About the 50mm "Normal" Lens

[VidThreeNorth]

"How the 50-mm Lens Became 'Normal'"
Allain Daigle, May 13, 2018
"How the 50-mm Lens Became ‘Normal’"

 

[VidThreeNorth]

This article seems a bit tightly "focussed" on the 50mm and its equivalents. A fuller coverage of the topic should show that there have been other opinions about what is a "normal" lens.

Just after the SLR took over from the range finder cameras (around the time of the Pentax Spotmatic II) the range finder manufacturers found a market for smaller 35mm zone focus and range finder cameras. I mentioned this in another topic. The main brands were Konica, Minolta, Canon, and maybe Miranda and Ricoh. I think the lower priced zone focus cameras were mainly around 38mm, F2.8 and the more expensive range finder versions tended to be around "40mm, F2.0"-ish. I wrote before that the range finders were 38mm, F2.8, but I am fairly sure now that those lenses were more often on the zone focus versions.

But the point is that around that time, there was some discussion about whether the 38mm - 40mm lenses better represented what the human eye saw. If you study the human eye and perception it is hard to say.
[2018-05-28 edited for clarity]

The retina is very wide, maybe like a 20mm lens coverage, but what we are mainly "looking at" is around the coverage of the "fovea" which is very narrow. So there was also a counter argument that because we are really looking primarily at this narrow are, that a short telephoto is a better "normal" lens.
[2018-05-28 edited for clarity]

The Wide School:

My Panasonic Lumix GF-3 (Micro 4:3) came with a 14mm F2.8 lens (35mm equivalent view 28mm) and many people felt that it was a good viable day to day lens.
[2018-06-20 clarified lens type and 35mm equivalent view]

I think the Apple iPhones have generally been 35mm camera equivalent 28mm lenses (someone might correct me on that?). Other phones have other focal length lenses. I never really checked, but it seemed to me that they looked closer to that "moderate" wide angle field of view.

GoPro action cameras have tended to have "170 degree fish-eye" lenses. Do people using those cameras consider that a "normal" lens?

The Tele School:

When the Yi-M1 (Micro 4:3 format) came out the lens selection was either a 12-40mm zoom (35mm camera equivalent 24 - 80mm) or a 42.5mm (35mm camera equivalent 85mm) F1.8 "portrait - macro" lens. I have used the 42.5mm quite often and I find that for me, it is a viable "normal" lens.

[2018-05-28]
Here is Kai's video, posted Feb 18, 2018, wherein he says that 40mm is a good general use focal length. He does not use the term "normal" lens, but that seems to be the gist of it:

WARNING: I felt his humour was overly crude, but his reasoning is worth considering.

"5 Reasons 40mm is the Best Focal Length", "Kai W"
""

 

[Tim Tucker 2]

Kind of misses the well understood reason.

Perspective is assumed from the position of the observer. When you view an image from a position different from that from which it was taken you see a distortion of that perspective. Standard lenses do not approximate human vision but when viewing a standard print from a standard position in rough approximation you are viewing the print from the same position from which it was taken. Therefore you do not see any distortion of perspective. That is why it is standard because when viewing a standard sized print you see the same perspective as you did when you stood on the spot and took the photo.

However it really depends on how you view so sitting close to a large cinema screen your assumptions change and so does the focal length that gives the best approximation of no perspective distortion. But distortion may be your goal. It has been explored in cinema and photography since the beginning.

With the current variety of output media standard as a term is far less relevant and really dates back to when most images were viewed at a more standard size.

50mm was only ever a standard lens for the 35mm film format. I don’t think there was ever the confusion the author of the arrival has.

 

[Ysarex]

I find the article an unconvincing re-hash of unsupported claims. Trying to compare single lens single event "camera vision" with binocular over-time human perception is ultimately unsatisfying as they just don't compare well.

On the other hand: if your camera viewfinder has no magnification factor built-in (many do) and it's a 35mm camera, then a 55mm lens will allow you to look through the viewfinder with one eye and view the scene with the other eye and see the same thing -- no apparent change in image size. That's as good an argument for "normal" as any I've heard. The Pentax 67 was originally sold with a Pentaprism viewfinder that did not magnify one way or the other and Pentax sold the camera with a 105mm "normal" lens making the point that you didn't have to close the other eye to look through the viewfinder with that lens.

