Scheimpflug in Practice

[bieber]

I'm pretty much brand new to all of this. I've read The Camera, and a bunch of stuff on the Internet about camera movements, so in theory I have some idea how all this works. In practice though, this (https://imgur.com/a/INdjk4O) is the first thing I've really tried shooting with a view camera, as kind of a test, and I'm curious how more experienced folks would approach it.

My goal was to get the entire thing in focus, so conceptually I knew that I needed to swing the lens to the right to shift the plane of focus. My problem was that when it came to actually getting the right tilt to produce the right plane--especially combined with the fact that I needed the camera angled down to get it in view, as rise/fall weren't sufficient from that height--I found myself just kind of floundering around guessing at settings and never quite getting it right. Ultimately I did get the top part of the band a little more in focus than a normal camera would have allowed, but I totally lost the bottom.

How would you approach this more methodically? I've been thinking it might help to try to visualize the intersection of the subject plane and the image plane and aim the lens plane at that, but even then that seems like it would be very imprecise. Do you actually take measurements of the angles involved in this kind of situation, or is there a more intuitive way to do it?

[jp]

My intuitive non mathematical way is that at infinity or long distances, the bellows draw = focal length and bellows will be at their minimum. Things get reversed as they go through the lens. Things low in a scene end up high inside the camera. Sky ends up at the bottom of the camera. Simple empirical pinhole / camera obscura stuff so far. Now add a lens that focuses.

If the top of the lens is tilted down a bit like scheimpflug, there is a greater distance within the camera at the top so the focus distance will be shorter for things at the bottom of the scene. With some front standard tilt you can have top and bottom of the camera focus on different distances because the top and bottom of the lens are differnent distances from the film. Draw a line between those two distances in the scene and you have your plane of focus.

Here's one https://flic.kr/p/DGE5NY I did indoors with my 8x10 and Kodak 305.. I have the foreground hose in focus and the corner of the couch which should have included my daughter's face, but the lens does not have a flat field as it was not made for 8x10 and/or she may have moved a little bit as it was not a kid friendly exposure time.

In practice, the amount of tilt adjustment is much less than sales pictures for view cameras that look contorted. I reversed a front standard on my speed graphic and it's has very modest tilt and that's plenty for outdoor scheimpflug. My wooden B&J 8x10 seems like just a few small degrees of adjustment makes noticeable changes in the plane of focus. No measurement. Just learn to operate the adjustments of the front standard while viewing the ground glass, which is more about camera familiarity and inspector gadget arms than measurement and academics. With a precise SINAR camera you might learn certain lenses at certain distances need a certain adjustment and it might save you time to record angles, but it's not necessary and probably rarely practiced.

[williaty]

I'm pretty much brand new to all of this. I've read The Camera, and a bunch of stuff on the Internet about camera movements, so in theory I have some idea how all this works. In practice though, this (https://imgur.com/a/INdjk4O) is the first thing I've really tried shooting with a view camera, as kind of a test, and I'm curious how more experienced folks would approach it.

My goal was to get the entire thing in focus, so conceptually I knew that I needed to swing the lens to the right to shift the plane of focus. My problem was that when it came to actually getting the right tilt to produce the right plane--especially combined with the fact that I needed the camera angled down to get it in view, as rise/fall weren't sufficient from that height--I found myself just kind of floundering around guessing at settings and never quite getting it right. Ultimately I did get the top part of the band a little more in focus than a normal camera would have allowed, but I totally lost the bottom.

How would you approach this more methodically? I've been thinking it might help to try to visualize the intersection of the subject plane and the image plane and aim the lens plane at that, but even then that seems like it would be very imprecise. Do you actually take measurements of the angles involved in this kind of situation, or is there a more intuitive way to do it?

You can do it with math but (even coming from a physics background) I don't think you'd ever get around to taking the picture.

The plane of focus moves the same way as the lens board but a little more extremely than the lens board does. Tilt the lens board forwards, plane of focus tilts forwards. Swing the lens board to the left, plane of focus swings to the left. Generally, the plane of focus changes angle faster than does the lensboard, so you get a little bit more than you expect.

Of course, on most camera+lens combos, moving the lens also throws the focus out of whack. So you have to refocus, which throws the movements out of whack...

