How do you/did you learn to work with DOF?

[sully75]

Any tips on how to be more competent with knowing how much DOF you have in a scene? I'm shooting a bit more landscapes with the 5x7 and so far have been just going by feel ("I think this needs F22 and I think the focus should be here"). I'd like to be more sure of myself and be able to maximize my DOF when I need it.

Any thoughts or techniques for being more scientific about it?

[Bill_4606]

There are charts and tables that will tell you what "should" be in focus based upon the lens focal length, aperture, bellows extension, subject distance and the phase of the moon. But the best way for you determine what is in focus is to just look at your ground glass.
Think about what you want to have in focus then look for it in your image. Learn your camera's movements so you don't have to stop down to get all of your subject in focus - f:64 is not always your friend.

You mentioned that you are shooting more landscapes. Think about it... Aren't big scenes mostly at infinity? If you want to have the mountain and the grass at your feet in focus, force yourself to shoot at 5.6 and use your tilt. If you want the fence at the side to be in focus from foreground to dstant, use your swing at full open aperture.

Finally, don't listen to me... Listen to your own eyes and have fun... ;)
Bill

[E. von Hoegh]

Any tips on how to be more competent with knowing how much DOF you have in a scene? I'm shooting a bit more landscapes with the 5x7 and so far have been just going by feel ("I think this needs F22 and I think the focus should be here"). I'd like to be more sure of myself and be able to maximize my DOF when I need it.

Any thoughts or techniques for being more scientific about it?

Set the lens to the working aperture and look at the groundglass. Not enough DOF? Use a smaller aperture. A proper (meaning dark) darkcloth helps quite a bit when looking at the GG at f 45.

[Frank Petronio]

You could shoot something wide open, half way stopped down, and all the way stopped down and see which you like best. Why not "bracket" the aperture next shoot?

[Bruce Watson]

Any tips on how to be more competent with knowing how much DOF you have in a scene?

Yup: Practice, practice, practice. It's how we learn.

To do it right, you'll have to make prints. You need to see your own personal end-to-end workflow and how your capture choices effect the final print. With that kind of feedback, you'll learn pretty quickly I think.

[E. von Hoegh]

Yup: Practice, practice, practice. It's how we learn.

To do it right, you'll have to make prints. You need to see your own personal end-to-end workflow and how your capture choices effect the final print. With that kind of feedback, you'll learn pretty quickly I think.

What Bruce said.

Keeping a log of every exposure is perhaps the single best learning aid you could use.

[sully75]

Thanks...this was the (frustrating) picture.

https://lh6.googleusercontent.com/-useW0A6YQQ4/UDj1GrbfC8I/AAAAAAAABRo/SfYFC0qk6-E/s800/img468%2520-%2520Chromega.jpg

I was hoping to get almost everything in the picture sharp. But after reviewing this:
http://www.dofmaster.com/dofjs.html

I didn't totally get how narrow DOF on a 240mm lens is. (I didn't write down the exposure, lesson learned) but I think I did F22, focusing on the first rock on the right side that is not in shadow. But the calculator shows only 5 feet of DOF for F22 @ 10 ft.

I can't say I've had great luck figuring out what's in focus via the groundglass once the lens is stopped down.

Lesson learned. Haven't figured out a good record keeping system for my way of working but what I'm doing is obviously not cutting it.

Thanks for the thoughts.

[Leonard Evens]

I use the near/far approach.

I select a point marking the furthest distance I want in focus, focus on that (wide open), and note the position on the rail. Then I select a point marking the nearest point I want in focus and also note the position on the rail. I then measure the distance, called the focus spread, on the rail between those positions. I focus by placing the standard halfway between the near point and the far point. Finally, I choose the f-stop to use when taking the position using the focus spread.

There are a variety of ways to do that. The method I generally use is to multiply the focus spread by ten and divide the result by two. This method is based on optical thoery and assume a maximum circle of confusion of 0.1 mm. It gives a reasonable estimate of the needed f-stop, but it ignores diffraction. I often stop down an extra half stop or full stop to be sure. But if you want to choose a stop which balances defocus and diffraction, you can use the method described at
http://www.largeformatphotography.info/fstop.html.
That website also explains the rationale for different methods.

