What is a "normal" lens?

Discussion in 'Medium Format' started by david_holland|1, Jan 24, 2004.

  1. Linsday Robb posted a question yesterday about which lens to keep for
    a Pentax 645 manual out of a 120 macro and a 150. It was addressed to
    people who had both and answered well by several posters. I studied
    the question without experience with either, and bought only the 150,
    so I wasn’t part of the audience to whom the question was addressed.
    For what it’s worth, I have been happy with the choice for the
    reasons the posters responded.

    That said, the question and responses brought some thoughts to mind
    that I didn’t think were sufficiently responsive to post as a
    response, but I would like to share for comments.

    We all know that normal lenses get their label, at least in part,
    because their angle of coverage supposedly covers the same angle of
    view as the human eye. However, once upon a time, many, many years
    ago, in my 35mm days, I noticed that landscapes and travel images
    with “normal” lenses didn’t really look the way I remembered the
    scenes. I discovered that lenses in the portrait range more nearly
    recorded what I thought I saw.

    I recall (hopefully with some accuracy) an article in a photo
    magazine by someone who studied paintings by the old masters of
    landscapes that are still basically unchanged today. The author went
    to the various sites and determined which focal length of lens (for
    35mm) would record the apparent perspective that Gainsborough,
    Constable, or other artist had recorded in his oil. His conclusion,
    expressed in 35mm terms, was that their “focal length” was not
    the “normal” one, but rather 85 to 105.

    There is, of course, no greater confirmation than someone else,
    however studied, saying something that one already himself believes.

    Much more recently, in the computer era, I had my monitor refresh
    rate set a bit too slow. The flicker was not at all noticeable in the
    central area of my vision, but it was annoying in my peripheral
    vision. I realized that the peripheral area does not appear focused,
    but movement stands out more readily, while movement in the central,
    focused area is dampened to improve clarity of view. Hours in a blind
    either hunting or bird watching had already confirmed a consistent
    involuntary trait, at least of mine: any movement outside of my
    central vision caused my focus to immediately move to that spot.

    It all clicked. The human eye sees only a small area of the field of
    vision in sharp focus. Our evolution and heredity as predators, with
    our binocular vision, politically correct or not, causes us to see
    that upon which we focus and to exclude (if it doesn’t move) that
    which is not. Although the angle of coverage of the lens on our eyes
    may indeed approximate that of a “normal” lens, what the brain
    records in memory approximates the coverage of a portrait lens.

    Perhaps if what you want to record is what you see at first glance,
    the 150 for 645 or 180 for 6x7 is ideal. If what you want to record
    is what your photo eye can select from a scene (that the rest of the
    world), it may be better to get away from 150/180—far away. Does this
    explain why “normal” and wide angle lenses are such useful tools in
    the hands of the talented—those who don’t want to bring home what the
    next door neighbor saw when he was there?

    I haven’t seen this mentioned anywhere since that article came out,
    and the speculation in the preceding paragraph came to mind only
    after I wrote what precedes it. Are the author of that article and I
    the only ones who think the explanation for “normal” is misleading?
    Is the real explanation that, regardless of focal length, it is the
    angle of coverage that can be manufactured smallest and fastest?

    And, to depart from physics to metaphysics, any comments on my
    speculation as to how it relates to what we see as photographers?
  2. Advocating a specific focal length on the grounds that it conforms to a uniquely human vision just doesn't fly.

    Firstly because a photographic lens simply doesn't perform in the same way as the eye (or more accurately the eyes). And secondly because the comparison also depends on the size of the final print and the viewing distance, so lens choice is only one half of the equation.
  3. In 35mm photography, 85mm has long been considered the lens of natural perspective. Personally, I think we see in wide angle, but we focus our attention more on the angle of view an 85mm lens provides. So, a whole range of lenses from wide angle to short telephoto could be considered normal, depending on how you define normal. This is why some people, including me, feel that a 35mm lens (for a 35mm camera) is closest to providing the overall effect of using our own eyes - it's a good compromise between 24 and 85.
  4. This isn't a new idea; you will find references to this in a number of texts on photography and perception.

    I think most people go through two stages. First there is the realisation that their perception of a scene built from the eye and brain (and other senses - sound, touch, and smell mainly) is a composite that cannot itself be photographed. This leads to using what *can* be photographed as an abstraction to convey the key mood or aim of the subject. The tools we use - wide lenses, long lenses, or even no lenses (pinholes or photograms perhaps) often show us images that we cannot see with our eye. We can learn to imagine the effect with experience.

    The second stage starts in the imagination and uses the tools and the subject to fulfill the concept. It is much easier to do this when you are fluent with the first stage. We have wide lenses, long lenses, we can even photograph using light we cannot see. But none of them are much help if we cannot decide what to keep in and what to leave out.

    As to what is a 'normal' lens:
    a) It is the only one on the camera
    b) It is the lens you use most often
    c) It is a state of mind
    d) Any, all, or none of the above
  5. By definition a "normal" lens is of a focal length approximately equal to the diagonal of the frame.

