How accurately can you level your camera?

[Leonard Evens]

I've been thinking about different sources of error when using a view camera.

The first question that came to mind is how accurately one can level the camera and make sure the standards are as close to parallel as possible. Presumably, one can never get it perfectly right, but it seems reasonable that small errors won't make any significant difference in the angle the subject plane makes with the vertical. But some rough calculations seem to indicate that is not quite true.

I use levels for checking parallelism, so I did some measurement with my carpenter's level., which is about 600 mm long. The amount one end can rise above the other with the bubble still between the two lines seems to be about 2-4 mm. That would mean an angle between 1/5 to 1/3 degree. Say the angle between the two standards is 1/4 degree. With a 150 mm lens, that would mean the hinge distance would be about 34 meters. That seems so large it might be considered to effectively infinite. But suppose you focused at 10 meters, which is pretty close to the hyperfocal distance for such a lens at f/22. The angle by which the subject plane would depart from the vertical would be about 16 degrees, which seems pretty far from its being perfectly vertical.

I'm not sure just what that means. Presumably, in terms of what would be caught in the frame in focus, you couldn't really distinguish a 16 degree departure from the vertical from a zero degree departure. But it still seems paradoxical.

Any comments?

Also, how accurately do you think you can level your camera and how close to parallel can you make the standards?

[Jay Wolfe]

As Einstein said, "It's all realtive." In other words, if the image looks right on the ground glass, does it really matter? Of course, I primarily photograph landscape. If I were working on images that require precision, the response might be different. In addition, I'm interested in images, not mechanics.

[Andrew O'Neill]

In other words, if the image looks right on the ground glass, does it really matter?

Precisely!

[Vaughn]

I will agree with Jay...what is important is how everything looks on the ground glass.

I worked with a Gowland PocketView for many years -- no indents at all. And I did not miss them. I just set the camera up close to "perfect" and swung and tilt depending on the needs of the image. Even my Zone VI 8x10 has little in the way of indents. Since I rarely photograph a vertical plane perfectly perpendicular to the camera, I usually do a bit of tilt and swing anyway.

I even base my levelness to the horizon on the scene...not "reality" (especially since most my photographs don'e even have a visibale horizon line, let alone sky.)

So I do not see parallelism of the standards as a possible source of error, but as a means of image management.

Vaughn

[cjbroadbent]

Getting level and square is a headache (a frequent cause of re-shoots).
The system I worked out for table-top is this.
To get square with the set: Glue a pocket mirror to a carpenters set-square, and center it on the front edge of the set. Tack a tiny flashlight below the lens, get the flashlight reflection smack in the middle of the groundglass.
To get level: place two tiny flashlights (torches to in GB english) at the ends of the horizon edge and square the ground glass on the points of light.

The system for interiors is this:
Use push-pins and taught string across the front edge of the set. Use the string to drop a perpendicular centered fore and aft. Align the center vertical ground-glass line on this perpendicular (if you can't see the string, use two flashlights).
These are just variations on the age-old art-reproduction technique of hanging a pocket mirror in the center of the art-work and making sure you can see the lens in the reflection.

As to squaring the front and rear of the camera: The only cameras I trust are the non-folding Ebony and the the Linhof Technica 5x7, which is 100% square. Sinars were always somehow skewed, despite all the micrometrics. The Toyo View was best for 8x10. But there's no way to square-off a Gandolfi (maybe there's no need in a field camera - the brothers knew best).

[David A. Goldfarb]

For floppy cameras like the Gowland PocketView or my old wooden cameras, I use a Suunto Tandem clinometer-compass to check that the camera is level and the standards are parallel along both axes. It claims accuracy to 1/4 degree, and is commonly used for installing satellite dishes (you can often find them second-hand on eBay from satellite dish installers).

If I'm photographing something flat that isn't too far away, I can also use the Suunto Tandem to make sure the camera is square to the subject. If it is something far away, I make sure it's square using the grid lines on the groundglass.

It can also be used to find the angle of the plane of focus for use with a device like the Rodenstock calculator, but I don't usually do that.

Another handy thing, is that it can measure the tilt and swing angles of either standard, so it can be used like the scales on a Sinar or other modern monorail camera. For instance, if I have too much bellows extension to reach the front standard easily, I can find the tilt angle on the rear standard, and then transfer the tilt to the front standard by reading the angle from the rear standard, setting the corresponding angle on the front standard, and then setting the rear standard back to plumb.

The one thing to be careful about, is that if you have a metal camera or camera with metal parts that are magnetic, the compass won't be accurate for reading swing angles.

