The 50 mm as standard - how it came to be

You might find this interesting.

How the 50-mm Camera Lens Became ‘Normal’ - The Atlantic

As long as the 50mm is designed and built as a flat field lens it can pretty much show the non-distorted look of reality or WYSIWYG unlike wide angle lenses.

 

The Argyll graphics camera I used in the '80's had a 55mm fixed prime lens where I had to enlarge and reproduce a client's geometrically designed logo so it lined up perfectly square to the demarcation lines where I lined up cut to fit RC paper dead center on the exposure plane so as not to waste paper which was expensive back then.

 

All 90 degree corners of graphic elements of lines and geometric shapes lined up to an actual T-square and ruler after developing the RC paper.

There was no lens profile correction software back then.

 

Unfortunately the 50mm setting on my DSLR's kit lens and film legacy zooms and one 50mm prime don't give exact WYSIWYG what I see through the viewfinder. What looks like it would match looking through the viewfinder and then up at the actual scene doesn't get recorded exactly what I saw when I view the Raw file in ACR. Objects dead center of the frame through the viewfinder and reality tend to look smaller in ACR.

Unfortunately the 50mm setting on my DSLR's kit lens and film legacy zooms and one 50mm prime don't give exact WYSIWYG what I see through the viewfinder.

Is it the lens, is it the viewfinder, or both?

 

Lenses for general photography are designed to different standards than a process lens. The first emphasizes speed (f/stop) and convenience (size), whereas the latter for rectilinear accuracy and flatness of field. At my newspaper, the "graphics" camera was a Crown Graphic on an overhead rail, with a Goetz Apochromat, with a maximum aperture of f/6.3. The viewfinder was a ground glass in the film plane.

 

A process lens is also designed to work best near 1:1 magnification. At that distance and extension, the FOV is much wider than the same lens at infinity.

 

Accuracy of a DSLR viewfinder depends on perfect alignment of the mirror and size of the ground glass. Furthermore, the GG in a DSLR has features to concentrate the light toward the viewpoint, which may introduce distortion.

 

I suspect an EVF is about as true to the film plane as you can get. Even a ground glass is not guaranteed to have the same field of view as film in a holder, unless calibrated. That said, a good lens often has linear distortion on the order of 1%, which probably wouldn't wash as a process lens, nor for aerial mapping.

What if I'm using my 50mm lens on a Nikon D300. Is it still a normal lens?

What if I'm using my 50mm lens on a Nikon D300. Is it still a normal lens?

X 1.5 for DX if everything else is Normal. :D

What if I'm using my 50mm lens on a Nikon D300. Is it still a normal lens?

Yes and no. as far as the way 50mm treats foreshortening and depth of field, it is still a 50mm, but as far as coverage in frame it is cropped to a 75 on your D300. Its like you take a picture with a 50mm on a full frame camera, but then crop to the area of a 75mm. The major difference from cropping a full frame and a crop frame camera is the D300 will pack all its pixels into the frame, whereas cropping a full frame 50mm will decrease the amount of pixels in the picture. Why called normal? Photographers decided back when that the 50mm roughly sees things like the eye does in terms of foreshortening, etc. You can look it up, but I think the actual "normal" is like 43mm or something, but roughly 50 is an equivalent.

I am mistaken about the FOV of a process lens, confusing coverage with FOV. The Goetz lens would barely cover 4x5 at 1:1 and vignette badly at conventional distances. That said, the more the lens is extended, the smaller the FOV.

 

A "normal" lens for DX would be 35 (52 mm equivalent) or 28 (42 mm equivalent). That was then (the Day). Now I rarely use a 50 mm prime, tending to 35 mm or more often, 25 mm for unplanned photography. Statistically, about 50% of my travel pictures, mostly landscapes, are taken at 50 mm.

 

Part of the "50 mm" rule concerns perspective in a print. A rectilinear lens distorts perspective of 3D objects (spheres, heads) toward the edges of the frame, unless viewed from a distance proportional to the focal length and magnification. For an 8x10" print, that would be about 16" - arm's length. Do that, and egg-shaped heads look round and normal.

Edited July 2, 2018 by Ed_Ingold

A "normal" lens for DX would be 35 (52 mm equivalent) or 28 (42 mm equivalent).

 

I think of this as being the reason for existence of the great little inexpensive Nikkor 35mm f/1.8. It's a $200 lens that I think everyone with a Nikon DX DSLR should have.

 

BTW, as far as a "flat field" normal lens goes-often the 50-60mm macro/micro lenses fit this criteria quite well. It's also why enlarger lenses at least use to be somewhat common for high magnification work on bellows-they are flat field also.

I am mistaken about the FOV of a process lens, confusing coverage with FOV. The Goetz lens would barely cover 4x5 at 1:1 and vignette badly at conventional distances. That said, the more the lens is extended, the smaller the FOV.

 

A "normal" lens for DX would be 35 (52 mm equivalent) or 28 (42 mm equivalent).