    The lens you use most often? That's up to you.
  6. In movie cameras; a normal lens is many times twice the film diagonal dimension.
  7. Hi,

    It is that which gives the perspective you want in your print.

  8. "The human eye sees only a small area of the field of vision in sharp focus."
    Moreover, only that area we see in color. We have two different type of receptors in the yellow spot that detect the light; at its edge, the receptors which can distinguish between colours, are missing.
    I guess, it was more efficient like this, in the beginning of our evolution.
    Disregarding definitions, for me a normal lens is one that does not surprise me when looking through it, respectively, does not produce a weird perspective image.
  9. I suspect the idea of a normal lens came about purely on an economic basis. Out of the range of focal lengths, there has got to be ONE focal length that is cheapest to produce, and that one gets sold as the "standard". Wider-angle lenses suffer from light fall-off & require retrofocus designs, longer lenses require progressively larger elements for the same speed. On 35mm SLR's, this was 58mm, then 50mm (regardless of the fact that neither matched the diagonal of the frame). On modern point-and-shoots, it tends to be around 33-35mm.
  10. Great answers, all!

    As to what is normal as defined by those who manufacture our equipment, I tend to agree with Stephen--I mentioned that "normal" seemed to be what could be made smallest and fastest, and was thinking after the post that those same characteristics translate to cheaper.

    Diagonals, for those who are interested, are 35mm: 43; 645: 70; 6x6: 79; 6x7: 89; 4x5: 163. Based on 8x10 enlargement (i.e, the portion of the negative that wouldn't be cropped out), they are 38, 66, and 72, with the two larger ones being the same. Those seem a bit shorter than what we usually see, certainly in SLRs.

    Csab's concept equates with what I feel it should be: the apparent perspective that does not surprise us.

    Your more subjective expressions, it seems to me, get back to the concept that our lens selection only determines how we define the apparent perspective our viewer will perceive as we reduce three dimensions to two, regardless of enlargement or viewing distance (apart, of course, of the times when magification and working distance requirements limit our choices).

    Gary started us out by introducing enlargement and viewing distance into the mixture. My sense is that, in medium and large format, the ability to create images of sufficient size and detail that they must absorbed by moving our area of sharp focus about the image is what causes our "normal," as determined by what we ususally use, to shrink somewhat below the 85mm equivalent(as Pierre indicates, even into the wide angle range), but rarely to exceed it by much. I'm not sure I would agree, however, that 28mm in 35 terms is a "normal" focal length and certainly not a 24.
  11. I wonder if the "85mm corresponds to human vision" idea is based on the way a lens of about that focal length on a typical 35mm SLR will net back to x1.0 viewfinder magnification. In other words, what you see through the viewfinder is neither enlarged nor reduced compared to your unaided vision.
  12. Going back to the landscapes of the old masters and estimating the angle of vision of the painter to see which lens they might have used had they had a camera, sounds like one of those academic goose chases that get funded by the very rich universities. What a fun job. But it misses the point that a painter can compress or broaden the angle of view at will for the sake of composition, just as the photographer can by the choice of lens and position. In 35mm terms, I'd say that Rembrandt used a 200mm for many of his landscapes, but Jakob van Ruisdael would almost certainly have used a 28mm or even a 24mm. It was the Impressionists who started the vogue for the 50mm, but then they were painting out of doors ... except of course for Cezanne, who used a Holga. Piranesi was the first artist to use the fish-eye for architectural views, and this was later taken up by Escher. And while on the subject, El Greco didn't paint any landscapes because of his astigmatism, (or was it cataracts?) but auto-focus might have helped him if it had been available. Did van Gogh invent cross-processing? And why didn't Monet ever use a polariser instead of that blue filter? Turner, of course, should have kept a packet of Speedy Lens Wipes with him at all times - some of his landscapes look as though they were taken in a snowstorm ... (The sound of rain is keeping me up, I need to sleep :.)
  13. My apologies if this sounds tiresome, but I'll return to an earlier comment I made. Our perception of perspective is every bit as much influenced by print size and viewing distance as it is by focal length.

    Put an extreme wide angle lens on a 35mm camera, take a shot of say a car from close to the front of the bonnet (hood if you're in the US!), make an 8x10 print and look at it from about 20" away. It looks like a wide angle shot, even a modest family saloon will look like an E-Type Jaguar with a huge, extended bonnet.

    Now print the same negative as a 20x30 and view it from 6" away, and the modest family saloon once again looks like a modest family saloon, "normal" perspective has been restored purely by manipulating print size and viewing distance.

