Perspective control in the darkroom

[BetterSense]

Sometimes I have my speed graphic and can't use front rise/fall, so the negative has keystoning. Or, I have keystoning on my 35mm or medium format negatives. In the darkroom, I can tilt the easel and correct for the converging lines, if I left enough space around the subject to crop, but it never seems to look quite right. I have learned from digital people that when they correct perspective, they don't just do a keystoning, they also do a stretch. I don't see how I can do that in the darkroom.

What are the best practices for correcting keystone distortion in the darkroom, and is there a difference between doing it in-camera or in the darkroom, besides that you more efficiently use the negative if you do it in-camera?

[Bob Salomon]

Much esasier and far more consistent by doing it in camera. But to have complete control you need a camera with a back that can at least tilt, swing would also be useful. Then you have it on the negative and it is always there when you go to print it. Doing it your way can waste a lot of time setting it all up and go to print another day and it may not match the first time.

[Gem Singer]

It's much more effective to do perspective control in the camera, and thus in the negative.

That's the primary purpose for the perspective controls on a view camera.

Photoshop can be utilized for making perspective corrections in post-processing, but it is not as effective as the corrective movements that can be done in the camera.

In other words, there is no substitute for a good negative.

[Sirius Glass]

In other words, there is no substitute for a good negative.

You can say that again!


In other words, there is no substitute for a good negative.

Thank you,
Steve

[cowanw]

google is your friend
http://www.largeformatphotography.info/forum/archive/index.php/t-66580.html
tilt the lens andor the easil.

[ic-racer]

I have learned from digital people that when they correct perspective, they don't just do a keystoning, they also do a stretch. I don't see how I can do that in the darkroom.

With a large format negative you can't tell if you tilted in the darkroom or did front rise. Massive coverage lenses are expensive; enlargers are cheap! The nice thing about doing it in a darkroom under the enlarger is that the 'stretch' is built in. As the rays from the enlarger lens get farther away the image grows in two dimensions. It gets both wider and longer. Optics are cool!

$100 210mm lens or $4000 210mm Super Angulon??? Only you know :)
http://img.photobucket.com/albums/v670/ic-racer/2011/corrected.jpg

[BetterSense]

Photoshop can be utilized for making perspective corrections in post-processing, but it is not as effective as the corrective movements that can be done in the camera.

Effective in what way? Is there a geometric difference?


tilt the lens andor the easil.
I know; that's what I do now. Since my negative stage doesn't tilt, I have to stop down a lot. But I'm not always happy with the results, because it sometimes looks "off". I wasn't sure if doing it in the camera is any different than doing it in the darkroom, or if there are better, more accurate techniques for correcting converging verticals in the darkroom. Does it matter how long of an enlarger lens you use?

[Brian Ellis]

It's much more effective to do perspective control in the camera, and thus in the negative.

That's the primary purpose for the perspective controls on a view camera.

Photoshop can be utilized for making perspective corrections in post-processing, but it is not as effective as the corrective movements that can be done in the camera.

In other words, there is no substitute for a good negative.

I have to disagree. If all we're talking about is keystoning alone, and not keystoning combined with other distortions, in my experience it's as fully effective to fix it in Photoshop as it is in camera, as long as the need for correction was known and planned for when the photograph was made.

[Gem Singer]

Brian,

I think I used the wrong word when I stated "effective". I probably should have used "easier".

You're probably much more skilled in Photoshop. Whenever I try to correct perspective in PS, the results just don't look natural to my eye.

It's easier for me to use tilt and swing movements in the camera.

[Brian Ellis]

Brian,

I think I used the wrong word when I stated "effective". I probably should have used "easier".

You're probably much more skilled in Photoshop. Whenever I try to correct perspective in PS, the results just don't look natural to my eye.

It's easier for me to use tilt and swing movements in the camera.

I agree, it's certainly easier in camera if the camera has the movements.

[Doremus Scudder]

BetterSense,

Control in the darkroom is optically the same as control in the camera, however, when you correct keystoning, as you are, you can lose a lot of the negative. What you are doing, in essence is adding "reverse keystoning" to bring the converging verticals in the image to parallel. This should not add other distortions to the image. The fact they don't "look right" to you is likely from seeing the original convergent image in comparison and feeling that the squared-up image looks a little "dumpy." Nevertheless, barring lens aberrations, etc. it should be "correct."