[evan clarke]

You can use precise equipment. I use an Arca mounted on a C1 Cube and leveling is virtually automatic.....EC

[Robert A. Zeichner]

Also, how accurately do you think you can level your camera and how close to parallel can you make the standards?

One thing I do is to eyeball the front and rear standards from an angle that permits me to visually abut one to the other. It's not a perfect system for measurement certainly, but it does allow me to make them pretty parallel without spending a tremendous amount of time doing so. I also have made my own GG and scribed same with a grid that helps in leveling. Adjusting yaw, I think is hopeless beyond a certain point because the film holder and the film inside it are rarely perfectly parallel with the camera.

[cjbroadbent]

Evan, my Cube just tells me the camera base is level, whis is a good start, but it won't tell me about the back and the front and whether it's pointing square to the room/table/building. You can fix verticals in Photoshop but it's a lot harder to straighten the floor boards.
Though I will admit that Arca joints are more precise than Sinar's.

[Walter Calahan]

"sources of error"

There are no errors, simply sources of opportunity. Grin.

Most buildings and rooms aren't square, so why do we expect our photography to be 'perfect?'

What's more important is that the resulting image is what you envision before setting up the camera.

Use a simple carpenter's angle finder.

[Frank Petronio]

With many cameras you can adjust the standards to be parallel either by design or force. If it pains you to have them off why not fix them?

I think what throws a lot of them off is too much bellows compressed and pushing the standards apart.

[Michael Graves]

I was at the car dealership getting an oil change and found a bubble level on a key chain in their accessories department for five bucks. It lets me check level on all dimensions quite easily.

[ic-racer]

Seem like you ask 2 different questions. Leveling the camera and making the standards parallel.

Making the standards parallel will be near impossible without detents. The detents need to be set in a manner similar to collimating a lens (as the lens is the important thing here, not the front standard). Yes, it will make a difference if it is off, especially with short lenses!

You may need a camera tech to do it for you, but I get good results bouncing an enlarger alignment laser off the front lens element and lining up the reflected diffraction pattern with the center of the laser. Then I take the lens off and bounce the laser off a optical glass plate at the film plane. When the detents are correctly set, the laser will bounce right back on itself.

Then check with your other mounted lenses and see that they are also correct. Since you don't want to re-set the detents when you change lenses and lensboards, I sometimes shim a corner of the lensboard or just set it up for the widest lens and let the longer lenses be a little off.

[tgtaylor]

What about using the bubble levels that are on each standard? My Toyo-Views have horizontal and vertical bubble levels on each standard that seem to be pretty accurate.

[Bruce Watson]

I designed a set of bubble levels for the rear standard of my Toho so that I could level and plumb the film plane. SKGrimes did the work to make and install the levels. They seem to be quite accurate -- if the bubble is centered the horizon is exactly parallel to the lines on my ground glass.

I could have done the same thing to the front standard, but couldn't justify it. The rail of the Toho is pretty stiff so if the rear standard is level then so is the front. And I use some amount of front tilt in just about every photograph I make, so having it plumb or even just parallel to the rear standard is of dubious value to me. What is of value is seeing the ground glass and insuring that what I want in focus is in fact in focus (as we all know, easier said then done ;-)

The reason I want the ability to easily level and plumb the film plane is that I like my trees to be straight (or at least not distorted by the camera). Not strictly necessary, and can often be done by eye within an acceptable tolerance. But I wanted to be able to do it quickly and to take most of the judgment out of the setup. For that it has succeeded admirably.

[Brian Ellis]

Getting things perfectly level is a concern for me only with architecture (especially windows and doors) and very occasionally in landscape photography (horizon lines, e.g.). I use a Bogen 410 geared head. I place a small level on the top and on one side of the camera. That allows me to level the camera from side to side and front to back. I get the camera square to the subject by trial and error, looking at the ground glass. When I open the image in Photoshop I view it with Rulers. If there are any mistakes they're usually no more than a degree or two and I fix those in Photoshop.

[Frank Petronio]

yeah your film probably shifts in the holder enough to be off a little too...

[BradS]

Wow. I think such attention to tiny details would take all the fun out of it for me. For me, close enough is perfect. If it looks good on the GG - it is good enough. But, I am fortunate in that I do not depend on any revenue from Photography.