    Extending the principle further, it's interesting to note that what we usually class as normal focal length lenses tend to give a very natural perspective when viewed as contact prints from an easy "arms length" viewing distance. And for at least the first 100 years or so of photographic history, when so many of our photographic pre-conceptions were enfranchised, contact prints were the norm.
  14. The most famous and sucessfull lens made in the last century is the 4 element Carl Zeiss Tessar. The Schneider Xenar is a close clone. Most of the Kodak Ektars PRODUCED; ie quantity produded are 4 element "Tessar" type designs. The Kodak "Ektar" name is used for their premium lenses; many DESIGNS are not "Tessars"; and have more elements; and wider coverage. <BR><BR>These "Tessar" type designs are darn sharp in the central core; but the lens has only about a 40 to 45 degree coverage; which depends on ones criteria of sharpness.<BR><BR>Since the "Tessar" was used in some many cameras as the "normal" lens; it defines what is "normal". For movie cameras the sharpeness criteria are more strict; and thus most "normal" Tessars are alot longer than the film diagonal; than still cameras.
  15. To get the correct perspective when viewing a print, you should adjust you viewing distance and position so that you have the same angle of view that the camera lens had - in all directions... up, down, sideways or whatever, especially if camera movements were used.

    A good photograph looks like the view through an open window.

    When a friend asked me what I thought of his new audion syste, I said it was like looking at a view through a frosted-glass window - but I could have said the same with reference to low-res digital C**P.
  16. Quote from someone above: “By definition a "normal" lens is of a focal length approximately equal to the diagonal of the frame.”

    That’s total nonsense. How could it be “by definition”? There has to be a scientific reason behind it (which I’m going to give you).

    It will once and for all dispel these “myths” about normal lenses and what they mean. Also the “normal” lens on a 35mm camera is not even that close to the length of the diagonal. The diagonal of the frame is 43mm the “normal” lens is 50mm. That’s a 16% difference! Hardly a good match.

    If you take a 35mm photo with any focal length lens (this applies equally to any other format camera too), there has to be a unique position somewhere in front of you where you could hold up the print and it would fit EXACTLY and perfectly over the scene just shot.

    The distance might be 6” in front of you or it might be 60” in front of you but it will be somewhere. It’s probably best to imagine a print printed on to a plastic sheet so that you can view the real scene through it.

    Well “normal” lenses are all to do with viewing the print (or the plastic sheet in my example) at the NORMAL READING DISTANCE of the eye; usually taken to be around 10” (250mm). Of course we need more information to do a calculation, like what size print are we talking about and what angle of view is the camera seeing.

    This angle has NOTHING to do with the angle the naked eye sees views at. As someone above pointed out the area of clear focus is very small (it’s actually only about 7 degrees) while the angle of peripheral vision is huge (> 120 degrees).

    (As an aside the reason the eye sees like this is that the brain would require far too much processing power if it had to cope with perfectly focussed vision at all times over a large angle of view. So instead the eye scans a scene to make the best use of brain “bandwidth” and concentrates only on a small central region - the "fovea").

    Right, we have our distance (250mm) but how large is the print? This will make a difference obviously to where we can hold the print in front of us and cover the scene. So an arbitrary “normal” size print is chosen by the industry of approx 7” x 5” (whose diagonal is therefore 8.6” or 218mm). It can be shown mathematically, by trigonometry, that the angle subtended by the diagonal of a 7" x 5" print at a distance of 10" is 47 degrees. For those interested it is

    tan(x/2) = (print diagonal)/(2 x distance)

    where x = the angle we are trying to find

    therefore x = 2arctan[(print diagonal)/(2 x distance]= 47 degrees in this case

    So the calculation becomes, what focal length lens on a 35mm camera will produce a 7” x 5” print that exactly covers the real scene at a distance of 10” with a 47 degree angle of view?

    From trigonometry it can be shown that:

    tan(x/2) = 218/(2x250) for the print

    and also, tan(x/2) = 43/(2 x F) for the camera

    where x = 47 degrees and F = the focal length of the lens

    Therefore by equating these two and rearranging:

    F = (43 x 250)/218 = 49mm and the industry happens to round it to 50mm. QED.

    Generically therefore, F = (diagonal of frame) x (reading distance) / (diagonal of print)

    Of course, if you decide that an 8” x 5” print is a more realistic “normal” size (and you are perfectly entitled to choose this or any other size) then a “normal” lens would be:

    F = (43 x 250)/240 = 45mm (much closer to the frame diagonal dimension by the way).

    So there is a scientific basis to “normal” lenses but I see and read so many fairy tales being perpetuated on the subject I thought I would give the mathematical basis for it.

    Incidentally it is interesting to do the calculation for other focal length lenses (eg: 100mm and 200mm etc) to ascertain what distance the resulting print needs to be held from the eye to exactly cover the scene and for it to look "normal".

    For a 100mm lens on a 35mm camera, substituting back into my equation (and using a 7" x 5" print), the distance is about 20 inches.


    Chris UK
  17. Shoot your sixties/seventies/eighties typical 35mm SLR with BOTH EYES OPEN, and you'll immediately know why a focal length of approximately 50mm is going to be considered "normal" for most SLR viewfinders.

    SLR designers have established (or re-enforced) the image diagonal as the "normal" focal length, by standardizing the viewfinder to match normal vision with a 50mm lens.

    This discussion thread (and many like it) is just an indication of how many people shoot with one eye closed.

Share This Page