[sully75]

Leonard,

That page is helpful but I was under the impression that most lenses are made so that if you pick a point of focus, your DOF will be 1/3 in front of that point and 2/3 behind it (supported by that DOF calculator site). So if you are splitting your focus halfway between the two points, seems to me that you are going to have to use a smaller aperture to get your foreground in focus.

[Bill_4606]

I really like like the scene you selected....

For that image I would have focused wide open on the tree on the right center side of the path.
Use front tilt to get the perspective that I liked.
Double checked the focus on that tree.
Use the rear tilt at wide open aperture looking for the top and bottom to focus.
Then stop down to my len's sharpest aperture ( not it's smallest). With the film plane aligned for the scene, you effectively extend your DOF out into the scene - not straight up and down the film plane.

I'm not sure you would have ever gotten the down river tree bank sharp due to haze or water vapor but for this image you might not want it sharp.
Again, I like what you have going in that image. The lighting is focusing the viewer's attention to the path. Not the sharpness.
Bill

[Vaughn]

F64 and be there...:) and f90 on some of my lenses.

I contact print alt processes, so max sharpness (F22 vs f90) is not a big issue for me.

But basically, I place focus about 1/3 of the way into the scene, play with swings and tilts if I need to, then looking at the far and near points on the GG, I start closing the lens down. As I close reduce the aperture, I look for both the near and far points to come into focus at the same time. If the near points come into focus before the far points, then I know I need to set the focus a little farther out. And if the far points come into focus before the near points, then I need to place the focus closer to the camera.

If I get everything looking good on the GG at a certain aperture, I usually close down one more stop to account for operator error, bad eyesight and for good luck.

Vaughn

[Frank Petronio]

It's a fine photo as it is... if I were shooting it I would leave the camera straight except for rise/fall as needed and focus on whatever is most important - that tree trunk in the foreground needs to be sharp above all else, IMHO. So I might focus a couple of feet beyond it, knowing that DOF would carry me the rest of the way and get at least that sharp. Then I would ponder whether f/32 or 45 is enough to carry the DOF of back far enough... I'd look through the camera with the lens stopped down/louped and see if it was in the ballpark, knowing I probably want an extra stop to carry.

The classic Ansel desert/mountain scenes with a horizon line are easier to do a front tilt and get the foreground to infinity in focus at f/16-22. That doesn't apply once you have trees going top to bottom like this. You might get away with a little tilt figuring that DOF might hold the tree but it's iffy unless you really can check it.

I wouldn't hesitate to work at f/32 with a 240, even f/45 will be fine.

If you have Type 55 Polaroid you can pull the neg all goopy and loupe it to see for sure but that's impractical nowadays.

[Greg Miller]

Thanks...this was the (frustrating) picture.

https://lh6.googleusercontent.com/-useW0A6YQQ4/UDj1GrbfC8I/AAAAAAAABRo/SfYFC0qk6-E/s800/img468%2520-%2520Chromega.jpg

I was hoping to get almost everything in the picture sharp. But after reviewing this:
http://www.dofmaster.com/dofjs.html

I didn't totally get how narrow DOF on a 240mm lens is. (I didn't write down the exposure, lesson learned) but I think I did F22, focusing on the first rock on the right side that is not in shadow. But the calculator shows only 5 feet of DOF for F22 @ 10 ft.

I can't say I've had great luck figuring out what's in focus via the groundglass once the lens is stopped down.

Lesson learned. Haven't figured out a good record keeping system for my way of working but what I'm doing is obviously not cutting it.

Thanks for the thoughts.

In this scene, I would not feel the need to get the distant path in focus; the brighter foreground is where the eye will gravitate, so that needs to be in focus. The distant shaded portion of the path is, imho, better left a bit soft.

I also would not use tilt on this image; the near trees on the left would have their upper portions go oof which I think would be unnatural and distracting.