[Nathan Potter]

When doing general photography with a Technikardan 4X5 I don't worry about front and backplane parallelism at all - can't be bothered and don't really need to. But in my days of industrial photography of planar parts (IC wafers, hybrid substrates, etc.) where high resolution was required I used an extra long vernier calipers to adjust the standards to parallelism. Tricky business due to backlash motion in the 4X5 and 8X10 clamps. But with care and sweat I could get parallelism within about 0.001 inch (25 micrometers). Orthogonality of the lens axis to the film plane was generally not a problem with a high quality process lens. Accuracy in the location of the film plane relative to the focus on the GG always was a bigger variable so I eventually used a specially designed aluminum film holder with the film plate surface ground. Shooting was done with lens close to wide open and electronic flash in various lighting configurations.

Nate Potter, Austin TX.

[Scott Knowles]

Interesting responses. I level the tripd (built-in bubble) and camera on the ballhead (two-way bubble in hotshoe). I have a prism effect with my eyes and all too often when shooting with the digital camera find handheld shots slightly off perpendicular or level (even constantly reminding myself to line up the vertical or horizontal). With the Horseman I know it's starting from level.

[Once]

I start to se a pattern. First it was almost unbearable to find the distance 1500 x the focal length or behind (to focus at infinity). Now it is difficult to even start taking pics because how the hell the standards can be perfectly parallel down to 1/x degrees... Hmm. Do you ever take pictures or you just sit down and calculate degrees of error in this and that part of your equipment? Hundreds of guys out there take good pics while you argue that finding something 225 yards or more away is not precise to focus at infinity...

[Bruce Watson]

Personal attacks aren't welcome here. Everyone works differently, everyone is entitled to express their curiosity and creativity as they see fit.

The man asked a question. It's OK for you to not have anything to contribute, or to choose not to contribute. What's not OK is to belittle someone for asking the question in the first place.

[Once]

Well Bruce, I express my opinion. Everyone works differently, you know...

[Frank Petronio]

How is pointing out something that is common sense and a matter of proportion "personal"?

[Andy Eads]

A poor man's way of assuring parallelism is to take a length of plastic irrigation pipe and cut it to about your favorite focal length on a table saw with a good fence. You can check that the ends of the pipe are parallel with a simple ruler. Tape the pipe to the lens board then rack it back till it touches the ground glass. Adjust till the pipe touches the glass evenly and everything will be parallel. Remove the pipe and go to town. Unless you have major slop in the lens board or your focus track is not straight, everything will stay parallel.

[Bruce Watson]

Well Bruce, I express my opinion. Everyone works differently, you know...

Well bless your little heart Sweet Pea, we've all got opinions. If you feel a desperate need to express yours, head for the lounge. That's what the lounge is for after all.

The purpose of this space, however, is for people to get answers to their questions. Judging the questions themselves is uncalled for. Judging what you image are people's reasons for asking the questions is unconscionable. You can't know what motivates other people to ask the questions they do. There are nearly as many motivations are there are people asking questions, because we all have different experiences and as a result different world views. Whatever their motivations for asking their questions, the questions themselves are valid. At least as valid as your questions are.

In the face of this, the best you can do is answer their questions in hopes that one day they will answer yours.

[Colin Graham]

Careful with those plastic bubble levels. You'd be surprised how inaccurate they are. Not only does the casting and reference marks have to absolutely on the nut, but also the installation. Even the bulls-eye level on my Ries head isn't very accurate. I found that out when I tried to set up a transit on it once as a test. Gridlines on the GG are still my favorite way of setting up. Siting along standards against known plumb and level points in the composition is a good quick and dirty method too.

[Leonard Evens]

Thanks for all those who actually answered my questions rather than criticizing me for asking the question. I found the answers useful.

Perhaps I should say why I got interested in questions like this in the first place. First, I do a lot of architectural photography, where being square to the scene and level can make a difference. I certainly rely strongly on what I see on the ground glass, but sometimes that doesn't suffice. When I first used my 75 mm lens, I encountered something unexpected. I thought I had everything in focus and even stopped down an extra stop to be sure, but when examining the negative closely I found that it went badly out of focus at the edges. I thought at first that the vendor had sent me a lemon, but I decided to look more closely, and I found that my standards were not parallel,. In effect I had a slight swing, and that was most likely the cause of the fuzziness. I find that when I understand things, it saves me time in the field because I don't waste time or film doing things by trial and error. Because of arthritis, the amount of time I can devote to taking pictures is limited, so I like to be able to get things right.

Let me explain how I make sure the standards of my Toho FC-45X are parallel. I can't rely on the detents since there is quite a lot of play in using them. So, I turn the camera 90 degrees on my tripod, making sure the rail is plumb. I then make sure both standards are level both in all directions. This doesn't depend on how accurate the level is as long as I use it consistently for both standards. But it does depend on how well I can manage to get the buble in the same relative position in both cases. It is my guess that I can get it easily to better than 1/2 degree and if I work it to better than 1/4 degree. So the question is how much could such an error affect the resulting pictures. My experience suggest that the answer is not by enough to matter, but being a curious type, I always want to know just why that should be the case. I am not suggesting that everyone else should find such questions interesting, and I hope others won't begrudge me the right to do it myself. The reason I ask questions is because often sharing a question with others is helpful in finding an answer.

