where to focus? Newbie needs help.

[niubi]

Hi all,

An embarassing newbie question...(Please note: I am only practicing camera movements at this stage and have not even used any film as yet (still waiting on darkroom equipment to arrive).

When photographing a landscape, I can't seem to make everything look sharp on the ground glass. That is, I first sharpen the background, then use front (or back) tilt to sharpen the foreground but (depending on the scene) the middle section is usually still out of focus on the gg.

1. Can I assume that when I stop down (the aperture) that the dof will sharpen the middle section? When I stop down I see some improvement on the gg but then it becomes too dark for me to accurately assess.

2. if stopping down solves the problem, how do I know which aperture to use? I have thought about just using f22 for my 90mm and f32 for my 150mm and 210mm. I'm sure there is a more scientific method but is there a simple but accurate method?

Thanks.

[Juergen Sattler]

There are no embarassing question! It is perfectly normal that the middle of your picture will go out of focus as you shift the focal plane - which is what you are doing when using tilt. In LF you use very small apertures to render everything perfectly sharp - for landscape photography I use f32 and even f45 all the time.

You need a very good darkcloth that you can close around you, so that stray light stays out. Then you stop down the lens and using a good loupe you will actually see how DOF improves. You are right that the GG gets very dark, but after some practice you will be able to see the change in DOF.

Very often you don't even need any movements (I know, you bought the camera because it allows for movements, but more often than not, they are not necessary) and in that case you can just use hyperfocal distance focsuing to render everything in focus.

Keep playing with your new camera - you are discovering the ups and downs of LF - perfactly normal. I would recommend that you get some polaroids and a holder and expose some sheets - this way you can see in the field what works for you and what doesn't.

[Vaughn]

Hello!

First, welcome and congrats on being smart enough to ask good questions!

Second...you should not start out by focusing on the background. As a real general rule, focus about 1/3 of the way into the scene...then apply a slight forward tilt with the front standard (or a slight backwards tilt of the back standard). In most cases, closing down to f22 while looking at the GG should give you a good idea if all is well...and there should be enough light to still see the image without problem.

After the initial focusing, apply the tilt to get the fore and background into focus as best as possible. Then as you close down the lens, see what areas (fore or background) come into focus first. In theory, they should come into focus at about the same time. If the background comes into focus first, you need to focus a little closer -- and visa versa.

When I find the f/stop that gets everything into focus, I tend to close down another stop or two to make sure. Lenses are sharpest in the middle f/stops and if one can use them and get everything in focus that is the best way to go. My scenes in the forest tend to need more DoF than f22 can give me (300mm lens). I prefer a slightly less sharp in-focus image than a very sharp out-of-focus one.

If your GG seems too dark to check the focus at f32 or higher, try angleing your head so that you are looking at the GG along the same path the light is coming from your lens. In other words, look at the corner of your GG, but towards the center of your lens. It will be brighter that way.

You probably get more in-depth answers from others so I will leave that to them. This is how I focus...others' milage may differ.

Vaughn

[David Karp]

Not a dumb question.

Try using the focusing method discussed here: http://www.largeformatphotography.info/fstop.html

Focus on the table at the end that shows the distance (spread) in mm between the two points you want sharp, and the appropriate f/stop to use.

This system works very well. It is especially helpful if your landscape has some of those pesky vertical objects in it (like trees) that mess up your attempt to use tilt to gain DOF!

[Bruce Watson]

Not a dumb question.

Try using the focusing method discussed here: http://www.largeformatphotography.info/fstop.html

Focus on the table at the end that shows the distance (spread) in mm between the two points you want sharp, and the appropriate f/stop to use.

This system works very well. It is especially helpful if your landscape has some of those pesky vertical objects in it (like trees) that mess up your attempt to use tilt to gain DOF!

What Mr. Karp said.

Note also that focusing is a compromise in many situations. Unless of course you find something planar to photograph, and there are more planar situations out there than I would have believed when I was just starting out.

I often approach this situation in one of two ways. First, get the foreground and background in focus as you have, then use the main focus adjustment on the camera to move the plane of focus into the scene some. What you want is for the middle ground and the foreground/background to be equally out of focus. Then use a small fstop to pull everything in.

An alternative is to use a little more tilt so that you have the foreground in focus, and the middle of the background in focus (say, half way up the mountain in the background). This brings the plane of focus closer to the middle ground rendering it more in focus. Again, fstop to pull everything in.

Which method I use depends on the scene (and to be honest, probably on my mood as well). As the old sage said: "There are many paths to the waterfall." You just have to find the path that works best for you, and the only way to find your path is to make lots of photographs so that you can learn from your mistakes and your successes.

Good luck, and happy experimenting.

[niubi]

Ok. So let me get this right...
1. Focus in the middle of the scene (either one or two thirds into the scene).
2. Use front or back tilt to get the foreground in focus.
3. Stop down to get the background in focus.

Not sure about the HANSMA technique as I tried the calculation but came up with some very small apertures. I would have thought that diffraction is rife at around f45+.
Anyway...more practice...

