correct verticals with a wide angle !

ader · Sep 13, 2003

Hi,

I was assisting an interior photographer the other day and he said that whilst doing interiors you had to get the verticals correct. Which I thought was fair enough but then when I went home I realized he was using a wide angle approx. 50mm (medium format), and as wide angles distort verticals, how does one get them correct using a wide angle ??. He was using a mamiya 67 medium format with a 50mm...

ta ader

Lex Jenkins · Sep 13, 2003

Wide angles don't necessarily distort verticals. We're dealing with two kinds of "distortion" here, tho'. One is an optical flaw, usually barrel distortion. The very best (and even some quite affordable) wide angles have little or no perceptible barrel distortion.

The second type is more a problem of perspective than distortion. When a camera is aimed upward or downward the converging lines of a tall structure seem to be exaggerated. In the worst cases a building can appear to be falling over backward or collapsing inward. This can occur even with lenses that are well corrected for barrel or other distortions.

The common method for minimizing this effect is to avoid tilting the camera, lens and film plane up (or down, as the case may be) along the same axis (hope I'm not fouling up my terminology here and making matters worse).

The simplest method is to keep the film plane the same as the building (or whatever vertical surface, be it a wall, fence, etc.) and move the lens up (or down). Technically this is call rise and fall, tho' in common parlance the term shift is used. Strictly speaking shift refers to side-to-side movement.

There are other methods but describing them is a poor substitute for demonstration with illustrations.

I'm not familiar with the Mamiya system but chances are good that the photographer you were assisting used either a shift lens or elevated the entire camera to a point where it would not be necessary to tilt the camera in order to take in the entire scene.

bjorke · Sep 13, 2003

...or he could have corrected in the printing, either by tilting the enlarger lens, baseboard, and/or head -- or by using PhotoShop.

dr bob · Sep 13, 2003

The closest I can come to your interior photographer is the 55mm lens on my Mamiya C330f. As Lex has said, this lens has never “distorted” parallel lines. When one observes a tall building, looking up at an angle for example, his eye sees the convergence just as camera lens does, however his brain works to “correct” this because it “knows” from experience that the building is not really falling backward. Then when he makes the picture and observes the same scene his brain tells him the lines should be parallel. But the flat image shows all the convergence vividly just as recorded. It is not really distortion.

Again as Lex has pointed out there are ways to eliminate or minimize convergence. With a 35mm camera your choices are limited. I have used Nikon’s perspective control lens a couple of times (borrowed) but I have found the best policy is to keep the film plane parallel (as Lex said). One way is to get as far from the offending subject as possible then enlarge to suit. I have also resorted to tilting the enlarger base (explained by bjorke), stopping down and exposing for (a long time).

The bottom line is that most lenses do not distort (much) - photographers sometimes do. Perspective control or loss of control is often referred to as distortion but it really is not if you think about it. It can produce bizarre images.

Lex Jenkins · Sep 13, 2003

BTW, one common mistake I see, one which I made myself when I first got my 28/3.5 PC-Nikkor, is to overcorrect for converging verticals.

It's natural to see converging verticals with tall buildings. The eye expects it. By overcorrecting these lines from ground level using a shift lens or camera movements we create a perspective that doesn't exist in nature. It looks odd.

A better method is to gain elevation. That's why you'll see photos of Ansel Adams working from the roof-mounted platform of his vehicle when photographing Yosemite and other favorite spots - he was trying to gain just a little more elevation to prevent having to overcorrect with the camera.

Poco · Sep 13, 2003

When not using a view camera, I'll sometimes use a wide on my 6x7 to avoid converging lines. The 50mm lets me shoot with the camera level to avoid converging lines and still get enough of the top of a scene in the frame so I can then crop the bottom in the darkroom and have my shot. In effect, rise on a view camera is nothing but in-camera cropping. This might be what your photographer was doing.

bjorke · Sep 13, 2003

Is that really why Adams had a roof platform? I thought it was just to avoid having his view obscured by other viewers/cars (and providing him with a steady workplace most anywhere). The 5-6 foot difference would be pretty miniscule against a mountainside.

Donald Miller · Sep 14, 2003

bjorke said:
Is that really why Adams had a roof platform? I thought it was just to avoid having his view obscured by other viewers/cars (and providing him with a steady workplace most anywhere). The 5-6 foot difference would be pretty miniscule against a mountainside.

