Jump to content

The 50 mm as standard - how it came to be


Recommended Posts

That said, the camera was not exactly in the center between the speakers and normal to the wall.

As I've said several times this is not caused by the camera lens. What you see in those images shot at 50mm is what I see between the two speakers. It doesn't matter if the camera was centered or not. I shot them separately straight on keeping both speakers perpendicular to the lens by moving latterly along the floor tiles that run parallel in front of the speakers. I did not stand in one place and turret from a center position between the speakers.

 

I believe I shot both speakers correctly to mimic the rigid straightness and perpendicularity I employed with a graphics camera.

 

I'll repeat again. This is an optical illusion created by viewing similar objects in different surround framing.

Link to comment
Share on other sites

  • Replies 96
  • Created
  • Last Reply

Top Posters In This Topic

Back in the day when I "managed" the photo department in an Photo / Electronic Store (That because I was the employee, ex. possibly the store manager who knew something about photography) 35mm Nikon, Pentax, Canon, Hanimex, etc., etc. came with a 50mm lens. Anything else was a special order. Even the medium format was pretty much the same, came with the standard lens for the format. This was a pretty decent store with a reasonable inventory. No Leicas or Hassies, except special order. I only worked there for a few years, but can't recall a single order for just a camera body, and we wouldn't have sold one that way, because we'd have been stuck with the lens.

 

Kits seem to be popular now.

 

When I bought my FM in 1979, bodies and lenses were priced and sold separately.

 

Some of my dorm friends had recently bought their FM with the 50/1.4, but I decided for the 35/2.0

As well as I remember, the prices were pretty close.

 

Indoors, I mostly used flash, though sometimes (especially in museums) would handhold at low

shutter speeds.

-- glen

Link to comment
Share on other sites

I tend to think of "normal" as a line drawn from a point, perpendicular to a plane ;) In statistics, to "normalize" is to group data into non-overlapping categories, expressed as a ratio relative to the modal value. In digital audio, "normal" is the loudest level without distortion, except when recording, where "normal" is about 20 dB below maximum. In education, "normal" is a curriculum intended to produce teachers. By definition, "normal" is conventional, something expected, or average. That usage allows for considerable flexibility.

Not to beat this to death, but I think in photography, "Normal" in terms of lenses, is a term of art. One that refers to lenses that roughly approximates perspective as the eye sees it, ie, "normal" vision. Its not totally accurate scientifically, but it doesn't need to be to convey the idea.

  • Like 1
Link to comment
Share on other sites

Not to beat this to death, but I think in photography, "Normal" in terms of lenses, is a term of art. One that refers to lenses that roughly approximates perspective as the eye sees it, ie, "normal" vision. Its not totally accurate scientifically, but it doesn't need to be to convey the idea.

 

The only plausible explanation I can see for this idea, "perspective as the eye sees it," is the occasional contention that people "tend to view prints from a distance roughly equal to the diagonal of the print." If this is true then it makes the case that a "normal" lens ought to have a focal length roughly equivalent to the diagonal of the film/sensor.

 

This follows from the principles that Kingslake discusses, and I don't see how it can be otherwise.

Link to comment
Share on other sites

Bill I've read that an actual "normal" perspective as the eye sees it is around 40 mm in 35mm film terms, actually 39+, I've also read it is a 43 mm. All I'm saying is the industry has somehow over the years picked the 50mm lens to be the standard "normal" lens. And it roughly works.

 

Here's from Wikipedia.

 

"In
and
, a
normal lens
is a
that reproduces a
that appears "natural" to a human observer.[
]

 

However, to find a photographic lens equivalent to the human eye, which has an effective focal length of approximately 17mm,
is problematic due to the nature of human binocular
, being mediated and processed by the cortex, and because of the structure of the
which has a concave
, rather than a flat sensor, with variable sensitivity and resolution across its wider-than-180° horizontal field-of-view.[
]

 

A normal lens then, is one that renders a printed (or otherwise displayed) photograph of a scene that when held at 'normal' viewing distance (usually arms-length) in front of the original scene and viewed with one eye, matches the real-world and the rendered perspective.