The process I'd use to shoot your watch, with the intent to lay the plane of focus down parallel to the table to get all the watch in focus, is this:

1) Set everything up.
2) Focus on whatever's dead-center in the ground glass.
3) Confirm that the two opposite ends of the watch strap are more unsharp than I'm willing deal with.
4) Think "OK, I need to tip the plane of focus forwards to get it to lay down flat on the table. That means I need to tip the lens board forwards"
5) Tip the lensboard forwards MUCH LESS than I think I'm going to need to. If I think I'm going to need 45* of tilt, I start with maybe 10-15*
6) Everything is now out of focus
7) Focus on whatever's in the dead-center of the ground glass.
8) Confirm that the two opposite ends of the watch strap are more unsharp than I'm willing deal with but that the region of acceptable sharpness in the middle got a little bigger (since the plane of focus is now slightly tipped).
9) Think "OK, that helped, so I'm moving in the right direction but I need a little more"
10) Tip the lensboard forwards MUCH LESS than I think I'm going to need to.
11) Everything is now out of focus
12) Focus on whatever's dead-center in the ground glass.

etc.


At first, tiny movements of the lens board and a dozen cycles through the "which way, small tilt, refocus, check it" loop is the fastest way to get where you need to go. It's really easy to think you need to move the lens more than you actually do and overshoot. You end up frustrated and wondering why it doesn't work the way the book says. With some practice, it's pretty easy to get down to 2-3 cycles the adjust-focus-check loop to get it dialed in.

[Vaughn]

I would not have used any swing...it does not seem to be needed in that example...just front (and/or back) tilt.

Once I think I have the plane of focus where I want it, and if I still have unfocused areas with the lens wide open, I close down the lens slowly watching the out of focus areas. If both near and far areas come into focus as the same time (same aperature), then I know I am focused properly. If the front comes into focus first, then I move the focus back until I can get far and near to come into focus at the same aperature.

[bieber]

Thanks for the tips, everyone. Another interesting thing this brings up is, in a scenario like this where the object is both tilted horizontally at an angle and pretty flat vertically, how do you decide between using tilt or swing? Do you just go for whichever one is closest to the natural focal plane? Or is tilt inherently easier to work with than swing?

[williaty]

Thanks for the tips, everyone. Another interesting thing this brings up is, in a scenario like this where the object is both tilted horizontally at an angle and pretty flat vertically, how do you decide between using tilt or swing? Do you just go for whichever one is closest to the natural focal plane? Or is tilt inherently easier to work with than swing?

Tilt and swing are identical. Roll the camera 90* and one becomes the other without anything changing.

If the camera is plumb in roll (level left-to-right) and the table top is level and the camera is just pitched down to see the watch, you'd need pure tilt. If you've got both pitch and roll in the camera position, or if the table top is not at right angles to reality, then you'll need some combination of swing and tilt together to get the plane of focus to match the plane of the table top.

Now, that assumes that your goal is to lay the plane of focus down on the table top. I think that's how most people would approach the problem.



However, with the specific cast of the watch, there's a second, less obvious, way to tackle it. You can also swing the plane of focus until it lies parallel to the strap. If you do that, you can get the whole watch in focus by using the swing movement to get the opposite tips of the strap in focus and then stopping down until the Depth of Field is fat enough to keep the entirety of the watch case sharp. The potential downside (or potential awesomeness) to doing this is that the plane of focus would slice through the table. This means that, while the watch would be sharp, the table down and to the left (relative to your picture) of the watch and up and to the right of he watch would be out of focus. With the all-white table you used, this won't be apparent. However, if the watch were resting on something with texture, pattern, or print, you'd see this weird cockeyed slice of sharpness the watch would be lying in. This would be either horrible or awesome.

[Vaughn]

Mentally place a flat plane over the surface you are photographing -- in your example, it is parallel to the same plane the watch is laying on. I do not see this plane as titled right/left, even though the object goes right to left in the scene.

If the watch was on a textured surface, using swing would throw some of that texture out of focus.

[Mark Sawyer]

Scheimpflug works in theory, but it seldom works in practice. There's always something sticking up or out where it shouldn't be. God made the world that way just to torment large format photographers.