Shooting from an elevated platform has many advantages. When using a view camera, the near-far relationship (Adams used this a great deal) can be exploited more effectively without resorting to extremely small lens apertures. Additionally the entire view changes a great deal from an elevated position. From a height of 6 feet the point of diminishment (on the level plane) will be four times nearer then from a height of 12 feet.(equivalent scene and object size). The rear back tilt is most effective from an elevated position.

Lex Jenkins · Sep 14, 2003

What Don said. ;>

Even tho' it gains me only two or three feet in height sometimes I'll set up in the bed of my pickup. It really helps make the most of my PC-Nikkor for architectural shoots.

Ed Sukach · Sep 14, 2003

Lex Jenkins said:
It's natural to see converging verticals with tall buildings. The eye expects it. By overcorrecting these lines from ground level using a shift lens or camera movements we create a perspective that doesn't exist in nature. It looks odd.

True.

The perspective "sensed" by the human eye is approximately equivalent to the image from a 100mm lens in the 35mm format, which translates to something like 160mm in 2 1/4 and ... whatever would be mathematically the same for 4" x 5" and 8" x 10" (Sunday morning and I haven't had breakfast yet). Everything is approxinmate, as no one yet has found a way to make accurate measurements of human perception.

Any image taken with a lens *greatly* longer or shorter in focal length (tilts and swings not considered) results in an unnatural convergence of line and area scale... NOT to be confused with distortion (i.e., barrel, pincushion, or random abberation) - all lens - optical - errors.

Never have I had to amke morecorrections to a message. I NEED coffee.

Donald Miller · Sep 14, 2003

Ed Sukach said:
Lex Jenkins said:

It's natural to see converging verticals with tall buildings. The eye expects it. By overcorrecting these lines from ground level using a shift lens or camera movements we create a perspective that doesn't exist in nature. It looks odd.

Click to expand...

True.

The perspective "sensed" by the human eye is approximately equivalent to the image from a 100mm lens in the 35mm format, which translates to something like 160mm in 2 1/4 and ... whatever would be mathematically the same for 4" x 5" and 8" x 10" (Sunday morning and I haven't had breakfast yet). Everything is approxinmate, as no one yet has found a way to make accurate measurements of human perception.

Any image taken with a lens *greatly* longer or shorter in focal length (tilts and swings not considered) results in an unnatural convergence of line and area scale... NOT to be confused with distortion (i.e., barrel, pincushion, or random abberation) - all lens - optical - errors.

Never have I had to amke morecorrections to a message. I NEED coffee.

Ed,

That is interesting. I had always heard that the so called normal lenses (related to equivalent human vision) were those that were near the diagonal of the format. That would be near 50 mm for 35, 75 or 80 for 6X4.5, 150 for 4X5 and 300 for 8X10. I don't know where this "normal' designation came from. However, I don't know that I personally see the world at the 150 focal length in 6X 4.5, 300 mm length in 4X5, or 600 mm in 8X10 which is what your comment would indicate. That seems to be a fairly strong telephoto view. (It could be that maybe I am just not focused enough in my perceptions)...

At any rate, explain further...I have a feeling that I am about to learn something here.

Lex Jenkins · Sep 14, 2003

At the risk of thread drift I'd consider my sense of vision - and by that I mean with my nekkid eyeballs - to be akin to a well corrected 17mm lens (for 35mm film format) with heavy spherical aberration.

If that sounds peculiar, consider that most of use have very good peripheral vision that matches an ultrawide very closely. Yet we can only clearly see whatever we're actually looking at. Hence the reference to spherical aberration, which in a lens diminishes the clarity at the periphery of an image.

Oddly enough, the photos I've seen over the past few years that seem to most closely mimic my sense of vision were illustrations in Shutterbug taken by Roger Hicks and Frances Shultz with their Alpa WA (which, I believe, stands for Wallet Annihilator) and 38mm Biogon (as in, "gone," referring to all of one's remaining money after buying the camera).

Ed Sukach · Sep 14, 2003

Lex Jenkins said:
At the risk of thread drift I'd consider my sense of vision - and by that I mean with my nekkid eyeballs - to be akin to a well corrected 17mm lens (for 35mm film format) with heavy spherical aberration.
If that sounds peculiar, consider that most of use have very good peripheral vision that matches an ultrawide very closely. Yet we can only clearly see whatever we're actually looking at. Hence the reference to spherical aberration, which in a lens diminishes the clarity at the periphery of an image.

It doesn't sound "peculiar", but it would apply to "field of vision", not "perspective" within that field. That human central vision has more clarity is due to two major reasons; the distribution of rods and cones in the retina of the eye, and the ... I was about to write "psychlogical" .. but that is not quite correct ... uh... "Visual Preconditioning", where the information in the center of the visual field is more important, therefore we are conditioned to pay the most attention to it.