But what really makes senses is what Tim pointed out -- perspective. Normal should be whatever maintains perspective between viewing the scene and viewing the image. NOTE: Normal in this context should not have any kind of value attached. We use non-normal lenses all the time for good reason.

Unfortunately relying on perspective is complicated (as Tim likewise pointed out) by the fact that we don't have a standard metric for viewing the image. That shouldn't stop us. When we turn our photos loose for all to see we have no color management control over how all will view our photos. That shouldn't stop us from adopting proper color management practice. So how do we find a print viewing metric? Leslie Stroebel addressed this from the point of view of photographic artists and came up with a fairly convincing argument. If given the freedom to stand and view a print on the wall the average person will chose to stand back from the print a distance approx. 2x the long side of the print. So for example with 16x20 prints hanging in a gallery the average person will chose to comfortably view the prints standing back 40 inches. Obviously this falls apart with very tiny and very large images but it's the best we've got especially now with folks using everything from their phones to their giant TVs to view images.

If we adopt Strobel's suggested metric we can then do the math. Using a 35mm camera with the intent that the perspective as seen from the camera position is maintained through to the final print when viewed from a distance of 2x the long side of the print the lens focal length on the 35mm camera would be 75mm. In this case the definition of "normal lens" is -- this is what I saw from the camera position the way I saw it when I took the photo.

Joe

 

[ac12]

The other reason for the wider lens is that you have more D0F for a given aperture, so focusing is less of an issue.
Some early cameras were focus by guess. There was no rangefinder, you had to estimate/guess the distance and turn the lens to that distance to focus.
In the case of some of the older cameras, there was no ability to focus at all, so DoF was the only method to have subjects in focus. Every camera I had before my first SLR had a fixed lens (no ability to focus), essentially a box camera.

 

[Jeff15]

It is a very useful focal length...

 

[VidThreeNorth]

. . .
NOTE: Normal in this context should not have any kind of value attached. We use non-normal lenses all the time for good reason.
. . .

Reading the article it struck me that the whole idea of the "normal" view lens was nothing more than a Zeiss marketing idea. We perpetuated it in the still camera industry because it made it easier to sell cameras: "No, no, you don't have to think about what lens to buy, just take this here 'normal' lens and you'll be happy." There is a valid justification that it is a useful versatile lens that can cover a lot of situations. Really, the theoretically "best" first lens is the one someone prefers and uses the most for ones own preferred pictures, but then a beginner is not going to have the background experience to make that judgement, and even a well intentioned and knowledgeable sales person is not going to be able to help.

 

[Tim Tucker 2]

And your complete acceptance and understanding that wide angles exaggerate and telephotos compress perspective goes completely out of the window when presented with a far more juicy *conspiracy* that it was all a marketing ploy?

There really is a preference among even the *experts* to look for an find the answer in the kit because the kit is absolute and can give an absolute answer. But they ignore, as many do, that creating a photograph requires TWO optical systems. You will never see the results unless you look at them with your human eyes and human vision is more relative than it is absolute. So although there is a great deal of vagueness, a photograph TAKEN WITH 35MM CAMERA and a 55mm lens will produce a perspective that is most similar to the actual scene. If you took the shot, printed it and then stood in the same spot with with the photo at arms length you should see roughly the same perspective. Try that with a wide angle or a telephoto and they won't line up. And yes I know that perspective only changes with camera position and not focal length which is entirely correct, and if you examine only the photos by cropping and zooming etc. you would find that they *absolutely* do have the same perspective. But try it and see and you'll see that I'm correct. Making a photo requires two optical systems and the effect of the distortion of the perspective is in the way the human eye interprets the data and tries to re-build a 3D understanding within a different FOV. Yes the FOV is not captured by the camera an transferred to the print where it's easily recognised and understood by the viewer. The FOV is guessed by the brain and is roughly equivalent, (but not exactly), to the FOV of a standard photo at standard distance, (vision is not absolute and you can view an image in many different print sizes and you will get slightly different impressions but not as big as you think, this is because one of the main functions of vision is to present a *consistent* view of the world not an *absolute *one). You can shoot with a lens of 120 degree FOV, but if you view that on a computer screen where it only takes up 35 degrees of you field of vision then the basic assumption that your brain makes when re-building a 3D understanding is incorrect. This is why you *see* the distortion of perspective. And if you can prove it's not in the photo with zooming and cropping but is visible when you look at the print then there is only one other optical system left...