It's perhaps the best example of Yogi Berra's old observation, "In theory, practice and theory are the same thing. In practice, they're not."

[Emmanuel BIGLER]

Hi!

You can have a look at the two diagrams below

https://www.flickr.com/photos/43175600@N00/6245674051/

https://www.flickr.com/photos/43175600@N00/6246188810/

An important thing to keep in mind is that for landscape photography, tilt angles required for a proper Scheimpflug setting are small.
For example, with a 150 mm lens, located at 5 feet above ground [1.5 m], tilting by 6° brings ground level to sharp focus, this is a small angle!
So, do not start with a tilting angle of 30° like in this legendary advertising picture, (http://www.badgergraphic.com/opencart/image/cache/data/product/266-456x456.jpg) you'll have really hard times to find the proper focus ;-)

However, tilt angles can be high with long focal lengths, for taking pictures of objects located close to the camera, like for pictures of objects for catalogues in the good old days whre the LF camera was the proper tool for this kind of commercial photography.


My goal was to get the entire thing in focus

Another thing to remember is that tilting + Scheimpflug changes the shape of the portion of object space considered as sharp to a wedge, so objects outside the wedge will not be seen sharp.

For example, for this kind of "tunnel-style" picture, again taken from an advertising image (for a local well-known French product), (http://en.franche-comte.org/var/ezflow_site/storage/images/media/images/marques/urbain-et-grands-sites/cave-a-comte/91414-1-fre-FR/cave-a-comte_format_924x367.jpg)tilting the lens or the film plane will be useless to achieve "infinite depth of field".

You can find a nice Scheimpflug simulator by Tim Parkin here

http://static.timparkin.co.uk/static/focus/

See this discussion with examples on the British LF forum.
http://www.lf-photo.org.uk/forum/viewtopic.php?f=3&t=2859

Yes, simulation will be useless in the field, but at least a simulation at home can tell you what is possible, or not, to achieve in the field, in the real, non-mathematical world.

[Doremus Scudder]

Mentally place a flat plane over the surface you are photographing...

Vaughn's point here is right on. It's really important to mentally determine the where you want the plane of sharp focus to be; only then can you really decide what movements you need.

So, if the plane you want to focus on is not parallel to the back, but displaced top-to-bottom (e.g., like a near/far landscape of flat ground), you need tilts. If the plane you want to focus on is displaced right-to-left (e.g., a wall is close on one side and stretches into the distance), you need swings. If the displacement is oblique, you'll need both swing and tilt.

Once I've determined where I want the plane of sharp focus to be, I choose focus points: top and bottom of the ground glass for scenario 1, right and left of the ground glass for scenario 2 and both top/bottom and right/left for scenario 3.

Before applying tilts or swings remember the old adage: "The back backs away from; the lens looks at" the plane you want to have in focus. Now, unless you have asymmetrical tilts/swings, my procedure would be:

For top/bottom points, I always focus first at the bottom of the ground glass (usually the more distant point) since I use field cameras with base tilts a lot. With axis tilts it doesn't matter. Anyway, I focus on the first focus point then tilt lens stage or camera back in the proper direction (i.e., "backing" or "looking") till both my chosen points are equally out of focus. Then refocus on the first point and then focus on the second point. If you have to rack out the bellows to get point 2 in focus, you need to move back or lens stage in whatever direction is needed to make the bellows a bit shorter. Vice-versa if you have to shorten bellows extension to get point 2 in focus. A couple iterations of this and you'll have both your points in focus.

The above works for left/right focus points too, just substitute "left/right" for "top/bottom."

For an oblique plane of sharp focus, you'll have to make adjustments for both top/bottom and left/right. It doesn't matter which you start with, however your focus points need to be directly opposite each other on the ground glass (i.e., vertically and horizontally aligned for top/bottom and right/left respectively). Apply one movement (say tilt), then apply the second (swing) using the method above. Then go back and check.

It's really easier and faster to do than to explain.