In case anyone is interested, the conclusion I came to was reassuring but also a bit perplexing. As all of us would have predicted, it is what happens in the vicinity of the ground glass that matters. The position of subject plane can vary by quite a lot without producing a significant effect on the image. So although it is almost never true that the exact subject plane is parallel to the film plane, or even particularly close to being parallel, we still get the right answers for depth of field by assuming it is.

[Mattg]

I didn't take offence at your comments Once but it's worth remembering that everyone's photography is different, as well as our ways of working.

I square the camera before using it each and every time, it's transported with all the movement locks relaxed. This takes me all of 60 seconds and it means that I am able to take the shot I want, even when it's dark and I'm struggling to focus. I rely on the spirit levels on an Arca and find that the degree of parallelism achieved is sufficient to avoid nasty suprises with a 65mm lens. As you say Leonard, the short hinge distances experienced with short focal lengths makes the "squareness" of the camera critical.

I do this because I spend an aweful lot of time making pictures and printing them, I don't need to spend more time making the picture I should have made the first time. In the past I haven't bothered and I have occasionally been disappointed. I do what I need to ensure the final image is what I want.

On the other hand I find I'm not always able to perfectly level the camera either horizontally or vertically; despite my careful efforts many scenes require some small correction in enlargement.

[neil poulsen]

I find it difficult, or at least awkward, to tell from a ground glass whether or not the camera is pointing too far up or down. (It's much easier to tell if the camera is level left to right.)

My 3039 (same as 229) Bogen head has two adjustable levels, one in each direction. I mount the camera and adjust them so that, if the camera's level in each direction, then these two levels show the bubbles in the center.

This makes it easy to level the camera. On the outside, it takes about 10 seconds.

[Once]

I've been thinking about different sources of error when using a view camera.

...
I'm not sure just what that means. Presumably, in terms of what would be caught in the frame in focus, you couldn't really distinguish a 16 degree departure from the vertical from a zero degree departure. But it still seems paradoxical.

Any comments?

Also, how accurately do you think you can level your camera and how close to parallel can you make the standards?


...

Let me explain how I make sure the standards of my Toho FC-45X are parallel. I can't rely on the detents since there is quite a lot of play in using them.
...

In case anyone is interested, the conclusion I came to was reassuring but also a bit perplexing. As all of us would have predicted, it is what happens in the vicinity of the ground glass that matters. The position of subject plane can vary by quite a lot without producing a significant effect on the image. So although it is almost never true that the exact subject plane is parallel to the film plane, or even particularly close to being parallel, we still get the right answers for depth of field by assuming it is.


Leonard,
the common sens is as important as honed mathematical skills. Every architecture photographer worth that name knows that short focal lengths have small depth of focus at the film plane. You know it too. Therefore they need a camera capable of good parallelism of its standards. A camera with "a lot of play" in 0 detents is not good for that. A modern Arca Swiss is, to give an example.
Trying to take pictures with very short focal lengths on a camera with crappy 0 detentes is of course more difficult than to do so with a camera with appropriate capabilities.
If you said in the first place that you use a 75mm lens on a camera with "a lot of play" in 0 detents and you sometimes get pictures that are not sharp you would get answers telling you what "it means" right away. Just my opinion.

[russyoung]

Methinks it bears pointing out that the needs for accurate leveling depends both on the subject and the photographer's artistic vision.

A landscape photographer who works at infinity and closes down to f/64 has far less stringent requirements than someone who copies artwork (where the camera must be perfectly square to the subject). Likewise, those who use shallow depth of field as a creative tool in portraiture have fairly stringent requirements and cannot afford to be fiddling with alignment while a client fidgets and grows impatient.

Briefly, I check my camera system with a laser level. I keep the level permanently in one position in the studio, well out of the way but in a line straight behind the camera. For the first time, once the laser is leveled, a small mark is made on the studio wall where the laser hits (this will never be in a photograph because it is always going to be behind the subject or object, by definition). For all subsequent use, I turn the laser on, verify that it still hits that target mark (ie, it has not been accidentally disturbed), then move the camera into the beam. The exact center of the ground-glass is marked with a cross-hair and the camera is moved until the beam hits that mark. Having already leveled the camera as closely as possible with a machinist's bubble level (owning one of these is very illuminating when checking other sorts of bubble levels), then I examine where the laser beam strikes the target on the wall. If it is spot on, then the front board must be parallel ...

