfocal distance, the extent of the band above and below the plane of exact focus is very close to the distance of hinge line to the lens. At other distances, you just multiply by the relevant scaling factor, so if you are 3 times as far away as the hyperfocal distance, you multiply the hinge distance by three to find the extent above and below in focus. The hyperfocal distance is of course determined by the aperture (as well as by the focal length and the coc).

 

So supposing you really want to do what you say, here is how I would go about it. First choose the tilt as described above with the lens wide open. Then find the hyperfocal distance at the desired aperture. For a 90 mm lens at f/32 with a coc of 0.1 mm, the hyperfocal distance is 2.53 meters or 8.3 feet. (Generally, you square the focal length and divide by the product of the coc and F-number.) Estimate the distance to where your exact plane of focus passes directly under your lens. Measure up and down that distance from the plane of exact focus at the hyperfocal distance. For other distances, scale accordingly.

 

Calculating the distances on the rail is a bit more complicated, but for sufficiently distant points, a reasonable estimate is the product of the F-number and the coc. Twice that is close to the total focus spread in most cases. I go into this in somewhat more detail in my essay. You have to include a factor of one plus the magnification and another factor related to the tilt angle.

[stephen gale uk]

Thanks for all the information. It is really useful.

 

So, I now understand how to set up the exact panel of focus. I understand how to calculate the DOF using the appropriate charts. But what I am slightly confused about is where to measure the DOF from when using a tilt. Is it from the middle of the film plane, the middle of the lens panel, or some other point?

 

Thanks in advance.

[troyammons]

>Is it from the middle of the film plane, the middle of the lens panel, or some other point?

 

Some other point.

 

As explained above when you tilt the lens forward/down, the plane of sharp focus goes from being vertical and equidistant, (IE for instance maybe 10 feet in front of and 30 feet behind the point of focus)

 

to a tilted plane of focus that would ideally go from the ground at the base of the camera out, but wedge shaped and maybe only a few inches of DOF at that nearest point expanding out to and infinite dof somewhere far out perpendicular to that plane.

 

Everything in the wedge will be in focus. If you are in the desert and the ground plane is within the wedge everything would be in focus. If you are on the east coast with the same setup and a tree top 40 feet away is outside of the wedge, even though that may be your focal point prior to tilting it will be OOF, the tree top that is.

 

WO the wedge would be narrower than stopped down, but how much is beyond me. I never have figured that out although I am sure it could be calculated. Not sure if I really even want to know.

 

For landscape I typically tilt WO and then stop down to where that particular lens is sharpest.

[leonard_evens]

Stephen,

 

I already told you the answer to your question, and Troy expanded on it. From you question, I think you haven't yet understood what the zone of focus looks like in case of a tilted lens.

 

You should read what I said very carefully, and draw yourself some pictures. At the large format web page www.largeformatphotography.info

you can find a variety of links which have pictures of the wedge, the hinge line and all the rest. That includes my essay, which I referred to previously. Look also at the Merklinger references, and somewhere---I'm not sure where---there is an animated version. The geometry is a bit complicated and in some ways counter-intutitive. I think when you get it straight in your head, it will all become much clearer.

 

If, after that, you still have further questions, please feel free to ask.

[john mackay]

<p>Hi Stephen,</p>

<p>I also had trouble understanding how the plane of focus transitioned with

tilt and subsequently made my own diagram (relatively accurate) so I could see

the effect for a 1 in 2 gradient for a 90mm lens (below)</p>

<p><img src="http://www.wildframe.net/images/dof_wedge.gif"></p>

<p>However, I think that the best visual demonstration of this is

<a href="http://www.trenholm.org/hmmerk/ViewCam3.mov">Harold Merklinger's

QuickTime Movie demonstrating base tilt</a> (34kb). You can view other QuickTime

Movie demonstrations by Harold

<a href="http://www.trenholm.org/hmmerk/HMArtls.html">here</a>.</p>

<p>Hope this helps either you or someone else.</p>

<p>Cheers...John.</p>

[michael_briggs2]

Stepen, re your first question states that you are using a DOF chart, and your second question implies that you are still interested in measuring the DOF. There is an easier way to figure out the correct aperture, which Leonard mentioned in one of his answers: "Usually, you start with a near point and far point which you locate on the rail and you set the focus half way in between. You then choose the f-stop based on the total focus spread between near and far points." With this method you make easy measurements on the camera: the positions of the standard when the near and far objects that you want in focus are in focus. You then difference these two values to get the focus spread and consult a table to get the aperture to use to get the two objects in focus. This is easier then trying to judge the distances to the objects; you don't need to use a DOF chart at all. One focus spread table works equally well for all of your lenses. The method is explained in greater length at http://www.largeformatphotography.info/fstop.html

 

Understanding this method of setting the aperture also makes very clear that the goal of setting up a tilt for focus is to minimize the focus spread.