 

Part of the "50 mm" rule concerns perspective in a print. A rectilinear lens distorts perspective of 3D objects (spheres, heads) toward the edges of the frame, unless viewed from a distance proportional to the focal length and magnification. For an 8x10" print, that would be about 16" - arm's length. Do that, and egg-shaped heads look round and normal.

Actually, Its tricky. A 35mm in DX covers the area of a 50mm, but it still has the same fore-shortening and characteristics of a 35mm (in full frame). The crop factor doesn't change the nature of the lens in terms of angles etc. So a 35mm lens on a crop camera is really still a 35mm lens.

Actually, Its tricky. A 35mm in DX covers the area of a 50mm, but it still has the same fore-shortening and characteristics of a 35mm (in full frame). The crop factor doesn't change the nature of the lens in terms of angles etc. So a 35mm lens on a crop camera is really still a 35mm lens.

The perspective (foreshortening) is determined by the distance, not the focal length. Distance being constant, an FX camera with a 50 mm lens is the same as FX cropped to 35 mm, is the same as 35 mm on a DX camera. The first has a wider FOV than the other two, which are identical (within roundoff error).

35 on a DX has the same area coverage as a 50, but it is still a 35mm lens, only the area is cropped. It has all the characteristics of a modest wide angle lens, not a normal lens, and the 35 will exhibit a different Depth of field scale on a crop camera than a 50 will have on a full frame. The perspective part I think you are correct there. But take the same shot with two cameras at the same distance. 35 on a DX and 50 on a FX and I bet you will see 2 different looks.

FWIW, the author of the piece is

Allain Daigle is a doctoral student in media, cinema, and digital studies at the University of Wisconsin at Milwaukee.

 

I remain very skeptical of the 'historic' or 'tradition' argument here as I do of similar other 'social' interpretations by postmodernists and the like. The 50mm is only for 35mm, of course, and there were empirical reasons why some views were and are seen as more approximating human vision.

 

Look through the viewfinder on a Leica M3 with one eye and keep the other eye open.....

Edited July 3, 2018 by JDMvW

But take the same shot with two cameras at the same distance. 35 on a DX and 50 on a FX and I bet you will see 2 different looks.

The apparent depth of field at the same relative aperture will differ - greater for the DX sensor, but the field of view and perspective will be identical. DOF is inversely proportional to the absolute magnification at the film plane. A cropped image using a 35 mm lens on an FX camera, and an uncropped DX image with the same lens, aperture and distance, will be identical in all respects.

 

The actual field of high resolution view for the human eye is defined by the size of the fovea centralis, which subtends an angle of about 2 degrees. You can see light/dark and motion outside this area, but in order to read a book, for example, the eye most move continually. Any correlation between visual FOV and that of a 50 mm lens is contrived.

No they won't be exactly the same, because the 50 mm will compress the image more than the 35 will and that is a real difference besides just DOF. I think normal doesn't technically and scientifically mean the same exact as the eye, actually most see through 2 eyes and of course the FOV will differ, but the way the image is compressed etc., is roughly similar. This isn't a solve for a physics problem, normal is a basic inexact rule of thumb.

No they won't be exactly the same, because the 50 mm will compress the image more than the 35 will and that is a real difference besides just DOF.

 

Perspective, the underlying principle of depth compression, is a function of distance only. Longer lenses "compress" only when the distance is changed so that the subject is the same relative size on the film plane. This principle is explained in practically every book on photography, including "The Camera" by Ansel Adama. In the citation below, see the example where images taken from the same distance with a 28, 35 and 50 mm lens are cropped to display the same field of view. There are other illustrations of "compression," where the distance is altered proportional to the focal length used.

 

Perspective distortion (photography) - Wikipedia

Perspective, the underlying principle of depth compression, is a function of distance only. Longer lenses "compress" only when the distance is changed so that the subject is the same relative size on the film plane. This principle is explained in practically every book on photography, including "The Camera" by Ansel Adama. In the citation below, see the example where images taken from the same distance with a 28, 35 and 50 mm lens are cropped to display the same field of view. There are other illustrations of "compression," where the distance is altered proportional to the focal length used.

 

Perspective distortion (photography) - Wikipedia

Look at the 3 photos with 3 different lenses from 18 mm to 55mm from the Wiki article you cited.

"How focal length affects perspective: 18mm (wide-angle), 34mm (normal), and 55mm (modest telephoto) at identical field size achieved by different camera-subject distances. Notice that the shorter the focal length and the wider the angle of view, perspective distortion and size differences change."

This is from the Wiki article you cited. Is this different than what you are talking about. I had to paste the text that accompanied these photos separately. I didn't see an example that matches what you are talking about Ed. Just trying to figure this out. Isn't the "different camera-subject distances used to achieve the same field size, effectively the same as by cropping?

I think what you're seeing Uhooru is how space is compressed using longer focal lengths, which causes the illusion that the object in the background is registering as different sizes at different focal lengths. None of what you're seeing through a camera lens actually looks like it appears to look. Remember, a one lens setup cannot give accurate depth of field or perspective. That's why almost all mammals have two eyes.