The idea of *standard* lenses was around long before the 35mm format, and it was only with the 35mm format that 50mm was standard. Ziess made lenses long before the 35mm format, (I have a 1910 Ziess Tessar which is slightly wide of standard for a 1/2 plate camera). There were a whole class of different formats before the 35mm including 120 and they all have their own *standard* focal length.

*Standard* focal length lenses have one massive advantage over all other focal lengths, they produce images that look most like what you saw, the was you saw it and remember it. This is why they are sold with cameras, because they are the easiest lenses for beginners to use and produce lifelike photos with.

That Ziess, and many other manufacturers, used the trusted formula that the diagonal of the film format is roughly equal to the lens that gives the most neutral perspective, then rounded that into a lens the was most easy and economical to produce for what was aimed as a mass production camera, well before computer aided design is not a marketing ploy. But it is a mistake in visual imaging to try to find absolute answers in *what you see* while ignoring *the way you see*. Vision is not absolute, and understanding why *standard* lenses are named that is based on the way WE see images and not how the camera records them.

But as said before, output media is anything but standard now so how you view the image has a great effect on the lens used as *standard*. Ask any film-maker what the effect of panning has at the edges of the frame is when you use *wide angle* lenses. Try it and then try it with a *standard*. Now you will understand the definition of standard.

P.S. Most cameras are now sold with zooms which reflects not only the different ways in which we view images but also how our acceptance and understanding of photographs has changed. The distortion of different focal lengths no longer looks odd, it looks and is accepted as far more normal. The boundaries and definitions of standard can’t really be judged and understood by how we view images today.

 

[pixmedic]

reading between the lines of condom and butt plug jokes, he didnt really offer very many compelling reasons why a 40mm was "better" than a 35 or 50....
ok, so maybe a 50 would be "too tight" in some situations, but a 35 being "too wide" doesn't make any sense in the digital world of cropping.
the 40mm pancake may be a little cheaper than a 35 or 50 that isn't an f1.4, but not by a lot. its a little smaller and lighter, which might be a bonus for some.
im not really sure i would have much use for a 40mm if i already had a 35 and a 50.

 

[ac12]

BTW, the mathematical normal lens based on the diagonal of a 24x36 frame is a 43mm.
a^2 + b^2 = c^2
if a = 24, b = 36, then c = 43.
So closer to a 40 than to a 50, but even closer to a 45.

Interesting side note, the Nikon 43-86 ends up starting on this diagonal measurement. It always seems like an odd number to start a lens on I would have thought 40-80 or 45-90.​

The 6x6 comes out almost exact to 80mm.
a and b = 56, c = 79.

Comparing the two formats 35mm film and 6x6, the 50mm lens comes close to matching the horizontal angle of view of a 6x6, interesting.

35mm 43mm lens; diagonal =53 degrees, horizontal = 45 degrees, vertical = 31 degrees
35mm 50mm lens; diagonal =47 degrees, horizontal = 40 degrees, vertical = 27 degrees
6x6 80mm lens; diagonal = 52 degrees, horizontal and vertical = 38 degrees​

Interesting point about the eye.
The angle of view of the eye when LOOKING is much narrower than what the eye can see. And based on various diagrams, like this one:

angle of view of human eye - Bing images​

There is a broad range of focal lengths that would fit into the 30-60 degree horizontal arc.
And depending on your training and behavior, you may have a narrower or wider arc than others.

But in practice, whatever the lens, you get used to what you have to work with, and don't think about it.

With my 35mm SLR it was a 50mm (before I switched to the 43-86 then the 35-105).
With my 35mm P&S it was a 35mm fixed lens.