[brian_ellis16]

"My question is how do I accurately predict the near/far focussing points when using the front tilt for a given aperture"

 

You first tilt, then determine the near/far points with the tilt in place.

[brian_ellis16]

In re-reading my previous answer I realized that in the interest of brevity I perhaps wasn't clear. My point was that after first focusing on the farthest object that you want to appear to be in focus in the photograph you then tilt to bring the nearest object as closely into focus as possible, then focus on that object if the tilt alone wasn't sufficient, then with the tilt in place measure your near and far points (sometimes the tilt alone will bring the near into focus without disturbing the far, in which case there's nothing to measure).If you attach a milimeter ruler somewhere on the bed of your camera, and place a little pointer somewhere on the part of the camera that moves as the lens moves forward and backward, you can then easily measure the distance in milimeters that the front moves between the near and far focus points. Once you know the distance in milimeters between your near and far focus points (after having tilted if necessary) then you can use a table to determine your optimum aperture and you don't have to worry about measuring depth of field at all or trying to tell what's in focus by looking at the ground glass after stopping down, something that's often impossible to do if you've stopped down to f32 or if you're in dim light.

[leonard_evens]

John's diagram shows you the wedge, but remember it is really three dimensional and the diagram only shows a plane cross-section. Also, I believe that the best focus should be centered vertically between the acceptable focuses above and below. That is confirmed by the many diagrams in Merklinger's files.

 

Let me add one additional point. The usual advice is first to choose the tilt so that the plane of exact focus passes through a previously chosen near and far point. The aperture is then related to the vertical extent of the wedge above and below the exact focus plane. But how are you supposed to choose the near and far points to choose the tilt in the first place? One method is to look at the vertical extent you want in focus in the foreground and the vertical extent you want in focus in the background. You might then choose the near point to be the middle of the foreground vertical range and the far point to be the middle of the background vertical range. Unfortunately, the foreground and background vertical ranges may not be consistent with a wedge centered on a horizontal line below the lens. Increasing the tilt moves the hinge line down and decreasing it moves it closer to the lens (never closer than the focal length), but generally, no choice of tilt will exactly encompass the foreground vertical extent and the background vertical extent. So usually we have to compromise and concentrate on which is most important. Usually, it is no problem getting the background entirely in focus, but we have problems with the foreground because the vertical extent of the wedge is so small near the lens. So, it is important to realize that the geometry place limitations on what is pssible, and not everything is feasible.

 

 

In practice, I find that I often have to experiment a bit with the choice of the near point, i.e., with the tilt angle. After I've tried it one way, I stop down to get some idea of what is in focus at the taking aperture, which I determine by the usual rules using focus spread for high and low points about the plane of exact focus. (Unfortunately, one can't usually see much below about f/16 to f/22, so it may be necessary to extrapolate from something less than the taking aperture.) I may then decide to modify the tilt because of what I consider most important in the foreground. so the process becomes a bit iterative. And, in many cases, I just decide the shot it not possible, so I give up and try some other vantage point and concept of the scene.

[pvp]

<I>how do I accurately predict the near/far focussing points when using the front tilt</I><BR>

<I>...I am slightly confused about is where to measure the DOF from when using a tilt...</I><P>

To oversimplify the answers already given: you don't. You choose the nearest and farthest points that you need to have in focus, then tilt the lens so both points are in perfect focus with the lens wide open, then stop down. At that point, you will have perfect focus everywhere on a plane that includes both of those points, and your DOF will extend more or less <I>vertically</I> above and below that plane. <U>Published DOF charts are useless</U> once you add tilt into the equation, because they seldom include calculations for tilted lenses. For typical landscapes, the effective DOF can often be infinite.<P>

My suggestion is, don't obsess over the numbers. Measuring DOF with tilts is usually best done with the focusing loupe and the lens stopped down. If it's in focus on the groundglass, it will be in focus on the film.

[wieslaw1]

I am not sure if everybody reading this discussion is aware of the

fact that tilting the lens in respect to the film plane offers focus

PLANE which extends from the lens towards the infinity (on the

horizon), while stopping down the lens works differently than in

the case when there is no tilt, i.e. in all the ordinary single reflex

or range finder cameras. In the latter case, as the lens is

stopped down or opened, the depth of field increases or

decreases respectively, in front of the lens, from short to longer

distances.

 

In a view camera, once you tilt the lens, the stopping down the

aperture increases the depth of field but in directions (plane)

PERPENDICULAR to the viewing direction along the horizon line.

On the graph offered by John this will be manifested in a

progressively more diffused contours of the subject away from

the �focus plane�, more or less symmetrically towards the close

and far focus planes indicated. The more lens is stopped down

the sharper will be the subjects located away (above and below)

from the mid-focus plane.