 

Also, each lens by each manufacturer is a little different in terms of what it "sees". I found that out tonight as I was making some tests with two 135 lenses to see which one I wanted to keep. I was quite surprised to see what I saw in the viewfinder (and on an SLR, what I saw was what the lens was actually recording). Placing the camera at exactly the same distance from a stationary subject, one lens recorded the subject as being quite a bit further away than the other lens! Has anyone else seen this sort of thing?

 

If it happens in 135mm lenses, it surely happens at other focal lengths, including 50's. All I can guess at is that some lens makers, for example, call a 47mm lens a 50, or the other way 'round as well.

Edited July 4, 2018 by steve_mareno|1

I think what you're seeing Uhooru is how space is compressed using longer focal lengths, which causes the illusion that the object in the background is registering as different sizes at different focal lengths. None of what you're seeing through a camera lens actually looks like it appears to look. Remember, a one lens setup cannot give accurate depth of field or perspective. That's why almost all mammals have two eyes.

 

Yep, understand all that and that is basically what I'm saying. Never said a lens actually depicts things the way the eyes said. What I'm talking about is how a 50mm on so called "full frame" 135 camera presents differently than a 35 on a APC or Dx body.

Look at the 3 photos with 3 different lenses from 18 mm to 55mm from the Wiki article you cited.

 

This is the second example, where the distance to the pink bottle is adjusted so that it appears the same size for each focal length. The compression (or expansion) is solely due to the distance, not the focal length.

 

In the first example (the garden), the distance is the same. The total field of view changes, but the angles and proportions of the subject do not. This is true even if the format changes as well (e.g., FX to DX) - the proportions stay the same for any object in the image, regardless of the focal length or cropping factor.

 

A better example would be to photograph geometric shape like a cube, obliquely, from the same distance with various focal lengths and cropping factors, then crop to the same field of view and display at the same size. You would see no difference in shape or relative sizes of near and far objects. Cropping and enlargement doesn't change the shape of an image, only the field of view. In fact, your eye sees exactly the same thing, viewed from the same distance as the camera. You just have to learn to "see" objectively like an artist or draftsman.

 

It is considered bad form to present images you don't own in this forum. That's why I only posted a link.

Also, each lens by each manufacturer is a little different in terms of what it "sees". I found that out tonight as I was making some tests with two 135 lenses to see which one I wanted to keep. I was quite surprised to see what I saw in the viewfinder (and on an SLR, what I saw was what the lens was actually recording). Placing the camera at exactly the same distance from a stationary subject, one lens recorded the subject as being quite a bit further away than the other lens! Has anyone else seen this sort of thing?

Focal lengths are approximate. Zeiss, for example rounds focal lengths near 24 mm to 25, consistent across their product lines. Telephoto lenses often have internal focusing, which works by altering the focal length. The size of an object may change, but its proportions stay the same if taken from the same distance.

 

Binocular vision is irrelevant. Each eye sees the object from a slightly different angle, giving a sense of depth. The principles of perspective are unaltered. Some animals (e.g., anoles) don't have an overlapping field of vision. They sense distance by bobbing their heads up and down. Your homework assignment is to analyze how that works for the lizard ;)

Well excuse my bad form and you are right to point it out, so errr pointedly. But you are right, not supposed to do it.

Where in the article in Wikipedia does it actually say what you are? Don't think it does, but I will empirically try to test this tomorrow.

So tried the experiment today, and I think Ed is basically correct. For some reason, P.net is not allowing me to post the images, but when that changes I will. I'm not sure they're exact equivalencies, but they look pretty similar. I used a 14mm and a 50 mm on a crop lens camera. Taken of the same thing from the same position. I then cropped the 14mm to match the view area of the 50mm, and while there is a little difference, they look pretty close in terms of angles. The main difference being DOF. Both taken at f11, ISO 400. But it will be better when I post the images so you can see for yourselves and also let me know if the methodology is flawed. I did it pretty simply.

take the same shot with two cameras at the same distance. 35 on a DX and 50 on a FX and I bet you will see 2 different looks.

 

I've never tried the experiment but I'd bet twenty dollars that if the distance to the subject is unchanged, and the field of view is the same for both setups, the images will be indistinguishable in the terms you have in mind.

 

EDIT: Sorry, I wrote before seeing your post above.

No worries.

I've never tried the experiment but I'd bet twenty dollars that if the distance to the subject is unchanged, and the field of view is the same for both setups, the images will be indistinguishable in the terms you have in mind.

Taken from the same distance, perspective of an object, including its apparent size relative to objects nearer or further away, is unchanged. Field of view, focal length and sensor size are irrelevant.

 

In a practical sense, we "zoom with our feet" in order to change the perspective. Moving closer to the subject makes the subject larger with respect to things in the background. A wide field of view keeps more of the background. A narrow field of view (e.g., telephoto) does exactly the opposite, emphasizing things in the background (landscapes), or rendering them out of focus (closeups). A "normal" lens is a compromise.