There are a few fine points left out but you get the idea. This is relatively important for me because I tend to use vintage wooden cameras which have no detents. I have found that attempting to level by using the wooden body surfaces is hopeless. The GG and the lens board are the only two surfaces that matter and must be directly measured if any accuracy is to be obtained.

Apologies for the length of this... but I actually left out a fair amount!

Russ

[poco]

I had edge softness due to non-parallel standards too so I started a simple routine at the beginning of every day's shooting: apply rise until the top of the front standard is about on level with the back standard and then simply rack the bellows back (or forth) until you can squeeze a credit card between them. Adjust your standards so they're both flush with a card edge and you're good to go. Parallelism on the tilt axis is more difficult to figure with my camera, but I generally don't bother since I'm usually using slight tilt anyway.

[Leonard Evens]

Leonard,
the common sens is as important as honed mathematical skills. Every architecture photographer worth that name knows that short focal lengths have small depth of focus at the film plane. You know it too. Therefore they need a camera capable of good parallelism of its standards. A camera with "a lot of play" in 0 detents is not good for that. A modern Arca Swiss is, to give an example.
Trying to take pictures with very short focal lengths on a camera with crappy 0 detentes is of course more difficult than to do so with a camera with appropriate capabilities.
If you said in the first place that you use a 75mm lens on a camera with "a lot of play" in 0 detents and you sometimes get pictures that are not sharp you would get answers telling you what "it means" right away. Just my opinion.

Different people use terms differently, so let me be as precise as possible. By `depth of focus', I mean the total variation in where you place the standard when you try to focus on a target. There is a simple well known formula which says it depends on the f-stop, the power of loupe you use, and the scale of reproduction. It also depends on the quality of the ground glass and how well the viewer can see fine detail. So although it is a personal thing, possibly different for each person, it does vary within a fairly narrow range. It also scales inversely with the power of the loupe. You can check it by estimating the distance on the rail between where you just come into focus and where you just go out of focus. Using a f/4.5 75 mm lens, with the equivalent of a 2 X loupe, my personal depth of focus is about 1 mm, and using a 3.6 X loupe is is a little more than 0.5 mm. Using additional tricks, I can reduce the focusing error below that, but not by too much.

Depth of focus does not depend on focal length. But as Mattg pointed out, the hinge distance is directly proportional to the focal length, so for the same tilt angle, it is smaller. That tends to mean that the subject plane is more inclined to the vertical. (Change the language appropriately for a swing, which was the problem in my case.) So an inadvertent tilt or swing may be more obvious with a shorter focal length lens. You can say that I learned that first by experience, and then subsequently I figured out why by analyzing the geometry. As I noted, experience alone would have taken me much longer, since my initial assumption on the basis of one example, that something was wrong with the lens, was wrong.

I would love to have an Arca Swiss. I got the Toho for two reasons: first it was very light and easy to transport, and second it was within the price range I envisioned at the time. The Arca was more expensive, and while light, didn't seem as easy to transport. Since then, because of the decline of the value of the dollar, its price has increased manifold. Still I might decide to make the plunge and get one, but being retired with fixed resources, it is not an easy decision to make. For the while, I have to make do with what I have.

But note, it is impossible to make a view camera which doesn't have the kind of problems I've described to some degree. For most purposes there are workarounds.. Using levels, I can get around the play in the detents, and similarly for other defects the camera has. It just takes a bit longer. If I do get an Arca Swiss, I am sure it will be a pleasure to use, but it won't allow me to avoid dealing with the kinds of problems I've described.

[Leonard Evens]

I had edge softness due to non-parallel standards too so I started a simple routine at the beginning of every day's shooting: apply rise until the top of the front standard is about on level with the back standard and then simply rack the bellows back (or forth) until you can squeeze a credit card between them. Adjust your standards so they're both flush with a card edge and you're good to go. Parallelism on the tilt axis is more difficult to figure with my camera, but I generally don't bother since I'm usually using slight tilt anyway.

That;s an interesting idea. I'm not sure I can get the standards that close, even with the lens removed, but I might be able to use my dial caliper, which supposedly can measure distances to within 1/100 th of an inch. That together with my method of turning the camera 90 degrees and using levels on the standards should do it. Fortunately, once the standards are set on the Toho, they don't move, so I only have to do it after using a stilt or a swing.

[poco]

I'm not sure I can get the standards that close, even with the lens removed,

Use the credit card lengthwise -- works just as well.