Yes I knew it was wider than my SLR, but after a short while I simply got used to it, and it was a non-issue.​

If you have a fixed lens camera like a TLR, well you have no choice. It was 80mm.
With an interchangeable lens, YOU can make your own decision what YOU like as a normal lens.
BTW, I used a normal/standard zoom for so long, since the 1970s, that I do not have a 50mm reference in my head. For me, the fast prime was and is simply to have a faster lens than my zooms.

 

[Derrel]

Some interesting points have been brought up in this thread. I have read pretty much all of the ideas put forth, plus others. Another little anecdote regarding the standard or "normal lens" issue for 35mm cameras was that back in the Nikon F era, Nippon Kogaku (the camera,lens,and instrument maker's original name for the first several decades of the company's existence) created a 58mm lens that allowed perfect, two-eyes-open viewing through the then-current Photomic-series interchangeable pentaprisms. Depending on the camera viewfinder's magnification level, the focal length needed for same-size or life-size apparent viewing magnification varies; at the time, a slightly longer than 50mm lens was needed. Interestingly, there was a one time a slight difference in measured focal length and specified or nominal focal length on Nikkor 50's; the number I've most often seen was 48.5mm actual measured focal length. Anyway, in the early days of the F's reign, there were 50mm,55mm,and 58mm "normal" lenses available for purchase, as well as the high-speed 58mm f/1.4 apherical-element Noct~NIKKOR for low-light photography without coma on point light sources like stars or lights at night seen against dark backgrounds.

Speaking of which...I started of my serious photography habit with a 58mm f/2 lens. My experience with that is that 58mm is much more-selective than is a 50mm. In the same time frame that I had the 58mm, I also had a 45mm Zeiss, and a 51mm Kodak. I also had a Canon 40mm. Each of these focal lengths offers a different look; the difference between a 45mm and a 50mm and a 58mm is very noticeable.

1) Angle of view, and also 2) the way foreground and background spatial relationships are both pretty important when using a prime lens. On a 24x36mm film frame, on on full-frame or "FX" digital as Nikon calls it, the 50mm lens offers a fairly human-eye-vision-like rendering of the world with regard to the foreground/background relationship as far as on-film or on-sensor sizes of objects go. The 50mm does not make foreground objects appear large and background objects tiny, but instead, as one poster mentioned, it renders the scene fairly closely to the way the human eye and brain sees and interprets the world.

As one goes shorter, the size of background objects becomes smaller-than-in-real-life, while as one goes telephoto, the on-film size of background objects appears larger-than-in-real-life. The 50mm lens on 24x36 capture medium looks...pretty normal.

It might sound odd to some people, but to me, the 55mm macro lenses from Nikon, and the 60mm macro lenses, and the 58mm prime lenses are very good at allowing an ever-so-slight degree of selective seeing. Pop a 55mm or 60mm macro on a camera that shoots to a 24x36mm sized capture medium, and you have what is in effect, a very,very short telephoto length lens...it's got a narrowish angle of view, and it magnifies the background objects sizes just a tiny bit,and it ever-so-slightly compresses apparent distances.

Anyway...yeah...the 50mm normal lens...

 

[nerwin]

There are 5 reasons for just about any focal length you can think of.

 

[jcdeboever]

It might sound odd to some people, but to me, the 55mm macro lenses from Nikon, and the 60mm macro lenses, and the 58mm prime lenses are very good at allowing an ever-so-slight degree of selective seeing. Pop a 55mm or 60mm macro on a camera that shoots to a 24x36mm sized capture medium, and you have what is in effect, a very,very short telephoto length lens...it's got a narrowish angle of view, and it magnifies the background objects sizes just a tiny bit,and it ever-so-slightly compresses apparent distances.

Anyway...yeah...the 50mm normal lens...

I agree with your commentary on the 55mm. I love using my Nikon F with DE-1 (eyelevel) finder. I have experimented with the 43-86 zoom as well on it but not really into it all that much, even though it's zoom range is rather handy when your feet hurt. I pretty much use the 55 f/3.5 micro exclusively on this camera for it's close focus ability, crisp subject render and tight OOF backgrounds. It really is a favorite lens of mine. I recently acquired an F2 with the DE-1 and considering putting a 58mm on that and two gunning it, one with B & W (F), and the other with color (F2).