 

A normal lens typically has an angle of view that is close to one
(~57.296˚) of the optical system's
. For
format (24 x 36 mm), with an
image circle diameter equal to the diagonal of the frame (43.266 mm), the focal length that has an angle of one radian of the escribed circle is 39.6 mm; the focal length that has an angle of one radian of the horizontally-bound
image circle, is 33 mm; the focal length that has an angle of one radian of the vertically-bound inscribed circle, is 22 mm. This correlates with the popularity of 35 and 24 mm lenses, and the existence of 40 mm lenses, albeit the latter in a more restrained offer. A 50 mm lens has a vertical-bound inscribed circle angle of view of ~0.5 radians. A 70 mm focal length (typically only available in zoom lenses) has a horizontally-bound inscribed circle angle of view of ~0.5 radians. An 85 mm lens has an escribed (frame diagonal) circle angle of view of ~0.5 radians. Effectively, the 24, 35 and 40 mm trio have a 1:2 relation to the 50, 70 and 85 trio of focal lengths. "Normal" lenses, those that cover one radian in at least one of their inscribed or escribed image circles, belong to the first group, with 35 and 40 mm lenses closer to one radian than 50mm lenses."

 

Note: that the Wikipedia has remarks in the mast of the article saying that there may be inaccuracies.

 

All I'm saying is lens makers have generally, for whatever reason, chosen to call the 50 as a so-called normal lens. And it roughly, very roughly it is. Its a term of usage. Part of it is possibly a 40 - 50 mm is easier to correct in terms of the expense of the glass elements needed to translate the image to a flat plane. But I really don't need to discuss lens design. When someone asks me if I have a normal lens they can borrow, I will hand them a 50mm and they won't go on a screed as to how its not actually normal. Here's a link to an article in the Atlantic that discusses how the 50mm lens became identified as "normal" by manufacturers.

How the 50-mm Camera Lens Became ‘Normal’ - The Atlantic The article branches off into some interesting areas and presents perhaps a more modern view of the whole idea of "normal". Let me know if the link doesn't work. Thanks.

Edited by Uhooru
Link to comment
Share on other sites

Here's a link to an article in the Atlantic that discusses how the 50mm lens became identified as "normal" by manufacturers.

 

Thanks, I've previously read that. I didn't think it was that authoritative, and as I recall there was some discrepancy with respect to the "normal" focal length used for motion pictures. (I'll reread if you'd like me to critique it more seriously, but otherwise ...)

 

The earliest reference that I know of for a "normal" focal length, although the word "normal" was not used, was by C.E.K. Mees in his 1921 book "Fundamentals of Photography." He says, "It is a good rule to secure a lens which has a focal length at least equal to the diagonal of the film. A little more focal length is still better." This would work out to roughly a 50mm focal length on a standard 35mm camera frame.

 

Mees doesn't state why this is so (at least I don't notice that he says so), but he does comment on both longer and shorter focal lengths, in both cases saying that each lens "represents the perspective that the eye saw from the viewpoint of the lens."

 

In other words, Mees has pointed out that each of the lenses gives the same perspective view that the human eye would have from the same position. I don't think that anyone who has considered this carefully would argue with this. Yet only one of these focal lengths, the one that approximates the diagonal of the film, was thought to be the "good rule" by Mees.

 

Now I'm sure there are more explanations in literature as to why such a focal length is preferable, but one I can put my finger on right away is C.B. Neblette's 1952 edition (I finally got a copy, Conrad) "Photography; It's Materials and Processes." He says, "In general, it is desirable to view the final picture from its center of perspective so that the view may have the proportions of the original." (The "center of perspective" is discussed in the Kingslake reference I gave earlier.)

 

I think if one puts this little string of ideas together, essentially that we should view a print from its "center of perspective," and that any lens gives the same perspective that the eye sees from the same position, and finally, the somewhat questionable idea that people tend to view a photo from a distance roughly equal to the diagonal of the image, I think the obvious conclusion is that, lacking some other priority, one should generally use a focal length roughly equal to the diagonal of the film/sensor as a "normal" lens. (I would say that 50mm for a standard 35mm negative is close enough to fit into this description.)

 

Anyway, in my view it all comes down the normal print viewing distance as the basis for the "normal" focal length.