[Once]

Different people use terms differently, so let me be as precise as possible. By `depth of focus', I mean the total variation in where you place the standard when you try to focus on a target.
--
Depth of focus does not depend on focal length.
--

I would love to have an Arca Swiss. I got the Toho for two reasons: first it was very light and easy to transport, and second it was within the price range I envisioned at the time. The Arca was more expensive, and while light, didn't seem as easy to transport.
--.
Yes and no.
Yes, I'm speaking the depth of focus as measured on the image (film) plane. And yes, that depth is depending on the focal length of a lens. There are different formulas out there, one calculating it with the focal length value, among others, another one with the help of the magnification (which depends on the focal length) etc. Short focal lengths have smaller depth of focus for the given format than longer lenses.
And no, the Arca is very easy to transport, especially when you can leave both standards on one of its short rails and the monorail with the other rail you keep separately. To set it up is a piece of cake. Don't let yourself be discouraged by imaging the opposite (well, I have an Arca, so I'm partially partial...)

[ic-racer]

I start to se a pattern. First it was almost unbearable to find the distance 1500 x the focal length or behind (to focus at infinity). Now it is difficult to even start taking pics because how the hell the standards can be perfectly parallel down to 1/x degrees... Hmm. Do you ever take pictures or you just sit down and calculate degrees of error in this and that part of your equipment? Hundreds of guys out there take good pics while you argue that finding something 225 yards or more away is not precise to focus at infinity...

In terms of how parallel the standards need to be, actually it depends on format size. I found 6x9 (as with non-view camera MF cameras) needs a very precision alignment to set the detents. 8x10, though, can be set up just by looking at the ground glass or using bubble levels. 4x5 is somewhere in between.

[Once]

In terms of how parallel the standards need to be, actually it depends on format size. I found 6x9 (as with non-view camera MF cameras) needs a very precision alignment to set the detents. 8x10, though, can be set up just by looking at the ground glass or using bubble levels. 4x5 is somewhere in between.

Right. But because the very short focal lengths are usually with smaller image circle you use them more on the smaller formats, especially in the architecture photography. Cameras for architecture are less forgiving in this aspect. That's why the fixed standards architecture cameras (Cambo, Silvestri etc.) have an edge, at least speaking about this point.

[Ole Tjugen]

Right. But because the very short focal lengths are usually with smaller image circle you use them more on the smaller formats, especially in the architecture photography. Cameras for architecture are less forgiving in this aspect. That's why the fixed standards architecture cameras (Cambo, Silvestri etc.) have an edge, at least speaking about this point.

There's one exeption to that rule, just like any other: The Carbon Infinity. But as I have discovered it is the exception to just about any rule you can think of, and it was designed that way. :)

[Once]

There's one exeption to that rule, just like any other: The Carbon Infinity. But as I have discovered it is the exception to just about any rule you can think of, and it was designed that way. :)

Good enough, with one exception - the weight. ;)

[David_Senesac]

Most of my work is landscapes. My used Wisner Expedition has three bubble levels embedded in the wood frame, but I learned quickly they were not accurate. I'd expect RW accurately calibrated the camera when new but the orthagonality of the wood camera probably did not last long. The whole camera body is rather wobbly and elastic that is part of the issue. So the only real thing I could depend on is what I view on the groundglass. Which is fine because over long years of shooting 35mm SLR and 6x7, I developed a keen sense of camera leveling.

One does however need to understand that the nature of a subject and one's surroundings can trick one's sense of what is vertical and horizontal as anyone that has studied optical illusions would be well aware of. For instance if one is in front of a large lake with a far shore slightly further on the left side versus the right, one's mind's eye may try and trick one into that shoreline edge being horizontal. Generally under the dark cloth, my sense of having my head in a vertical orientation is fairly good, though that is far more prone to error than when I'm viewing a scene outside the camera. Thus I always tweak the vertical graticule lines on the groundglass by removing my head, looking up in the sky at about 135 degree altitude to grasp a sense of the vertical between my two eyes, and then move my vision down towards the 90 degree horizon above the camera, and check that versus those graticule lines. Sometimes that may take a few hunting tweaks of the ballhead to optimize.

A large format friend of mine has a mediocre sense of the horizontal. In the field we often amuse ourselves by looking into each others cameras during or after setting up shots. About half the time I immediately notice his horizontal is well off and chuckle. I have the same sense looking at images in Photoshop on my monitor. Being off even one degree is rather immediately noticeable. Sometimes I will use the rectangular marquee tool to quickly check about elements like trees that ought to be vertical. However off my own crude home flatbed scans, XY errors are almost always due to the position slop in my flatbed 4x5 film guide. I would encourage other photographers who think they have a natural sense of, to try and develop an internal visual sense of horizontal and vertical.