BTW, Thanks for all your help.

[riooso]

When you say tilt you mean just a couple of degrees? Right?

Richard

[steve simmons]

There is a very good article on just this topic in the Sept/Oct issue of View Camera and there was another very good article on movements and foccussing in the Jan/Feb issue of 07. The S/O issue should be out on the newsstands now.

steve simmons

[niubi]

When you say tilt you mean just a couple of degrees? Right?

Richard

It's usually more than a couple of degrees. That's why I imagine I am doing something wrong as everyone on this forum talks about how little tilt is required. I have a Tachihara so I can't measure exactly how much tilt but from my reading, I expected much less tilting required. :(

Steve, I will try to get a hold of the View Camera articles. Thanks.

[Leonard Evens]

The basic method is described quite well as David Karp told you elsewhere in this large format web page.

But let me outline briefly how to go about it. This should work fine for landscapes. It needs some modifications when part of the range is in the close-up region, usually defined as less than 10 times the focal length.

First, see if you can manage without using a tilt, just with DOF, at an an acceptable aperture. To this end, choose a near point and far point you want in focus. Focus first on one---say the far point---note the position of the standard on the rail or camera bed, then focus on the other---say the far point---do the same and determine the distance between the two positions in mm. That is called the focus spread. Set the standard half way between (but see below for variations of this method). Next, multiply that distance by 10 and divide it by two. That is the lowest f-number, i.e., largest aperture which you should consider. (This is based on a criterion of sharpness called the diameter of the the maximum circle of confusion (coc) of 0.1 mm for 4 x 5 film.) This choice ignores the possible effects of diffraction. You can also use Hansma's calculation or table to determine the aperture which gives you the best you can do combining defocus and diffraction. As you say, that will often suggest a very small aperture, and in my experience I find I usually don't have to go that far. But the two f-numbers will give you a range of apertures to consider. In deciding which aperture in that range to use, you also have to consider the resulting shutter speed and whether or not subject motion will be an issue. I find that just stopping down one to two stops further than the widest acceptable aperture will work.

The one third of the way into the scene rule, although often recommnded, works exactly for only one distance, namely one third of the hyperfocal distance for your chosen aperture. It doesn't make sense if you want infinity to be in focus. (What is one third of infinity?) I doubt that anyone actually calculates where one third into the scene is, and in fact it just means you focus closer to the near point in the scene than to the far point. Focusing halfway between the rail positions of the far and near points is the theoretically correct place to focus, although in practice you may want in some cirumcstances to favor either the foreground or the background. In that case you would move closer to the the corresponding point on the rail.

If the aperture and shutter speed combination you come up with is impractical, consider using a tilt. You can find references elsewhere in the large format page for how to do this, but let me briefly outline what I do. First visualize where you want the plane of exact focus as well as the upper and lower limits of what you want to end up in focus. The exact focus plane should split each vertical plane in this DOF region in half. Choose. a near point and a far point which will be in the exact focus plane. Guess an appropriate tilt. It doesn't much matter what it is as long as it is small, say about five degrees. If your camera doesn't allow you to measure the tilt in degrees, don't worry. Just eyeball it and keep the tilt pretty small. Now refocus on the far point, note the position on the rail of the standard, and refocus on the near point. If you have to increase the distance between the standards to refocus on the near point, then increase the tilt. If you have to decrease the distance between the standards to refocus on the near point, decrease the tilt. With some practice, you should find that a few iterations of this procedure will get you the tilt where both near point and far point are simultaneously in focus, or close neough that it doesn't matter.

Next choose an upper point above the exact focus plane you want in focus and a lower point below the exact focus plane you want in focus. They need not be at the same distance from the lens. Focus first on the upper point, then on the lower point and determine the focus spread on the rail between them. Focus halfway in between. (In some circumstances, you may want to favor either the region above the exact focus plane or below it, in which case you would adjust the focus position accordingly.) Now use the focus spread just as previously to determine the aperture..

When using these methods, you should also stop down and inspect the image on the gg to see if everything looks right to you. Unfortunately, as you have discovered, if the taking aperture is much smaller than f/16, you aren't going to be able to see much of anything. Even so, if you stop down to f/16 or f/22, you should be able to get some idea if you have it right.. Also, even wide open, as someone else suggested, when you are at your desired focus point, the limits of the desired DOF region should appear "equally out of focus". If they don't you may want to adjust the focus position.

[niubi]

Thanks Leonard. This is very comprehensive. I may have to read it a few more times with camera at hand, so I can practice. This is the type of explanation I was looking for.

Your calculation also comes up with more realistic apertures. I just read in another thread you posted, that your calculation is based on a given viewing distance of the final print. So I'm assuming if I wanted a closer viewing distance I would have to alter the calculation.

[David Karp]

. . . I tried the calculation but came up with some very small apertures.