 

If people don't recognize names, Mees started the Kodak Research Labs, Kingslake spent years as Kodak director of optical design, and Neblette headed up RIT's photographic technology program for many years. I'm using these people as reference rather than argue the case myself because this is the internet, after all, and by using them I can probably escape from the thread easier (I've been there before, as many others have also, I'm sure).

Link to comment
Share on other sites

I read that Wikipedia piece too, but I don't like it much, especially the paragraph with all the numbers in it. One radian seems to be just a magic number grabbed out of the air, and all the inscribed and escribed circles are just juggling to complicate what's basically a simple idea. Also, I don't look at photographs by holding them at arms length with one eye shut.

I'd mark the Wikipedia writer down again, for his/her five significant figures of precision in converting that one radian to degrees, and in giving the frame diagonal.

 

I think the angle of view of a normal lens is roughly what we see clearly without turning our heads: Wikipedia's first sentence is ok. Much more justification than that is inappropriate, because we're comparing apples to pears; as the article itself says, we have two eyes (if we're lucky), and each one moves all the time. Our image never sits on a flat screen; it's constructed in memory from what the 'good bit' in the centres of our two retinas saw last in each direction; and it includes interpretation of the double image as 3D. It's really nothing like a camera, so one sentence of comparison was enough.

 

Incidentally, on the term 'normal lens', I grew up hearing 'standard lens'. I think I started coming across 'normal lens' in discussions on the internet that involved more than one format.

Link to comment
Share on other sites

I had forgotten S. Ray's "Applied Photographic Optics."

(See https://www.amazon.com/Applied-Photographic-Optics-HonFBIPP-HonFRPS/dp/0240515404/ref=sr_1_1?ie=UTF8&qid=1531553105&sr=8-1&keywords=applied+photographic+optics)

 

It turns out that he has a short chapter on the Standard Lens where he discusses the basic issues, much as what I'm saying, but referring to two other chapters for background. He also allows as how there is some variability in the focal length. He says that the 50mm as standard originated with the Leica around 1926 (I think).

Edited by Bill C
Link to comment
Share on other sites

My personal definition of normal focal length would be the length at which I take most of my pictures, which has changed as I age. I was perfectly happy with the 50mm I used on my Retina for most pictures grew to prefer the 35mm on my Nikon L35AF and now like the 28mm on my Sigma. I find the iPhone a little too wide and wind up cropping most of the pictures it takes.
Link to comment
Share on other sites

There is Data and then there is Lore: All lenses project a circular image. Masking in the camera crops this illuminated circle into a square or rectangle. In any event, the longest dimension of the image is the corner to corner dimension. Add to this, all lenses vignette. The usable portion of the circular image is called the “circle of good definition”. The diameter of that circle is approximately the same as the focal length. This is because the lens is essentially a portion of a sphere. In any event the image forming rays arrive square at the center of the image and oblique at the edges. The oblique rays must travel further and because of the angle of attach, the circles of confusion arrive as ellipsis’s. In other words, the bounders of the circle of good diffusion are due to angle of incidence of the image forming rays. To make a lens with a larger circle of good definition requires a longer back-focus. This is accomplished by shifting the rear nodal forward. This design is called “retro-focus” somewhat akin to looking through binoculars backwards. Thus a lens with a focal length about equal to the diagonal measure is considered to be “normal”.

 

As to matching the human eye perspective: Such a match is easy to accomplish. A picture is taken; the focal length is of no consequence. A contact print is made from that negative. The print is then viewed from a distance equal to the focal length of the taking lens. The image observes matches the “human perspective”. If a miniature camera is used it likely sports a short lens, say 50mm. Viewing such a contact print from a distance of 50mm (2 inches) is impossible without optical aid. Besides miniature films are enlarged for viewing. Now the viewing distance is the focal length of the taking lens multiplied by degree of magnification. As an example: A 35mm negative is enlarged to make an 8X10 inch print. The degree of magnification is about 8X. A 50mm lens was mounted. The viewing distance to view this print as see the “human perspective” is 50mm X 8 = 400mm = 16 inches. This is about our normal reading distance. Thus the “normal” for any camera is that focal length that corresponds to what magnification is applied and the viewing distance.