[Richard M. Coda]

I have two Arcas... 4x5 Field and an 8x10 F Metric, and an 11x14" custom back for the 8x10. I have a Gitzo CF tripod with an Arca leveling head (not the cube).

I can get it level in a few seconds, but it really depends on whether I've had anything to drink prior to that! ;)

[Leonard Evens]

Yes and no.
Yes, I'm speaking the depth of focus as measured on the image (film) plane. And yes, that depth is depending on the focal length of a lens. There are different formulas out there, one calculating it with the focal length value, among others, another one with the help of the magnification (which depends on the focal length) etc. Short focal lengths have smaller depth of focus for the given format than longer lenses.
And no, the Arca is very easy to transport, especially when you can leave both standards on one of its short rails and the monorail with the other rail you keep separately. To set it up is a piece of cake. Don't let yourself be discouraged by imaging the opposite (well, I have an Arca, so I'm partially partial...)

From `Applied Phtographic Optics" by Sidney F. Ray, on p. 224, there are two formulas for depth of focus. They are

t = 2CN(1 + m)

and

t = 2CNv/f

where C is the circle of confusion that is applicable when viewing the ground glass, N is the f-number, m is the scale of reproduction (also called the magnification), f is the focal length, and v is the bellows extension.

These formulas are of course the same since v/f = 1 + m.

The second formula seems to make the depth of focus depend on the focal length, and that would indeed be the case if the bellows extension were kept constant while varying the focal length. But the circumstances in which one would do that are rare. For most normal photography, the bellows extension depends strongly on the focal length. If you use the first formula, you don't have to worry about the relationship between the two. For photography of distant subjects the magnification m is very small, and a good estimate for the depth of focus is

t = 2CN

and that certainly doesn't depend on the focal length.

For close-ups, the factor m plays an important role. In most cases in close-ups, if one varies the focal length, one still tries to keep the scale of reproduction constant, e.g.. one tries to fill the frame with the subject. So, in that case also, the depth of focus won't change if you change the focal length since (I should have added) you keep m fixed. One can certainly envision circumstances in which m changes as the focal length changes. In that case, the depth of focus would depend on the focal length. But although i can think of cases where that might happen, e.g., when one in effect converts a distant scene into a close-up by using a longer lens (but keeping the focusing aperture fixed.), but they are few and far between.

There is one other point which might muddy the waters. If you fix the distance of the subject from the lens, then you get greater depth of field with a short focal length lens at the same aperture than with a long focal length lens. That means that focusing critically on one specific point, depending on the circumstances, may not be as important.. It may also be harder to see very fine detail since everything in the plane of focus is smaller, so you may need to look for something larger to focus on if you choose a smaller focal length while keeping the subject distance fixed.

But that is not what depth of focus is concerned with, which is how accurately you can set the exact plane of focus when it is important to do so..

Everyone should do the experiment I suggested. Focusing as you normally do, try to focus on a distant object and determine the distance on the rail between where the image just comes into focus and where it just goes out of focus. (This can be hard to do unless you have an expanded scale on the focusing knob, since typically the depth of focus is 0.5 to 1 mm.) Do this several times for each focal length that you use and take an average. Make sure you use the same f-number for focusing with different lenses for the purposes of the experiment. The number you come up with is your personal depth of focus.

[Once]

---
Depth of focus does not depend on focal length. But as Mattg pointed out, the hinge distance is directly proportional to the focal length, so for the same tilt angle, it is smaller. That tends to mean that the subject plane is more inclined to the vertical. (Change the language appropriately for a swing, which was the problem in my case.) So an inadvertent tilt or swing may be more obvious with a shorter focal length lens. You can say that I learned that first by experience, and then subsequently I figured out why by analyzing the geometry. As I noted, experience alone would have taken me much longer, since my initial assumption on the basis of one example, that something was wrong with the lens, was wrong.
---
.


From `Applied Phtographic Optics" by Sidney F. Ray, on p. 224, there are two formulas for depth of focus. They are

t = 2CN(1 + m)

and

t = 2CNv/f

where C is the circle of confusion that is applicable when viewing the ground glass, N is the f-number, m is the scale of reproduction (also called the magnification), f is the focal length, and v is the bellows extension.

These formulas are of course the same since v/f = 1 + m.

The second formula seems to make the depth of focus depend on the focal length,
---.

I'm glad you start to see than even mathematically the depth of focus depends on the focal length (the two formulas are the same as you correctly see). Not so much for the distant subjects, more in the macro photography. In the macro photography the shorter lenses give less depth of focus.