For this scene I would start with the lens' sharpest aperture (f22?). Then I would check on the ground glass to see if everything that I want to be sharp is in focus. If so, then great, we're done. If not, then start stopping down and checking on the GG until everything that needs to be sharp is in fact sharp. Conversely, starting with the lens' sharpest aperture, if something is sharp that I want OOF, then I would start opening up.

But always best to start with the lens' sharpest aperture, then open up or stop down if necessary to get the DOF that you desire (or better, use tilt, swing, or shift if they can be used for the given scene).

[Kirk Gittings]

Short of shooting allot of Polaroid and writing everything down-make every mistake in the book but actually pay attention and learn from it

[dave_whatever]

Any tips on how to be more competent with knowing how much DOF you have in a scene? I'm shooting a bit more landscapes with the 5x7 and so far have been just going by feel ("I think this needs F22 and I think the focus should be here"). I'd like to be more sure of myself and be able to maximize my DOF when I need it.

Any thoughts or techniques for being more scientific about it?

I started off just guessing an aperture like you're doing but found fairly quickly that I would find myself in a situation where I'd travelled hour for a shot, got there with great light, got the camera set up, spent time getting the composition spot on, got the tilt set up, metered the light carefully, and after doing all that then just had a guess at the aperture and later find out it was wrong and the shot was ruined by out of focus areas. I realised it was stupid to work in a precise methodical manner, and then blow it all by using guesswork for something so critical.

You can try stopping down to F/22 or F/32 and seeing if things are in focus, but if you're working in low flat light at dusk or dawn then good luck to you.

So based on this I would strongly recommend using a DOF scale on the focus knob on the camera. I would not recommend his website for most things, but this is one subject he's absolutely nailed - down near the bottom of the page he details rigging up a DOF scale for view cameras:
http://www.kenrockwell.com/tech/focus.htm

Honestly this will always give you sharp shots with no guesswork, and avoids the pitfalls of stopping down too far - i.e. losing shots to wind movement etc.

[Jim Jones]

Leonard,

That page is helpful but I was under the impression that most lenses are made so that if you pick a point of focus, your DOF will be 1/3 in front of that point and 2/3 behind it (supported by that DOF calculator site). So if you are splitting your focus halfway between the two points, seems to me that you are going to have to use a smaller aperture to get your foreground in focus.

Leonard is right. He is setting the focus halfway between the near and far focus points on the rail, not halfway between the near and far selected distances from the lens.

Many years ago (before the age of affordable hand-held calculators) I spent a few hours with pen, paper, and the DOF and hyperfocal formulae in Lenses in Photography by Rudolf Kingslake. Only then did I understand both topics, and learn to mistrust simple DOF charts and even the DOF scales on some lenses. It was well worth the time.

[Heroique]

Sometimes my negatives or prints reveal DOF mistakes that I didn’t catch on my GG.

But I take enough field notes to figure it out.

I learn a lot about the field back at home!

[IanB]

By far the best way of overcoming this issue is to make a depth of field indicator for the focusing knob of your camera as in this article:
http://www.largeformatphotography.info/dofknob/
It seems complicated, but is actually well worth it, as it gives you full control over DoF in an equivalent way to DoF scales on conventional helical lens mounts.
This is one I made about 15 years ago for my Technikardan out of some bits of Perspex:

795247952579526

In essence it is just a marked ring which is sprung lightly behind so it turns normally with the focusing knob, but can be moved separately to adjust. There is also an indicator which I shaped to work as a magnifying glass. The f-stop marks were transferred onto the ring from the DoF tables supplied with the camera using a dial gauge accurate to 0.1mm.

In use, you focus the scene on the furthest point you want to be in focus, then adjust the DoF indicator to zero. Refocus to the nearest point you want in focus - the indicator turns with the knob, and shows you what aperture you need to achieve that overall DoF. Set this on the lens. Then - and this is the key point - you turn the knob and indicator back to the indication which is HALF the overall exposure f-stop and lock off, as this will be the absolute correct hyperfocal distance. i.e: if you need f22 for the exposure, turn back to f11, etc. You are now correctly focussed for the exact scene you want to take, and set at the correct aperture.

