1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

something of an Error in "AskAP" this week

Discussion in 'AP Magazine Feedback & Suggestions' started by Terrywoodenpic, Jul 8, 2012.

  1. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    “Ask AP” ( tips on Portraiture) Page 42
    That section gives some very inaccurate information behind the “Perspective” obtained with wide angle and telephoto lenses, and the “perspective distortion” relating to their use.

    The facts are these. ...
    Neither the focal length nor the construction of a lens changes the perspective at all.
    Perspective is defined by view point alone. That is say the distance from the lens to the various objects.
    The geometry of perspective can be predicted and drawn with out any reference to a lens.

    If one were to overlay the image taken with a wide angle and telephoto lens, taken from the same spot, by enlarging the wide angle one, they would match exactly.

    So what changes? Mostly your feet... you get closer with a wide angle and change the viewpoint. You do this to fill the frame.
    So why do you get compression with a telephoto lens? You do not. The compression comes from the geometry of being far from your subject. The distant bits taken with a wide angle lens are equally compressed.

    So why “BIG NOSES” in Portraiture?..... because you are too close. So why use a long focus portrait lens? ….To fill the frame at a greater distance.

    You move your feet to get the best view point, and chose a lens or zoom to fill the frame.

    A fish eye lens is constructed to provide a different “Projection” to a rectilinear lens, and purposely introduce other optical distortions unrelated to normal perspective. However even these can be re mapped in stitching software to give normal perspective.

    (I have not covered depth of field which is related to aperture and also viewpoint (distance) and by inference: magnification, focal length, and the accepted circle of confusion.)
  2. P_Stoddart

    P_Stoddart Well-Known Member

    Mmm, I think not. That exactly the point is it not. If shoot your face filling the frame with it on a wide it will distort because the vanishing point are inside the frame.

    But if I shoot on a telephoto and crop to give the same framing as the wide no distortion because the vanishing points are probably outside the frame.

    It is by adjusting the vanishing points you can give a wide angle effect in drawing. That is what the lens does, it moves the vanishing points. That is what perspective is all about ie using vanishing points.

    Under certain situation you can get a very close match by using cropping say when boosting a telephoto to telephoto ie 18MP at 400mm crop to 9MP to simulation 600mm. But it is not exact because of DOF but close enought to work. But wide to telephoto crop I think would show up.

    By the way have not got a copy of magazine yet. I wait until in shops on Tues.
  3. Benchista

    Benchista Which Tyler

    You've completely missed Terry's point - if you're using the same framing, you're not in the same position. Terry is absolutely right about perspective. You're talking absolute nonsense about lenses moving vanishing points - the ONLY way you can manipulate perspective is by changing position, the lens is merely a means for an end in doing that, it has no magical properties of its own.
  4. P_Stoddart

    P_Stoddart Well-Known Member

    So you are agreeing that the magazine is wrong. :)

    All lenses have properties (not magical ones). That is why they differ. So colour shifts, different sharpness.

    Otherwise you would not get the same focal length from different designers.

    Distortion is one such properties, come on Benchista even you know that from shooting on different lenses from Canon.

    A 28mm wide prime will not give the same a 28mm on a zoom lens.

    Unless I misunderstanding Terry is suggesting a crop of a wide angle shot to the same as telephoto shot from the same position "would match exactly", you know that not true. You argued that point in the TC debate awhile back.

    And we know the DOF would not be the same. Would the prespective be the same, possible, but it would depend how good the wide angle is at it's edges for example.

    Let say the telephoto is point at something that is at the edge of the wide angle view of the same scene from the same spot. When you shoot the wide version the film/sensor plan would be parallel to say a wall. But to get the same area of the wall if it at the edge the film/sensor plan has to change it's angle. That changes the view you end up with. You are not changing your distance but you cannot help but change your angle.

    If the subject is in front then yes you don't change your angle. But some of the subject is going change angle. With a telephoto less of the film/sensor plane changes angle to the subject.