 

If you are photographic in the mountains at high elevations, and

you want to focus so that a deep valley, below your vantage

point, appears sharp as well as high peaks, than the above

graph has to be modified to account for this.

If the valley is 1000 m deep and you want to have sharp focus

starting from 2 yards to infinity, say, with a 150 mm lens, you

need f/64 aperture on the lens.

[leonard_evens]

"Measuring DOF with tilts is usually best done with the focusing loupe and the lens stopped down. If it's in focus on the groundglass, it will be in focus on the film."

 

This generally true, but there are a couple of problems with it. First, few people can see much of anything when stopped down below f/16-f/22 because the image just isn't bright enough. A second factor is that using a loupe in effect means choosing a different maximal allowable coc. Ideally, if you view a 4 x 5 gg from about 6 inches, that would be equivalent to viewing an 8 x 10 print at 12 inches or a larger print proportionately further away. If you use a 2 X loupe, that is roughly equivalent to viewing from 6 inches. A higher power loupe gets you even closer which also magnifies the blur discs and decreases the apparent depth of field compared to a print at normal viewing distance. Thus, relying on a loupe after stopping down leads you to overestimate how far you have to stop down for desired DOF, so it is generally safe to do. But, when subject movement is an issue, you may want the largest aperture which will work.

 

Since DOF is roughly proportional to the F-number, one could always balance these different considerations by making simple calculations, but I suspect few people would want to do that.

[pvp]

<I>"...few people can see much of anything when stopped down below f/16-f/22 because the image just isn't bright enough."</I><P>

That may indicate that a better darkcloth is in order?<P>

 

<I>"...one could always balance these different considerations by making simple calculations, but I suspect few people would want to do that."</I><P>

 

More people might be willing to take a measurement and refer to a chart. Another way to describe depth of field is, "optimal sharpness over a range of distance." Hansma and Peterson covered that subject quite well, proving that for any given focus spread, there is a particular f/stop that will yield the smallest possible circle of confusion, taking into account both defocus and diffraction.<P>

I keep a small chart taped to my GG protector, which gives the optimal f/stop for any given amount of focus spread. After applying tilts and swings to best advantage, simply focus -- with the aperture wide open -- on the nearest point that must be sharp and note the position of the lens on the camera bed. Then focus on the farthest point and measure the distance the lens moved. After measuring the spread, set the lens midway between the near and far points, set the f/stop defined by the spread and take the photo. This method also eliminates the darkcloth problem, since everything is done wide open until you stop down to shoot.<P>

The optimal f/stop is given by <B>N= SQRT(375*dv)</B> where dv is the focus spread in mm. Note that the optimal f/stop depends ONLY on the distance the lens must be moved to sharply focus both near and far. Also, in this case "near" and "far" don't necessarily refer to distance from the camera, but rather the movement of the lens to focus everything of importance in the scene. With a tilted lens, "near" might mean the top of a tree, and "far" a rock at its base. The groundglass reveals all...<P>

Obviously, this ignores the problem of a moving subject, but that is seldom the issue in landscapes.

[david_senesac]

The above answers directly apply to your question as you probably intended stephen, and I'll dampen such a bit with another landscape focus reality.

 

More often than not in landscapes, the planes and objects to focus in one's frame are impossible to all get in sharp focus. Thus one needs to estimate from experience where in a given frame to focus the near and far, and to a lesser extent left and right points in a frame for whatever intended image one is creating. Its mostly guessing from experience and then stopping down as much as is appropriate to take care of the compromises. In many non ideal situations focusing on the furthest point at the top of a frame and the nearest foreground point at the bottom of a frame, then optimally tilting is not the wisest choice.

 

As a simple example, if the top half of one's frame is a 30 foot tree 30 feet away on level ground, and the near point is a short single flower 5 feet away, it is impossible to get the foreground, the bottom of the tree, and the top of the tree all in sharpest focus even if stopped down to f/45. Thus one might choose to place the sharpest far focus half way up the tree maybe where detail is most interesting and the near sharp focus 10 feet away where a larger group of other flowers provides more detailed interest. By doing so the bottom of the tree would also now be sharper and the near flower less so. Knowing how much one is going to give up on the near focus versus the middle is an experienced guess. Additionally one might stop as far down as a lens provides, but if there are moving objects like wavy water, or breeze on vegetation, one has to compromise that with considerations of adequately stopping motions. Thus focusing on landscapes is often requires making decisions that are not at all mechancal and exact but relie on one's experience. Thus the notion of accurately predicting optimum focus in some cases may not be as important as just making wise decisions on how to compromise.

 

...David

 

www.davidsenesac.com