The next step for me is finding the focus spread between the nearest and farthest objects I want in focus in the scene (like Mark points out, there are always pesky objects that don't lie in your chosen plane :) ). If you've applied tilts/swings you need to really remember that the "near" is not necessarily what is nearest to you and the camera, but what is in front and/or above the plane of sharp focus and the "far" is what is behind and/or below the plane of sharp focus. I find the focus spread and note the bellows extension for each extreme. I then position focus exactly halfway between the two extremes and use the total distance to find the optimum f-stop to keep everything acceptably sharp (as described in the article on the LF home page, "choosing the optimum f-stop").

Hope this helps,

Doremus

[Bob Salomon]

Vaughn's point here is right on. It's really important to mentally determine the where you want the plane of sharp focus to be; only then can you really decide what movements you need.

So, if the plane you want to focus on is not parallel to the back, but displaced top-to-bottom (e.g., like a near/far landscape of flat ground), you need tilts. If the plane you want to focus on is displaced right-to-left (e.g., a wall is close on one side and stretches into the distance), you need swings. If the displacement is oblique, you'll need both swing and tilt.

Once I've determined where I want the plane of sharp focus to be, I choose focus points: top and bottom of the ground glass for scenario 1, right and left of the ground glass for scenario 2 and both top/bottom and right/left for scenario 3.

Before applying tilts or swings remember the old adage: "The back backs away from; the lens looks at" the plane you want to have in focus. Now, unless you have asymmetrical tilts/swings, my procedure would be:

For top/bottom points, I always focus first at the bottom of the ground glass (usually the more distant point) since I use field cameras with base tilts a lot. With axis tilts it doesn't matter. Anyway, I focus on the first focus point then tilt lens stage or camera back in the proper direction (i.e., "backing" or "looking") till both my chosen points are equally out of focus. Then refocus on the first point and then focus on the second point. If you have to rack out the bellows to get point 2 in focus, you need to move back or lens stage in whatever direction is needed to make the bellows a bit shorter. Vice-versa if you have to shorten bellows extension to get point 2 in focus. A couple iterations of this and you'll have both your points in focus.

The above works for left/right focus points too, just substitute "left/right" for "top/bottom."

For an oblique plane of sharp focus, you'll have to make adjustments for both top/bottom and left/right. It doesn't matter which you start with, however your focus points need to be directly opposite each other on the ground glass (i.e., vertically and horizontally aligned for top/bottom and right/left respectively). Apply one movement (say tilt), then apply the second (swing) using the method above. Then go back and check.

It's really easier and faster to do than to explain.

The next step for me is finding the focus spread between the nearest and farthest objects I want in focus in the scene (like Mark points out, there are always pesky objects that don't lie in your chosen plane :) ). If you've applied tilts/swings you need to really remember that the "near" is not necessarily what is nearest to you and the camera, but what is in front and/or above the plane of sharp focus and the "far" is what is behind and/or below the plane of sharp focus. I find the focus spread and note the bellows extension for each extreme. I then position focus exactly halfway between the two extremes and use the total distance to find the optimum f-stop to keep everything acceptably sharp (as described in the article on the LF home page, "choosing the optimum f-stop").

Hope this helps,

Doremus

Or just pick your self up a Rodenstock/Linos pocket Scheimpfluf/depth of field calculator and just follow its instructions!

[minh0204]

Did you check out the guide?

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

Procedure I works very well for me, I can get the correct tilt within 2-3 iterations (after some practice).

[Keith Pitman]

If you are using a P2, you have asymmetric tilts and swings. Simply focus on one of the lines on the groundglass with the focus knob (lower line for tilt and right line for swing), then use the tilt or swing to bring the other line into focus. You should be using the fine tilt knob, not the coarse tilt lever. This effectively automates Scheimpflug. It’s covered in the Sinar manual, but it more clear in a specific publication Sinar had. It used to be on the Sinar website.

[bieber]

Thanks a bunch for the advice everyone, I've got a lot of practicing to do.


If you are using a P2, you have asymmetric tilts and swings. Simply focus on one of the lines on the groundglass with the focus knob (lower line for tilt and right line for swing), then use the tilt or swing to bring the other line into focus. You should be using the fine tilt knob, not the coarse tilt lever. This effectively automates Scheimpflug. It’s covered in the Sinar manual, but it more clear in a specific publication Sinar had. It used to be on the Sinar website.