Hmmm... "heavy spherical abberation" ... I don't think I have that in *my* vision... but I'm not sure I'd notice it if I did... the brain can correct all sorts of optical "strangeness" through experiencing and conditioning to make the end result "correct" ... perceptually correct.

That the "normal" lens has a focal length equal to the length of the diagonal (or diameter) of the field is a more or less arbitrarilly choen value... at that focal length/ field diameter, quite a few optical design problems are minimized - that is why "Normal" lenses can have the largest maximum apertures.

jtk · Dec 11, 2018

The word some may be looking for is "rectilinear".

pentaxuser · Dec 11, 2018

Wow! How high does this rank for length of break? Previous comment was Sept 2003 so 15 years 3 months.

Just out of curiosity, jtk, how did you come across a thread that was last seen 15+ years ago? Thanks

pentaxuser.

Sirius Glass · Dec 11, 2018

For wide angles photographs with the Hasselblad, I use the 50mm lens and the SWC [38mm] lens and make sure that the camera is level. That way I do not have the verticals converge.

Mainecoonmaniac · Dec 11, 2018

ader said:
Hi,

I was assisting an interior photographer the other day and he said that whilst doing interiors you had to get the verticals correct. Which I thought was fair enough but then when I went home I realized he was using a wide angle approx. 50mm (medium format), and as wide angles distort verticals, how does one get them correct using a wide angle ??. He was using a mamiya 67 medium format with a 50mm...

ta ader

A lens that wide will have some pincushion or barrel distortion. It all depends on what getting the verticals "Correct" means.

DXO makes software to deal with lens distortions.

Sirius Glass · Dec 11, 2018

Mainecoonmaniac said:
A lens that wide will have some pincushion or barrel distortion. It all depends on what getting the verticals "Correct" means.

DXO makes software to deal with lens distortions.

No, not the Hasselblad Zeiss 50mm and the Hasselblad SWC. Good glass is worth the money.

jtk · Dec 11, 2018

jtk said:
The word some may be looking for is "rectilinear".

pentaxuser said:
Wow! How high does this rank for length of break? Previous comment was Sept 2003 so 15 years 3 months.

Just out of curiosity, jtk, how did you come across a thread that was last seen 15+ years ago? Thanks

pentaxuser.

Slow internet connection. That, plus I was wondering how long it would take for somebody to mention the obvious.

pentaxuser · Dec 11, 2018

jtk said:
Slow internet connection. That, plus I was wondering how long it would take for somebody to mention the obvious.

Thanks but I am not clear how a slow internet got you to this very old thread. It may be that I do something different when I open Photrio but unless I were to research a title close to this one I cannot see anyway for me to stumble across a thread "lost" in the depths Photrio for 15 years.

I always have been curious how anyone discovers or stumbles across old threads that have lain dormant for even months let alone years like this one. It's usually good when they do as it can avoid a duplicate thread and keeps the information under one heading.

As an aside I found reading the names of the contributors to be a little sobering, given that nearly all "disappeared" many years ago and at least one, I am pretty certain, is dead

pentaxuser

ic-racer · Dec 11, 2018

ader said:
Hi,

I was assisting an interior photographer the other day and he said that whilst doing interiors you had to get the verticals correct. Which I thought was fair enough but then when I went home I realized he was using a wide angle approx. 50mm (medium format), and as wide angles distort verticals, how does one get them correct using a wide angle ??. He was using a mamiya 67 medium format with a 50mm...

ta ader

No 'software' is needed. Only 'hardware.'

jtk · Dec 11, 2018

My Canon 35pc was rectilinear.

Theo Sulphate · Dec 11, 2018

On a Hasselblad SWC, the external bubble level is visible in the viewfinder, so you can even hand-hold it to get correct verticals.

I've used mine both on a tripod and handheld. In regard to perfect verticals and perspective, I cannot think of a better lens than the SWC's 38mm Biogon.

--

"slow internet" - I like that

Alan Edward Klein · Dec 11, 2018

ic-racer said:
No 'software' is needed. Only 'hardware.'
View attachment 212733

Curious how you deal with the fact that lighting of the image will not be equal through the picture?

Sirius Glass · Dec 12, 2018

Alan Edward Klein said:
Curious how you deal with the fact that lighting of the image will not be equal through the picture?

Uh, use a light meter.

correct verticals with a wide angle !

Privacy & Transparency

Privacy & Transparency