If you make a double scale as shown in the article you can work near-far as well as far-near as I do it.

This works with all lenses, and even applies when you are using Scheimpflug adjustments - just at an angle, and flared out, but you can see that on the GG.

If you want to get heavy, you can actually make one of these for any circle of confusion size (hence format) you want if necessary - the equations are available in good texbooks such as Stroebel etc.

It seems complicated, but is actually very simple once you get the hang of it. This is now so important a part of my way of working that I get really irritated at having to use any camera without it. Investing the time in making this is probably the best investment I have ever made in view camera use generally.

Alternatively, though, you could buy a Sinar P2. They have these as standard...

I hope that's useful.

[Drew Bedo]

Any tips on how to be more competent with knowing how much DOF you have in a scene? I'm shooting a bit more landscapes with the 5x7 and so far have been just going by feel ("I think this needs F22 and I think the focus should be here"). I'd like to be more sure of myself and be able to maximize my DOF when I need it.

Any thoughts or techniques for being more scientific about it?

Well, first of all: a wise man once said, "The ground glass is truth". If its not the way you want on the GG it won't be the way you want it on the film.

Second, make a lot of exposures. No ammount of thinking about photography will beat informed and critical effort.

[E. von Hoegh]

Don't forget your notebook.

[Doremus Scudder]

... if you want to choose a stop which balances defocus and diffraction, you can use the method described at
http://www.largeformatphotography.info/fstop.html. ...

The best method I've found so far. +1

Doremus

[David Lobato]

My approach in a tough situation like this is to use a moderately sharp aperture, then get the foreground and middle ground in very sharp focus, and give up sharpest (while still keeping fair) focus on very distant objects. Sharp foreground and middle distance objects are more pleasing to the eye and atmospheric haze slightly softens distant objects anyway. The composition will retain a natural look with this in mind. Try it.

[Mark Barendt]

WYSIWYG, the ground glass DOF = film DOF when stopped down.

A really nice dark cloth and the patience to stay under it long enough for your eyes to adjust to a stopped down image helps.

[Brian Ellis]

Two points that I didn't see mentioned in quickly skimming this thread.

You need to remember that tables showing hyperfocal distances and others showing the aperture to use after setting the lens half-way between the near and far points of focus all assume a certain enlargement factor (print size), often 8x10 from a 4x5 and smaller negative. If you plan to make a larger print you need to stop down more than those tables typically show.

Looking at the ground glass to determine what appears sharp and what doesn't certainly isn't a bad idea but it's sometimes impractical because it has to be done at the taking aperture and that's often so small and/or the light is so dim that the ground glass isn't bright enough to tell exactly what looks sharp and what doesn't.

FWIW, I also use the method outlined in Tuan's article cited by Doremus in his message. It's quick, easy, and accurate. With the price of film being what it is these days there's no reason to guess or try to learn by experience over a long period of time.

[Jim Noel]

Practice, practice, practice

[E. von Hoegh]

WYSIWYG, the ground glass DOF = film DOF when stopped down.

A really nice dark cloth and the patience to stay under it long enough for your eyes to adjust to a stopped down image helps.

Yes. Those who use a black t-shirt as a darkcloth and a pair of reading glasses instead of a good loupe will never understand this, though. Getting rid of the Fresnel helps too.

[Lenny Eiger]

Thanks...this was the (frustrating) picture.

I didn't totally get how narrow DOF on a 240mm lens is. (I didn't write down the exposure, lesson learned) but I think I did F22, focusing on the first rock on the right side that is not in shadow. But the calculator shows only 5 feet of DOF for F22 @ 10 ft.


Have you actually tested what things look like in a print at 32, 45 and 64 ? i would go back to the scene, or find a similar one and do the comparison. I did, and my lens was just as good at 45 as it was at 22. (Sironar S).