Ohhh, so that's what those dashed lines on the ground glass are for. Thanks, now that you put it that way that part of the manual makes a lot of sense. So the way the rear tilt and swing work is that they shift about an axis such that they always keep either the right or the bottom dashed line in focus, if it was in focus before the movement began? I'm curious about this bit, though:



Direct tilting and swinging with the lens standard is not satisfactory as its optical setup always involves an overall sharpness shift


So what they're saying is that if you want to use front standard tilt or swing, you should first find the correct amount of movement with the rear standard, and then transfer the movement to the front standard, right? I can see why moving the lens around would be different from moving the standard, but won't copying the movement from the rear standard introduce the same optical effect that you would have seen if you'd just moved the front standard directly in the first place?

[Mark Sawyer]

(like Mark points out, there are always pesky objects that don't lie in your chosen plane :) ).

But if you apply the Scheimpflug Principle correctly, at least you'll also have vignetting from having your lens way off-axis... :)

[Thom Bennett]

"I was told there would be no math." Focus on the farthest point you need in focus. Tilt the lens towards the plane you want in focus until the nearest point and farthest point are equally out of focus. It's easy to see on the ground glass. Lock the tilt. Then refocus on the farthest point and everything along the plane will now pop into focus. Since you are dealing with a 3D object, stopping the lens down will now help to bring parts of the object that lay outside of the flat plane into focus. Again, watch the gg as you stop down and you will see just how far you need to go. Best to practice on something not so small. Try the top of a table and get used to the movements before moving to smaller objects. Once you get the gist of it you can tackle any subject. Same principles apply to the swing. In your case a bit of both tilt and swing would probably be ideal. Good luck!

[Peter Collins]

others have said it: Use very little tilt, and check.

[John Layton]

A really good fresnel (like a Maxwell) will allow for a decent amount of stopping down while still being viewable enough to visually evaluate DOF.

[Bob Salomon]

A really good fresnel (like a Maxwell) will allow for a decent amount of stopping down while still being viewable enough to visually evaluate DOF.

Maxwell is a fresnel? Or is it a ground glass? Or is it the same as the Linhof Super Screen or the Beattie screen?

A Fresnel is a non focusing screen with concentric grooves on one side. It is sandwiched, usually on top, with a ground glass.
A ground glass is ground on one side and that side is what the image is formed on.
A Super Screen or Beattie Screen, both were manufactured by Fresnel Optics in Rochester, NY, the parent company of Beattie, is a piece of acrylic, like a fresnel, with one side frosted to act as a focusing surface and the other side grooved like a fresnel Screen.

Which are you suggesting?

[williaty]

Maxwell is a fresnel? Or is it a ground glass? Or is it the same as the Linhof Super Screen or the Beattie screen?

A Fresnel is a non focusing screen with concentric grooves on one side. It is sandwiched, usually on top, with a ground glass.
A ground glass is ground on one side and that side is what the image is formed on.
A Super Screen or Beattie Screen, both were manufactured by Fresnel Optics in Rochester, NY, the parent company of Beattie, is a piece of acrylic, like a fresnel, with one side frosted to act as a focusing surface and the other side grooved like a fresnel Screen.

Which are you suggesting?

The answer to that is dependent on which size of ground glass you need. Depending on size, he can do anything from a simple fresnel with high transmission coatings to whatever magic pixie dust is used to make his complete ground glass replacement wonder-screens (I'm not mocking the product, the top of the line screens are stunning). The bigger stuff only gets the simple fresnel, the smaller screens can have your choice of price-and-performance.

[Bob Salomon]

The answer to that is dependent on which size of ground glass you need. Depending on size, he can do anything from a simple fresnel with high transmission coatings to whatever magic pixie dust is used to make his complete ground glass replacement wonder-screens (I'm not mocking the product, the top of the line screens are stunning). The bigger stuff only gets the simple fresnel, the smaller screens can have your choice of price-and-performance.

Sorry, I am not sure what your point is. As I was the Linhof, Rodenstock and Wista distributor from 1980 till 2015 and before that I was with Sinar I have never heard of your point. Perhaps you have more experience then I have. BTW, I was also a commercial photographer before all that and an USAF photographer.
So please enlighten me so I can understand your point.