That's the rub sometimes. Due to the nature of the scene, sometimes the choice is a very small aperture or no photo. I just decided that if diffraction really turns into a problem, I would not enlarge a photo that I like to the point that it becomes an issue. In practice, since most of my photos are not enlarged from 4x5 to more than 11 x 14 that it has not made a practical difference. In the meantime, I got some photos at f/45 that would not have worked as well with a larger aperture.

[Dan Schmidt]

That's the rub sometimes. Due to the nature of the scene, sometimes the choice is a very small aperture or no photo. I just decided that if diffraction really turns into a problem, I would not enlarge a photo that I like to the point that it becomes an issue. In practice, since most of my photos are not enlarged from 4x5 to more than 11 x 14 that it has not made a practical difference. In the meantime, I got some photos at f/45 that would not have worked as well with a larger aperture.

As i understand it at f/45 with say a 150 mm lens your aperture diameter works out to a bit over 3 mm. I find it hard to imagine diffraction would be an issue since this is a few thousand times greater than the wavelength of visible light.

For most apertures on our large format cameras I can't see how diffraction would be an issue. Am i wrong?

[steve simmons]

The advantage to getting the articles, or one of these books

Large Format Nature Photography by Jack Dykinga
User's Guide to the View Camera by Jim Stone
Using the View Camera that I wrote

is that there are illustrations and diagrams to go along with the written material. Some of us, and I am one of them, learn better visually, and the pictures will help a lot.

If you like the math fine but it is not necessary to use, or even understand it, to solve your problem.

steve simmons
www.viewcamera.com

[Leonard Evens]

Dan,

Diffraction is measured by the size of the Airy disc, or alternately in terms of effective lp/mm as measured on the film. These depend on the relative aperture or f-number, not simply on the physical dimensions of the aperture. Both the physical size of the aperture and the distance of the film from the lens enter into the calculation and it is the ratio of the second to the first that matters. When focused at infinity, this ratio is the f-number. When closer than infinity, this ratio is the f-number multiplied by one plus the magnification, which is called the effective f-number. But, except when one is in the close-up range, the effective f-number is so close to the f-number that it doesn't pay to worry about the difference.

The usual rule of thumb for photography is that the resolution limit imposed by diffraction at the film is 1500 divided by the (effecitve) f-number. For f/45, this is about 33 lp/mm. This is less than a good large format lens can deliver at its optimal aperture, so diffraction at f/45 will typically play a significant role. The way this enters in DOF considerations is that you won't get quite the desired resolution at the limits of the specified DOF region as you would think based just on defocus calculations. By stopping down further, you will in effect extend the diffraction-free DOF region, thus increasing the resolution because of defocus at the place you need it. The aim is to calculate just how far you have to stop down to counterbanace the effect of diffraction. To get this right requires sophisticated analysis based on MTF curves, But one can come close using certain rough rules of thumb, and that is what Hansma did. Perhaps someone who is a better physicist than I am can provide a more intutitive argument explaining this. But it doesn't really matter in the end whether the results of such a calculation square with our intuition.

[Leonard Evens]

Thanks Leonard. This is very comprehensive. I may have to read it a few more times with camera at hand, so I can practice. This is the type of explanation I was looking for.

Your calculation also comes up with more realistic apertures. I just read in another thread you posted, that your calculation is based on a given viewing distance of the final print. So I'm assuming if I wanted a closer viewing distance I would have to alter the calculation.

The usual calculations are based on an 8 x 10 print viewed at about 10-12 inches, which is the normal distance most people would view such a print. Indeed, it is as close as most adults can get. If you produce a larger print, and get proportionately further away, the calculations for an 8 x 10 print still hold. But if you still view the larger print from 10-12 inches, as many people would do, in principle, you would have to take that into account. But some sample calculations show that this would lead quickly to unrealistic apertures. For example, suppose a coc of diameter 0.2 mm (corresponding to 5 lp/mm) is acceptable at the level of the print. If you enlarge 2 x for an 8 x 10 print, that would require 0.1 mm (corresponding to 10 lp/mm) in the film. But if you enlarged an additional 3 X for an 24 x 30 print, and still viewed that at 10-12 inches, you would need a coc at the level of the film of 0.033 mm (correponding to 30 lp/mm). To obtain that at the limits of the DOF region would require stopping down an additional 3 stops. If f/32 worked for an 8 x 10 print (or a larger print viewed further way), then you would need something like f/90 for the larger print, and that is ignoring diffraction. Clearly, this is impractical.

The upshot is that if people insist on viewing large prints from close-up they are going to see much reduced depth of field. Even in the exact plane of focus, where the resolution would be limited by lens aberrations and diffraction, there is a practical upper limit for how far you can enlarge without loss or quality in the print viewed close up. For example, if you stop down to f/32, just from diffraction, the resolution at the film level would be about 47 lp/mm. The optimum resolution ignoring diffraction might be 60 lp/mm. The combination would result in something between 26 and 36 lp/mm. If 5 lp/mm were acceptable to you when viewing the print, you would be limited to enlarging between something like 5 and 7 times, and if you did, you would end up with zero DOF. It gets even worse if you take into account the resolution of the film.