    That is why I take issue with a general statement as exact match. Not in all situations.

    Bear mind Terry is being critical of a misuse in a article when he is IMHO making worse general statements.

    But as I said I have not read the specific piece in the magazine yet. So I may have to offer a apology down the line. :)
  5. Norman

    Norman Well-Known Member

    I'm afraid you are completely wrong. The ONLY thing that can alter perspective is the viewpoint. Take a look here for a very succinct explanation/example.

    Go out with a zoom lens and take a picture at various focal lengths from the same position. Take the longest focal length as your datum and crop each of the other images to match the view. The perspective will be identical in each.
  6. Bob Maddison

    Bob Maddison Well-Known Member

    I agree with both Terry and Nick. For any normal lens, the image "magnification" is equal to the lens to image distance divided by the lens to subject distance. In practice, for all but macro and micro photography, a good approximation is to say that the "magnification" is equal to the focal length of the lens divided by the lens to subject distance. Most importantly, this applies to ALL parts of the image whether in focus or not.

    It has nothing to do with the lens focal length, and it is the relative "magnification" of parts of the subject at different distances that give the image its "perspective".

    However, although Terry was careful not to deal with Depth of Field, there is a significant difference in this. Try taking a picture of a good 3D scene with a WA and Tele lens, from the same position and at the same lens aperture, and then crop the WA image to show the same field of view as the Tele lens. You will find that the cropped WA photo shows greater real DoF than the tele photo. However, the overall IQ will be worse and you might prefer the Tele image!!!
  7. mark_jacobs

    mark_jacobs Retired

    OUCH! I'll have the matter investigated :(
  8. RogerMac

    RogerMac Well-Known Member


    and does anyone have a "this thread is useless without pics" smiley

  9. beejaybee

    beejaybee Marvin

    Assuming the lens isn't distorting.

    Most lenses do distort, at least to some extent which is noticeable in extreme situations.

    However it is quite correct to assert that - with the model of a lens as a "fast pinhole", and ignoring diffraction effects - the perspective depends on the viewpoint and nothing else. The MAIN reason for using a longer focus lens is to change the image scale to map the desired field of view more accurately to the light sensitive area of the sensor or film gate. Control of depth of field is SECONDARY, control of perspective is not a factor at all.

    So far as portraiture is concerned, where realistic depiction of the subject is more important than artistic effect, lenses giving field of view between 20 and 30 degrees (corresponding roughly to 135mm to 85mm in full-frame terms) have long been regarded as the most useful, and with good reason. The focal length is long enough to encourage the photographer to maintain adequate distance from the subject to avoid significant perspective distortion without requiring an excessively large studio.
  10. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    I was not stating any facts that are open to any realistic contradiction. what I said was perfectly correct and is very "old Knowledge" It was one of the first things we were taught at college back in the 50's.
    It is just one of the facts known by the "Old Masters" for centuries, that allowed them to draw accurate perspectives. and has been used in optical drawing aids ever since.

    In three point Perspective drawing, the converging verticals seen when a camera is unlevel are also constructed and drawn.

    The Idea that a camera lens in some way manipulates vanishing points is ludicrous.

    The distorted face that you suggest is the result of lens distortion is simple geometry, and would be the same through a pinhole camera placed at an identical viewpoint..
  11. AlanClifford

    AlanClifford Well-Known Member

    I did once read an article insisting that the depth of field would be the same too. You just can't see it in the wide angle photographs.

    I'll see if I can find the URL.
  12. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    Very true.
    Lens distortions and aberrations Like Barrel, pincushion, astigmatism and spherical. are all down to the inadequacies of lens design, and do not produce the effects under discussion.
  13. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    It is dangerous to take articles as "Gospel" (they tend to perpetuate both fact and error.)

    But it can be shown that if the one is magnified to the same size as the other there will be the same depth of field for a given aperture and circle of confusion.