 

Suppose a 120 square format camera is used. The diagonal measure is about 80mm. To make an 8X10 the enlarger magnification is about 4.5X. Thus the viewing distance is 80mm X 4.5 = 360mm = 14 inches.

 

I say “normal” for any format is a focal length about equal to the diagonal measure of the format.

 

As to Edison and 35mm film: Edison and his chief engineer negotiated with Kodak for film. Kodak was making long rolls of 70mm for the Brownie camera. Edson purchased 70mm long rolls and sit them down the middle gaining two rolls of 35mm for the price of one. The 35mm film was punched with sprocket holes to help transport film in camera and projector. The usable space between the sprocket holes was 24mm. Edison’s chief engineer set the image size as 24mm wide by 18mm height. This remained the cine standard for many years.

 

The Litz Optical Works designed the Leica camera in 1913. It came to market in 1924. The camera uses cine film 35mm however its chief orientation was horizontal. Oskar Barnack, chief engineer set the height of the image at 24mm. He doubled Edison’s format height making this dimension (36mm) the length. Thus the Lieica’s frame size was set at 24mm height by 36mm length. The diagonal measure of this rectangle is 44mm. This value was rounded up to 50mm and 50mm lenses became standard for this format.

  • Like 5
Link to comment
Share on other sites

A cube with one face parallel to the film plane will be a square when rendered by a rectilinear lens, whether it is in the center, side or corner. Perspective of the visible sides will be distorted, but not in an obvious manner. Spherical objects are noticeably distorted into egg-shapes toward the edges. When viewed at a distance proportional to the focal length and magnification, this distortion is eliminated. Landscapes do not seem distorted because they are not composed of symmetrical objects easily recognized as being distorted. Trees and such do appear fattened when viewed with a practices eye.
Link to comment
Share on other sites

A cube with one face parallel to the film plane will be a square when rendered by a rectilinear lens, whether it is in the center, side or corner. Perspective of the visible sides will be distorted, but not in an obvious manner. Spherical objects are noticeably distorted into egg-shapes toward the edges. When viewed at a distance proportional to the focal length and magnification, this distortion is eliminated. Landscapes do not seem distorted because they are not composed of symmetrical objects easily recognized as being distorted. Trees and such do appear fattened when viewed with a practices eye.

The biggest problem is cosign error. Looking back at the lens from the center, you would see a perfect illuminated circle. Looking back at the lens from the edge of the frame you would see an ellipse. The ellipse has reduced surface area so the image produced is reduced in density. Lenses shorter than the diagonal measure tend to vignette more severely.

Link to comment
Share on other sites

Distortion due to rectilinear rendering is related to the angle of incidence on the sensor (cosine), but is distinct from illuminance (vignetting), which is related to both the angle of incidence at the sensor and narrowing of the exit pupil when seen off-axis. The apparent distortion of spherical objects is explained on most intermediate or advanced books on photography (e.g., The Camera, by Ansel Adams), or the topic of perspective in mechanical drawing.
Link to comment
Share on other sites

  • 4 weeks later...

I read the Atlantic article and found it a little odd. For one thing, why would editors at this magazine care about such an esoteric and obsolete topic?

The article neglects to mention the economics of the 50mm lens. They can be made cheaply, and are fast and reasonably sharp. In the 70s, when 35mm photography became a mass-market hobby, the salesmen at the camera stores always tried to sell you a 50, or they bundled it for a deal.

 

However, whatever the merits of the 50, I still find it the most boring of all lens choices. Too short for portraiture, too long for scenery. And I find the "normal" field of view of the medium format 75 or 80mm lenses to be the same: B-O-R-I-N-G.

Link to comment
Share on other sites

  • 3 weeks later...

An article in "Outdoor photography" explains the normal lens as, first, that it should be the diagonal measure of the frame, which is 43mm for a 24x36mm frame.

 

But that when Leica popularized the 35mm format, they did it with a 50mm lens.

I suppose no-one knows why now, but it seems to have stuck.

 

The article then says that 35mm to 60mm are within the normal range, which includes my always favorite 35mm lens.

 

But if Leica had started with 43mm, then we would all think today that 43mm was a normal lens!

-- glen

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...