If people want depth of field fall-off, that's fine, but if you don't, then you can just close down. In another thread on this very subject many people commented that the diffraction effect was minimal, and shouldn't be worried about. I would say test the limits of the lens. You might be surprised.

Lenny

[pdmoylan]

A cautionary comment and another suggestion. Contrary to some here, I find that landscapes require much rigor in movements to maximize DOF because I frequently choose scenes which combine a range of focus from close in to infinity. It is rare that my framing of a scene is entirely at infinity (this may be more likely with a 58MM lens than say a 150mm or longer lens).

Caution: There are many situations when you have a range of magnification within the scene where 100% DOF is not possible.

Case in point: there are images taken by Jack Dykinga in which he had focused on a dozen or fewer Columbine i(Colorado scene I beileve) in the immediate foreground and a mountain in the distance. It looked to me as if he was using a lens between 65mm and 90mm, due to the near far ratio. Though the flowers and mountain where in focus, you could see that areas near the ground were not. THis is an extreme situation and unless you shoot very near far subjects often in the same image with a large amount of magnification on the closest subject, obtaining complete DOF if mosty achievable.

Though not frequently recommended, to see if you have everything in focus at a particular aperture, simply stop down the lens to the desired f-stop after closing the diaphram and view the ground glass with a loop. I am able to see relatively well to about f22 but it depends on the amount of light and whether my bellows brightness is clipped by the use of movements. This will confirm that you have used hyperfocal technique and movements correctly.
It's like using a preview button on DSLRs.

Good luck.

[john borrelli]

I think the photo Sully75 has provided is excellent. I have tried to photograph similar scenes and the light here is just right. His comments about wanting more DOF reminded me of an experience I had many years ago.
I was shooting a 35mm camera. I took a photo that I liked, when I saw the slide I was disappointed that there was some detail that I felt I could have done better. I went back to the same place and by some luck, I took the same image with the "issue" corrected and everything else was remarkably similar; the ironic thing I discovered was that I liked the original "flawed" image better. Go figure!
A good image isn't always about doing things in the most technically "correct" way. I'm not advocating being sloppy but sometimes the best images have a quality that may transcend the reservations of the photographer.

[Cor]

My few Euro cents here (I we do not have them any more overhere..;-)..).

A talented Large Format shooter and platinum printer Carl Weese thought me long ago: the less movements you use the better, when you start out in LF you tend to over use the movements (at least I did, with noticeable ugly effects).

In the scene presented I would not use tilts to avoid tree trunk getting blurry at the top which I find ugly and distracting. Better focus (pun intended) on good sharpness from foreground to middle and sacrifice the back, it suits the image anyway.

All good tips have been mentioned already, but now I'am nit picking: if I would print the negative I would crop out the (featureless?) black trunk at the very left. I find it distracting from the composition.

Best,

Cor

[rdenney]

A few points, reviewing this thread:

1. DOFMaster assumes a print size of 8x10 inches. You should look at their information screen to see what the circle of confusion assumption is for the format you are using, and then divide that by the ratio of your desired print height to 8.

2. Movements are not used to increase depth of field. They are used to manipulate the angle of the plane of sharp focus. Where that plane is close to the camera, the depth of field will still be as limited as it would be if you focused on a subject that closely. I find I have to stop down dramatically in some cases even when I use extensive movements, just to tighten up apparent sharpness in the near field. In those cases, the movements made the image possible at a small aperture, where they would have otherwise not been possible at all without a different visualization.

3. Given that the focus plane is a plane (at least nominally--but pretty closely for most LF lenses), angling it works when the things you want in focus fall in a plane. Sometimes you can find a plane through a three-dimensional array of subject material that lets you bring to apparent sharpness what you want to bring to apparent sharpness. Sometimes not. The more three-dimensional the subject material, the more time you can spend chasing possibilities, and the more time it requires, the more likely you'd have done just as well without the movements. If you spend a while tangling yourself in knots trying to adjust that focus plane, start over by zeroing the movements and give yourself the chance to take a different tack.