    However the images are rarely enlarged so that the objects in shot, are the same size, so the question is academic.
    Last edited: Jul 8, 2012
  14. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    Only if there is a "PRIZE" :D

    This is a very common error these days, and gets repeated time after time.
    It became blatant when he said it applied at all distances......the complete reverse of the truth.
  15. P_Stoddart

    P_Stoddart Well-Known Member

    Across the whole image from the wide angle lens?

    Let say I do a crop off centre of the wide angle image to the same level as a telephoto. You sure that will be near identical as a crop taken centre frame? Leaving out lens distortion.

    That was my point. Terry did not say what part of the wide angle shot. I believe (and I could be wrong) that it only applied to objects/views directly in front of the camera.

    Once you take a crop from anywhere else of the wide angle I don't think it gives you the exact same when shot on a telephoto, even if you close down and have max DOF. I think there will be differences. Certain parts of the view will appear that are not there in the wide shot.

    Now the term perspective could be said to mean point of view (POV).

    Well when you change a lens it is change a POV.

    You certainly will move the camera position (or the cameras POV) even if you don't move yourself.

    Now as I wrote I not read the article or piece but I am not sure the use of the term perspective is incorrect.

    Which I believe is the thrust of Terry's complaint. :)
  16. Bob Maddison

    Bob Maddison Well-Known Member

    Surely, when we are talking about comparing views, we mean that the camera is in the same position (i.e. on a tripod), and that we are dealing only with the central part of the WA view (this is implied by the camera being on a tripod so that its orientation doesn't move when we change lenses). Without these terms of reference, discussion is irrelevant.

    Although distortion is a possible factor governing the appearance of a scene, most "normal" lenses are corrected so as to minimize distortion across the field otherwise it would be very difficult to create panoramas. (Just try it with a fisheye!).

    My dictionary defines 'perspective' as: "the appearance of objects with reference to distance, relative position etc."
  17. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    The peripheral image area, from a wide angle lens is rarely so poor quality, through design faults. that the perspective drawing changes significantly from center to edge. Not long ago AP gave the extent of pincushion and barrel distortion in their lens tests. In real terms this was only significant in low end super zooms.

    With that consideration in mind... wide angle lenses do not distort.
    The visual effect is caused by view point alone.
    That is the whole thrust of my argument.

    It matters not, which part of the image circle is considered.
  18. Benchista

    Benchista Which Tyler

    I've not read the article either, but I did bother to read Terry's original post - it seems you either didn't, or didn't understand it. We're talking perspective, and perspective only. You're just muddying the waters - for a change. ;) And you clearly didn't understand what I said in the TC debate - I can't remember what I said, but it certainly wouldn't differ from what I've said here.
  19. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    The reality of what we are discussing is why a wide angle lens appears to distort.....
    As I have explaind the perspective or "Drawing of a photograph" is defined by view point alone

    However a wide angle lens allows you to come very close to some parts of the subject.
    This makes those parts larger in proportion to parts further away. This is caused by the relative distances (viewpont) and follows the normal perspective constructions.

    This distorted view point has become associated with wide angle lenses ( which simply record the effect produced at the viewpoint)

    The same is true of the compression effect seen when using "Tele"lenses this is not caused by the lens but by the distance involved.

    When teaching technical matters in Photography or any other subject. It benefits no one by telling half truths or makebelieve. As the learner has to use the information he is given to build on, and inter-relate with other facts as they arise.

    Errors tend to be compounded by repetition and always confuse.
  20. daft_biker

    daft_biker Action Man!

    Aye, magnification and aperture are all you need to work out DoF and it's always half in front and half behind the point of focus. You just can't see it like that in low magnification shots;)

    Outside of macro it's possibly not a particularly useful way of working which is why folk apply a tolerance for what will look sharp for a given output size and substitute magnification for focal length and focus distance. It's just taking into account real world factors.

Share This Page