Distortion with extreme rise?

[Cor]

In another thread (http://www.largeformatphotography.info/forum/showthread.php?t=69485) Oren Grad mentioned:


Extreme front shift, just as extreme rise/fall, will have obvious effects on the geometry of most scenes.

Rather than hijacking that thread I would like to know what happens when you apply extreme (what is extreme?) raise to the image.

Recently I made an image on 4*5 Fp4+ of a building with my 90mm Super Angulon on my Linhof Technika III. I positioned the back vertical and square with the building, but I needed a rise as well to include the top part. The result looks fine to me, but what distortion would happen if one uses more rise, or perhaps I missed noticing that distortion?

thanks,

Best,

Cor

[Peter K]

Cor, distortion is an abberation of a lens e. g. the barrel distortion of a fish-eye lens with it's smaller magnification at the edge compared to the center.

But with a well-corrected lens like your Super-Angulon you can get a strange perspective if the point of view is to close to the subject together with rising the lens to the border of the image circle.

On the other hand also video-cams are equipped with short-focal lenses so this kind of perspective is more common today.

Peter

[Cor]

Cor, distortion is an abberation of a lens e. g. the barrel distortion of a fish-eye lens with it's smaller magnification at the edge compared to the center.

But with a well-corrected lens like your Super-Angulon you can get a strange perspective if the point of view is to close to the subject together with rising the lens to the border of the image circle.

On the other hand also video-cams are equipped with short-focal lenses so this kind of perspective is more common today.

Peter

Thanks Peter,

In this case I was about 75 meters away from the building, so I guess the distortion is not/hardly there. But suppose I was 5 meters away from the building, back vertical and aligned with the building, and I would apply maximum rise. Would the resulting image show a "bulging" effect as in your fish eye example? Less extreme though, I guess?

[Jack Dahlgren]

Thanks Peter,

In this case I was about 75 meters away from the building, so I guess the distortion is not/hardly there. But suppose I was 5 meters away from the building, back vertical and aligned with the building, and I would apply maximum rise. Would the resulting image show a "bulging" effect as in your fish eye example? Less extreme though, I guess?

Take your camera outside and check it out on the ground glass. Bring a ruler if you want to check if the lines are still straight.

[Leonard Evens]

I strongly doubt that your 90 mm Super Angulon would produce any visible distortion within the circle of coverage. But when the center of perspective is far off from the center of the image, which will happen with a large rise, the image may look strange to your eye. In principle, you should view the print or other final image so that your eye is in the same relative position to the image as the lens was to the scene. So, for example, if you were looking at an 8 x 10 print, your eye should be at distance 180 mm = 2 X 90 mm from the print and in line with the vanishing point for paralelel lines in the scene leading directly away from the lens. With a wide angle lens, one seldom places one's eye close enough to the print, and that, in itself, produces an apparent distortion, but the displacement from the center of perspective exaggerates the effect.

[Oren Grad]

The problem isn't linear distortion, it's that objects or portions of objects (for example, buildings) toward the edge of the frame will look unnaturally stretched. Think of what a rectilinear ultrawide lens on a 35mm camera does toward the edges of the picture; it's the same effect here.

[Andrew]

The problem isn't linear distortion, it's that objects or portions of objects (for example, buildings) toward the edge of the frame will look unnaturally stretched. Think of what a rectilinear ultrawide lens on a 35mm camera does toward the edges of the picture; it's the same effect here.


If I read you correctly, you're refering to extreme movements [esp in architectural work] being used to remove all convergance?

I've noticed this in many images and I'm seriously doubting the need to make all lines truely parallel because we're so used to seeing convergance every day that suddenly removing it looks unnatural.

[John NYC]

I've noticed this in many images and I'm seriously doubting the need to make all lines truely parallel because we're so used to seeing convergance every day that suddenly removing it looks unnatural.

It is traditional to remove all convergence. And not to be questioned.

Just kidding, but that is what some of the architectural photographers will say here.

I see this distortion of rectilinear and cylindrical shares (NYC water towers are my biggest issue in this regard since I do roof shots here in the city). It isn't something you can get away from, I don't believe, if you just have to have the rise. Maybe if you used a longer lens with a HUGE image circle it might be less noticeable.

[Oren Grad]

If I read you correctly, you're refering to extreme movements [esp in architectural work] being used to remove all convergance?

That's right.


I've noticed this in many images and I'm seriously doubting the need to make all lines truely parallel because we're so used to seeing convergance every day that suddenly removing it looks unnatural.

Yes. Both extremes look weird, just in different ways. If you're making pictures for yourself, do whatever works for you - there's no obligation to strike any particular balance between the buildings-falling-over look at one extreme and the buildings-stretched-like-silly-putty look at the other.

[ic-racer]

Harold M. Merklinger explains the details of this effect in chapter 5 (page 43) Perspective and Distortion ( http://www.trenholm.org/hmmerk/#FVC )

[Kirk Gittings]

Just kidding, but that is what some of the architectural photographers will say here.
Yes I will say it, I've been making my living at AF since 1978 and fully correcting convergence is expected (unless you are obviously doing an up or down view). It is a visual convention of the genre even though we all know that the eye doesn't see architecture that way. I think of it like B&W photography. When I first got started seriously around 1970. B&W was seen as more real than color, because b&w was the convention in newspaper photography.

[Brian Ellis]

If I read you correctly, you're refering to extreme movements [esp in architectural work] being used to remove all convergance?

I've noticed this in many images and I'm seriously doubting the need to make all lines truely parallel because we're so used to seeing convergance every day that suddenly removing it looks unnatural.

[marathon_racer]

I've noticed distortion of circular shapes near the edges of the frame, but also what you've mentioned, if I am very close to an object and closer to the ground with a wide angle lens. I think this creates the look of strong perspective, so effective at dramatizing structures at the expense of verisimilitude. It makes objects and buildings seem taller and narrower.