4. The fundamental technique of view camera photography is deciding what needs to be sharp and then finding a strategy for making it so. Any rule of thumb that says "you MUST use movements" or "you should NEVER use movements"--even when not worded as obviously as I've done here--will be wrong in some circumstances. The movements are there for a reason. They solve problems, but sometimes they are used to solve what is not a problem.

5. Generally, diffraction will be less of a problem than being out of focus. Sometimes you have to pick your poison. I'd rather have a good photo that looks right but is limited to 16x20 than a photo where the focus plane can be printed wall-size but much of what I need to be sharp is obviously out of focus. I think of diffraction as a fine effect, while managing depth of field is a coarse effect. Solve the big problems before worrying about the small ones.

6. Whoever said the truth is on the ground glass has it right. If you need an aperture too small to see through, put a light in the scene you want to check. I like those powerful green laser pointers, but I've also used plain flashlights sitting in the scene where I need focus--I can see it even at f/45 in a dimly lit room. DOFMaster provides a useful tool, but does not relieve you of knowing what you want to be sharp or what you have to do to make it so. And you have to understand its assumptions, including whether the lens really has the flat field such tools assume.

7. Sometimes the reality of LF forces you to consider a slightly different composition. Everything is a compromise. I made a photo of a water fountain on the Seattle waterfront using a small digital camera. That picture would have been impossible with large format, if I want to freeze the backlit droplets as I did. I needed a very high shutter speed to do so, which means a wider aperture. The small format meant a short lens which provided the required depth of field. But the compromise is that the small format can only be enlarged a certain amount. LF could be enlarged far more, but the combination of high shutter speed and large aperture would have provided grossly insufficient depth of field. So, like most other LF photographers, I'd have had to give up the frozen, glittering droplets in favor of letting the motion smooth out over a longer shutter time. Different set of compromises and different set of possible visualizations.

Rick "not all visualizations are possible" Denney

[Mark Barendt]

Hey Rick do you know if it would be safe to assume that the circle of confusion is directly related to the change in print area.

So if we move from 8x10 to 16x20, 4x on the area, would the COC be 1/4 of what the calculator suggests?

hope I said that right.

[David Lobato]

Rick,

That's an excellent summation of how to manage DOF with large format cameras. I have done all of them except number 6. Will have to remember that one.

[dave_whatever]

5. Generally, diffraction will be less of a problem than being out of focus. Sometimes you have to pick your poison. I'd rather have a good photo that looks right but is limited to 16x20 than a photo where the focus plane can be printed wall-size but much of what I need to be sharp is obviously out of focus.

+1

[Sylvester Graham]

By running up and down the steps to the Philadelphia museum with a tripod on my back in a jumpsuit.

[rdenney]

Hey Rick do you know if it would be safe to assume that the circle of confusion is directly related to the change in print area.

So if we move from 8x10 to 16x20, 4x on the area, would the COC be 1/4 of what the calculator suggests?

hope I said that right.

No, it's based on the linear dimension. So, 16x20 would need a circle of confusion on the film half the size of 8x10 to appear similarly sharp.

You can always work back from the typical person's ability to see, with the assumption that a print needs to display 5-8 line-pairs/mm with reasonable MTF. A 16x20 is a 4x enlargement of 4x5 (again, based on the linear dimension) and therefore for things to appear sharp on the print they need to be resolved to 20-32 line-pairs/mm with reasonable MTF on the negative. I've therefore always used a circle of confusion standard of 0.025mm for 4x5 (half the 20 lp/m pair = 1/40 mm). DOFMaster uses 0.1mm for 4x5, but I may be being too conservative assuming the circle of confusion is only one half of the line pair at the desired resolution. I don't think they make that assumption and size it to the whole pair sized to an 8x10 print. But the circle of confusion is a standard one sets for himself depending on how sharp he wants things to appear.

Rick "who often looks at DOFMaster, realizes it's unattainable, stops down to f/45, checks the ground glass, and then just prints it no bigger than it looks good" Denney