I often make small adjustments with the easel (shims, cardboard, etc.) to get things squared up, even thought I've done my best in the field with the camera. Getting an image square that has a lot of keystoning requires quite a lot of adjustment at the easel stage and you lose a lot of negative plus introduce an exposure difference from the highest to lowest part of the image (correctable by uncovering the image from bottom to top over a period of time). That's why camera corrections are better. To minimize the amount you have to adjust, you can shoot wider and try to keep the camera back as close to parallel to the subject as possible. This means you will have to crop the bottom off, but you will end up with less correction needed. Also, if you tilt the camera up to include the top of a building, etc., you can still bring the back parallel, or closer to parallel. If your lens has tilt, you can tilt it to parallel as well, thus imitating a front rise (typical point and tilt strategy). If the lens won't tilt, you can sometimes compensate with depth of field, or maybe compromise a bit to minimize the corrections you have to do in the darkroom.

A word about focus when doing this. My Beseler enlargers have a lens stage that moves, so you focus exactly like you would with a view camera, adjusting lens stage to match the easel adjustment. The Omega enlargers I often work on do not have this. However, I have found that the depth of focus at the easel is fairly large and, as long as you stop down enough, adequate for moderate adjustments.

Hope this answers your question,

Doremus Scudder

[Bob Salomon]

I have to disagree. If all we're talking about is keystoning alone, and not keystoning combined with other distortions, in my experience it's as fully effective to fix it in Photoshop as it is in camera, as long as the need for correction was known and planned for when the photograph was made.

But then you lose image size and composition over the original and it never seems to be as sharp as one that was not manipulated.

[BetterSense]

So does the focal length of the enlarger lens not matter? I had a hunch that the darkroom correction was only 'correct' if the enlarger lens was the same focal length as the taking lens, but I can't articulate why that should matter.

[IanG]

It makes no difference as the negative is two dimensional. It does make a great difference at the taking stage.


I agree, it's certainly easier in camera if the camera has the movements.

Equally the lens needs the coverage which is why even when the maximum practical movements are use it may still be necessary to correct at the enlarging stage.

Ian

[Brian Ellis]

But then you lose image size and composition over the original and it never seems to be as sharp as one that was not manipulated.

See the part of my message where I said "as long as the need for correction was known and planned for when the photograph was made."

I'm not sure what you mean by "never seems to be as sharp." There shouldn't be any loss of sharpness when you make this adjustment. What tools are you using when you do this in Photoshop? And what version of Photoshop are you using? CS5 added a few things that make the adjustment a little easier than earlier versions.

[engl]

See the part of my message where I said "as long as the need for correction was known and planned for when the photograph was made."

I'm not sure what you mean by "never seems to be as sharp." There shouldn't be any loss of sharpness when you make this adjustment. What tools are you using when you do this in Photoshop? And what version of Photoshop are you using? CS5 added a few things that make the adjustment a little easier than earlier versions.

It depends on how big adjustments are made. If you aim a camera up to capture a building, and then correct perspective in post processing, you are basically stretching the top of the image which will make the image look less sharp. If the base of the building occupies the full width of the negative, and the top is half of this, you will be doing a 1:2 stretch horizontally in that part of the image. Beyond this, the image also has to be stretched vertically.

Of course, with minor adjustments, there won't be much of a loss.

[Bob Salomon]

See the part of my message where I said "as long as the need for correction was known and planned for when the photograph was made."

I'm not sure what you mean by "never seems to be as sharp." There shouldn't be any loss of sharpness when you make this adjustment. What tools are you using when you do this in Photoshop? And what version of Photoshop are you using? CS5 added a few things that make the adjustment a little easier than earlier versions.

This explains it as well as anything else.

http://www.rodenstock-photo.com/mediabase/original/Entzerren_am_Computer_A4_e_Druck_7860.pdf

[Sirius Glass]

While I was in college, one of the guys in the dorm was getting married and he asked me to photograph the wedding. Both he and his bride were of the shall we say generous size persuation.

I make sure that I never had a round or square object in the background.

I tilted the easel for all the prints. I printed albums for the couple and both sets of parents plus a few extra prints. Everyone loved the photographs! They said no one had ever captured the "real" them! So the orders from reprints came from both families ... that one photo shoot paid for tuition, room and board, all my textbooks and my expenses for one year in college.