[williaty]

Sorry, I am not sure what your point is. As I was the Linhof, Rodenstock and Wista distributor from 1980 till 2015 and before that I was with Sinar I have never heard of your point. Perhaps you have more experience then I have. BTW, I was also a commercial photographer before all that and an USAF photographer.
So please enlighten me so I can understand your point.

Yes Bob, we're all well aware of, and very tired of, your professional credentials at this point. You shove them in everyone's faces at every opportunity.



You asked as question:

Maxwell is a fresnel? Or is it a ground glass?
I gave you a direct answer: He makes both. Whether both options are available depends on what size you need.


Any difficulties in communication and understanding are not mine.

[Bob Salomon]

Yes Bob, we're all well aware of, and very tired of, your professional credentials at this point. You shove them in everyone's faces at every opportunity.



You asked as question:

I gave you a direct answer: He makes both. Whether both options are available depends on what size you need.


Any difficulties in communication and understanding are not mine.

Good, but that was not how you originally phrased it!

[Bob Salomon]

Yes Bob, we're all well aware of, and very tired of, your professional credentials at this point. You shove them in everyone's faces at every opportunity.



You asked as question:

I gave you a direct answer: He makes both. Whether both options are available depends on what size you need.


Any difficulties in communication and understanding are not mine.

Visiting his web site, such as it is, shows focusing screens and not separate fresnel screens. So, exactly what does he offer? Ground glass? Fresnes? Combination sandwichs of both or systems like Beattie?

[williaty]

Visiting his web site, such as it is, shows focusing screens and not separate fresnel screens. So, exactly what does he offer? Ground glass? Fresnes? Combination sandwichs of both or systems like Beattie?

Bob, I answered this already.

If you have any more questions, I'm sure your decades of experience allow you to operate a telephone and call Bill. He'll be very happy to talk to you at great length about his products.

[Bob Salomon]

Bob, I answered this already.

If you have any more questions, I'm sure your decades of experience allow you to operate a telephone and call Bill. He'll be very happy to talk to you at great length about his products.

Although he is located just a few miles from me I have no interest in finding out more then he posts on his site, which is very little!

[williaty]

Although he is located just a few miles from me I have no interest in finding out more then he posts on his site, which is very little!

Like so many in the LF community, Bill Maxwell is older and would rather talk to you than write for a website. Could be worse, to order things from John Coffer, you have to write a letter and include the payment in the envelope. At least Bill has a phone!

[Bob Salomon]

Like so many in the LF community, Bill Maxwell is older and would rather talk to you than write for a website. Could be worse, to order things from John Coffer, you have to write a letter and include the payment in the envelope. At least Bill has a phone!
He has a web site, he could rent a junior high school student to flush it out.
I doubt that he is much older, if at all, then me and flushing out some short paragraphs is not that time consuming but can be very economically rewarding!

[aclark]

I have a simple method of using tilt to bring everything on a horizontal surface into focus .Let's say I am photographing a level path and want everything from the foreground to the distant part of the path to be in focus.

1. Apply front tilt. Guess the amount.

2. Focus on the distant part of the path, which will be somewhere near the bottom of the screen.

3. Check the focus of the near part of the path, which will be somewhere near the top of the screen. Let's say it is out of focus. DO NOT TWIDDLE THE FOCUS BACK AND FORTH TO GET IT SHARP. Instead, move the focus one way only. If you need MORE extension to get the near object sharp, then you need MORE tilt. If LESS extension gets it sharp, then you need LESS tilt.

4. Re-adjust the amount of tilt accordingly, and then repeat the whole sequence until near and far objects are both in focus.


I didn't think this up for myself; I read it somewhere. It is very simple to do and works very well.

Alan

[Maris Rusis]

Or to put it poetically:


Focus on the far,
And tilt towards the near.
Then fiddle with the camera,
Until everything is clear.

[andrewch59]

Fred Newman has a very good video on camera movements, he will email the written version if you contact him.
https://www.youtube.com/watch?v=0JU-eHpk97Y&t=87s

[MikeH]

Focus on the far,
And tilt towards the near.
Then fiddle with the camera,
Until everything is clear.

Very Good! Thank You!