Of course, a camera with crappy 0 detents is always more prone to create havoc on the film plane with the depth of focus than a camera where the parallelism of standards is more a question of solidity rather than that of wishes.

[ic-racer]

That's why the fixed standards architecture cameras (Cambo, Silvestri etc.) have an edge, at least speaking about this point.

Totally agree!

[Preston]

My camera levels adequately for me since I shoot landscapes, primarily. But most of my LF friends say that I am about a bubble-width out of plumb.

-P

[Leonard Evens]

I'm glad you start to see than even mathematically the depth of focus depends on the focal length (the two formulas are the same as you correctly see). Not so much for the distant subjects, more in the macro photography. In the macro photography the shorter lenses give less depth of focus.

Of course, a camera with crappy 0 detents is always more prone to create havoc on the film plane with the depth of focus than a camera where the parallelism of standards is more a question of solidity rather than that of wishes.

I assume you will respond to this, but I will give it one last try and then leave the last word to you, which I am sure you will avail yourself of

If you read what I said, I made it explicit that usually in close-ups the depth of focus does not depend on the focal length. (Neither by the way, does the depth of field.) In such circumstances, one attempts to fill the frame with the subject, which means the quantity m (which is the ratio of image size to subject size in the plane of focus) remains more or less fixed. Since the first formula doesn't include the focal length in it, if m remains more or less fixed, then so does the depth of focus.

Now you may very well do close-up photography differently than I do. If you do it so the bellows extension stays more or less fixed while you vary the focal length, then it will be true that the depth of focus will change. As you decrease the focal length, the depth of focus will increase. I don't suppose it it will make any difference in your take on the matter, but it might be worth pointing out that you originally claimed that short focal length lenses have less depth of focus. The only circumstance I can see in which that would be true would be if you decreased the bellows extension faster than you decreased the focal length. Perhaps you do that regularly in close ups. I don't believe many people do that, but I can't prove that I am right.

As to my `crappy detents', let me repeat once more that play in the detents doesn't keep me from aligning the standards. I don't rely on them to make sure the standards are parallel and I probably wouldn't even if I got an Arca Swiss. It just takes me a bit more time---I'm talking about 30 seconds to a minute---not a large fraction of the time it takes me to take a picture. I will often wait much longer than that for the wind to die down, for example.

I just looked at the review of the Arca Swiss on this website and at their product information. I have to say that its design gladdens the heart of a gadget nerd like me, but I really doubt that one absolutely needs it to do the kind of photography I want to do. The biggest advantage it has over my Toho is the possiblity of using a wide angle bellows allowing large movements for my wide angle lenses. But the Toho eccentric lensboard more or less solved that problem for me, since it allows me to rise/fall almost as far as the image circle allows with my 75 and 90 mm lenses. If I had $7-8,,000 to throw away---that is what I figure I would need to get the camera and all the extras I would want, plus a Schneider 72 mm Super Angulon XL to take advantage of the wide angle bellows---I would buy one in a flash, even if I ended up using it once a year.

[john borrelli]

Here is my 2 cents worth of personal experience with leveling my camera.

I admit I am not gifted at intuitively judging whether a scene is level.

I stopped using too many bubble levels because it was too hard to get them all to agree! I now use my tripod's level to level the tripod then I use my camera's 3 bubble levels.

When I first started out in LF photography, I bought the old style Bogen leveling head: essentially two overbuilt metal plates ,held together by 3 leveling screws, with a bubble level on the top plate. I screwed this on to my tripod then screwed my cameras tripod block on to the leveling head couldn't point the camera up or down, but the photos were more often level than with my present and more expensive Arca Swiss Ball Head. Go Figure!

I have noticed recently with that ball head that it minutely pulls the camera to one side when I am tightening the head. I don't tighten it as much now, it may need a service.

The thought occurred to me some time ago that there are some landscapes that are "not" level. Or at least do not appear level to the human eye. This may be what Dave S. was describing. I recognize this phenomenon when I have fiddled and fiddled with the scene on the groundglass and it still does not appear level even though my tripod bubble level and Arca Swiss camera levels tell me otherwise.

[Once]

I assume you will respond to this, but I will give it one last try and then leave the last word to you, which I am sure you will avail yourself of
---
?

[Gary L. Quay]

I know how to knock a camera out of level.

Also, I was a the beach a few weeks ago, and I didn't notice that a cap was missing from a leg of my tripod (which has hollow legs). As I made a 5 second exposure, the tripod slowly listed to the left.

--Gary