Therefore, I submit that perspective control can be achieved in the darkroom.

Steve

[Brian Ellis]

It depends on how big adjustments are made. If you aim a camera up to capture a building, and then correct perspective in post processing, you are basically stretching the top of the image which will make the image look less sharp. If the base of the building occupies the full width of the negative, and the top is half of this, you will be doing a 1:2 stretch horizontally in that part of the image. Beyond this, the image also has to be stretched vertically.

Of course, with minor adjustments, there won't be much of a loss.

Sorry but I'm not following you. First, the base of the building wouldn't occupy the full width of the negative, assuming as I said that you plan to make the adjustment at the time you make the photograph. So you don't do a horizontal "stretch" (at least I don't). And any loss of sharpness from the vertical "stretch" will of course depend on the size of the crop and the size of the print just like any other cropped photograph. But there's nothing in the methodology that inherently leads to a loss of sharpness. I've done this many times with prints up to about 16x20 and never seen any loss of sharpness.

[Brian Ellis]

This explains it as well as anything else.

http://www.rodenstock-photo.com/mediabase/original/Entzerren_am_Computer_A4_e_Druck_7860.pdf

Thanks Bob but I didn't ask for a cite to a magazine article about doing this in Photoshop vs doing it in camera since I've often done it both ways myself. You told me that making the adjustment in Photoshop leads to a loss of sharpness. So I assumed that you must do this yourself with some frequency and that you've experienced the loss of sharpness of which you spoke. And since I do it pretty often myself with no loss of sharpness I was just curious about what tools you use in Photoshop and what version of Photoshop you use. If you don't care to answer those two questions that's certainly fine, there's no obligation, I was just interested in knowing how you were doing it.

I did notice that the version of Photoshop that is described in the article appears to be some older version of Photoshop, not CS5 or even CS4. The article talks about getting to the adjustment by going to Edit, then Transform, then Perspective. That isn't the method I use to correct for keystoning and I don't think it's the method most people use today. I only skimmed the methodology described but that also didn't even look like the same method used today. But since I didn't read the article carefully I could be wrong.

[Thebes]

If you want to do it in an enlarger, rather than PS or in camera, wouldn't you need to tilt two of [lens, easel, film stage] in order to both correct the distortion and retain tight focus on the film grain?

[engl]

Sorry but I'm not following you. First, the base of the building wouldn't occupy the full width of the negative, assuming as I said that you plan to make the adjustment at the time you make the photograph. So you don't do a horizontal "stretch" (at least I don't). And any loss of sharpness from the vertical "stretch" will of course depend on the size of the crop and the size of the print just like any other cropped photograph. But there's nothing in the methodology that inherently leads to a loss of sharpness. I've done this many times with prints up to about 16x20 and never seen any loss of sharpness.

I don't understand how you'd do perspective correction without stretching horizontally. Neither do I understand your comment about the base of the image not occupying the full width of the negative. How much of the width you let it occupy makes no difference, you will still have to stretch the upper part of the building to match the width of the base, otherwise you won't have corrected verticals (in a "dead on" example).

Perspective correction crops away part of the image, then redistributes the remaining information unevenly. It is absolutely impossible to retain the same sharpness and in the entire image with the same final print size (which I always assume, anything is sharp if printed small enough). The different tools for correcting perspective all do the same thing (transform->perspective, the "perspective" crop option, lens correction tool etc.).

However, if the adjustments are not too big, the degradation might not matter.

[IanG]

If you want to do it in an enlarger, rather than PS or in camera, wouldn't you need to tilt two of [lens, easel, film stage] in order to both correct the distortion and retain tight focus on the film grain?

If the enlarger allows it then yes. However it is possible to just tilt the easel, then stopping down well will give enough DOF to give a perfectly sharp image with no drop in image quality unlike Photoshop).

Ian

[engl]

If the enlarger allows it then yes. However it is possible to just tilt the easel, then stopping down well will give enough DOF to give a perfectly sharp image with no drop in image quality unlike Photoshop).

Ian

Doing this you'd lose sharpness just like you do with Photoshop, for the same reason. You'd lose more even, since you are no longer at the optimum aperture of the enlarging lens, or you are at (or closer to) optimum aperture and have much of the negative out of focus.

Keeping the same print size, it is not possible to throw away part of the negative and keep the same sharpness, unless the print size is small enough to not display the loss.

[ic-racer]

With a proper tilted image, the focus is, of course, perfect. No more loss of sharpness than you would expect tilting your $5000 Fine Art XXL lens on your camera. Indeed the top portion of the my print is enlarged 1.4 time while the bottom is enlarged only 1.0 times from the original 8x10 negative. So there is a theoretical increase in grain and decrease in sharpness. This is only theoretical because that reproduction ratio (1.4 power) yields an image that is sharper than what can be resolved by the paper. So, unless you are a kludge at burning, there is no way the observer can tell how the image was made. In fact, I'll go out on a limb and say that, depending on your lens, you are likely to get more of a fall-off in image quality when using the far reaches of your camera lens circle.

The advantage to correcting converging verticals in the camera is one of convenience. Doing it in the darkroom involves using the center of both the taking lens and the enlarging lens (there is no shift during the projection printing, it is a tilt. The Durst lensboard does shift, but that is so you can re-center the image circle after the tilt).

I think correcting verticals in the darkroom has become a lost art and I think it is good to see an interest in these techniques. Even the small Durst enlargers can be set up to do this.

[Bob Salomon]

Brian,,
Whatever suits your nrrds.

[IanG]

Doing this you'd lose sharpness just like you do with Photoshop, for the same reason. You'd lose more even, since you are no longer at the optimum aperture of the enlarging lens, or you are at (or closer to) optimum aperture and have much of the negative out of focus.

Keeping the same print size, it is not possible to throw away part of the negative and keep the same sharpness, unless the print size is small enough to not display the loss.

You need to try it for yourself. With a LF negative you won't see any difference and MF is the same, even with 35mm it's only likely to be apparent on larger prints.

Ian

[Brian Ellis]

Brian,,
Whatever suits your nrrds.

??????

[sanking]

There is obviously some loss of resolution when you stretch an image file in Photoshop to make perspective changes. However, my experience is that this rarely results in a loss of sharpness on the print because in most cases when printing with LF and MF negatives there is a lot more sharpness in the file than is needed. So even if you stretch the image and reduce resolution in the file, the detail on the print will still be beyond the threshold of resolution of the human eye. Unless you print very, very large.

Sandy

[Brian Ellis]

I don't understand how you'd do perspective correction without stretching horizontally. Neither do I understand your comment about the base of the image not occupying the full width of the negative. How much of the width you let it occupy makes no difference, you will still have to stretch the upper part of the building to match the width of the base, otherwise you won't have corrected verticals (in a "dead on" example).

Perspective correction crops away part of the image, then redistributes the remaining information unevenly. It is absolutely impossible to retain the same sharpness and in the entire image with the same final print size (which I always assume, anything is sharp if printed small enough). The different tools for correcting perspective all do the same thing (transform->perspective, the "perspective" crop option, lens correction tool etc.).

However, if the adjustments are not too big, the degradation might not matter.

O.K., I see what you're saying about horizontal stretching, I misunderstood the earlier message. And I agree that the upper portion has to be stretched to some extent to match the lower. However, in my fairly extensive experience at making this adjustment in Photoshop any resulting loss of sharpness hasn't been noticeable. I suppose it would be at some much larger print size just as any cropping can lead to a loss of sharpness with a large enough print made at the same size as the uncropped print.

[Doremus Scudder]

Brian,

The "r" key is right next to the "e" key on the keyboard. Bob meant "needs." As kids, we used to type in "secret code" by moving our fingers one key one way or the other.

Example: ksbr s movr fsx"

Have a nice day!

Doremus

[Robert Opheim]

I have had great results correcting in the darkroom using an omega perspective control device. I shoot quite a bit of built environment shots and often there isn't enough coverage of the lens and rise on the camera. This was especially true before I bought technikardan and bag bellows. I used to shoot with a graflex view II and a Omega View both without bag bellows. With the correction control device - I have been able to go back through old negatives and make much better prints.

It sis much easier to get as much information on the negative in the camera - the rest has to be done in the darkroom.

Anyway enlarger perspective control works very well. Some of the image bottom (usually) is cropped away. A pivoting grain magnifier is required - as well as blocking up the easel.