Page 7 of 9 FirstFirst ... 56789 LastLast
Results 61 to 70 of 89

Thread: Portrait perspective: Quiz and two questions

  1. #61

    Join Date
    Feb 2004
    Location
    Scottsdale, AZ
    Posts
    489

    Re: Portrait perspective: Quiz and two questions

    Quote Originally Posted by Ralph Barker
    Juergen - the test is being conducted on Sunday afternoon at my place in Rio Rancho. But, you have to bring the tequilla for the pre-test margaritas.
    I can always be convinced of a good margarita - how many bottles Tequilla should I bring? I am sure all this math is much easier after the first 5 or 6 margaritas and maybe it'll even make sense.
    Juergen

  2. #62

    Join Date
    May 2002
    Posts
    1,031

    Re: Portrait perspective: Quiz and two questions

    The math discussed is extremely important when dealing with the science and craft of optics, and can provide and interesting diversion when one has time to kill. It has little to do with photography, IMO.

  3. #63

    Re: Portrait perspective: Quiz and two questions

    Can anyone make an intelligent guess as to how well Ansel Adams would do on my quiz?

  4. #64

    Join Date
    Apr 2006
    Location
    Hell's Kitchen, New York
    Posts
    525

    Re: Portrait perspective: Quiz and two questions

    "I will keep looking, but if someone else can clarify all this and provide a reference, I would appreciate it."
    Leonard,
    What specifically do you want a reference for?

    The first principal plane goes through the first principal point, not the first nodal point, though they may be coincident. Remember that the two principal planes and the two nodal planes are only coincident when the lens is in a single medium. They need not be coincident for the Nikonos underwater lenses, for example (but they would be coincident for the amphibian lenses, because those are really all-air lenses).

    Zeiss have a lot of information on their website if you want some examples of the locations of the various planes. For example the 60 mm f/3.5 CFi:
    Entrance pupil position 32.2 mm behind the first lens vertex
    Position of principal plane H 53.8 mm behind the first lens vertex.

    Where is the quote that says "that the distance from the entrance pupil to the front nodal point is often negligible"?

    "Helen appears to be right about something, but I'm still not sure what." Would it be easier to start with what I'm wrong about?

    Best,
    Helen
    Last edited by Helen Bach; 27-May-2006 at 11:10.

  5. #65

    Re: Portrait perspective: Quiz and two questions

    In Ray, third edition, p. 123, I found this: "The centre of the entrance pupil is the centre of perspective of the lens." I had missed it because it is in the index under "center" instead of "perspective." I too am also puzzled by this, because I thought I had worked out a demonstration that it had to be the front nodal point.

  6. #66

    Join Date
    Nov 2003
    Posts
    1,219

    Re: Portrait perspective: Quiz and two questions

    Helen,

    I was aware that the nodal points and the principal points might be different, but I assumed the medium was the same on both sides of the lens, thus, as you point out excluding underwater photography. I don't know anyone who does large format underwater photography, but I suppose it is possible.

    I don't know of anything you are wrong about. I just don't understand the signficance of the difference between the principal planes and the entrance/exit pupils, so although you are certainly right in what you said, I don't fully see what its implications are beyond panoramic photography. The only thing that makes any sense is that there is some sort of shift in the position of the image when computed using the principal planes, but I don't understand what that might mean. I've looked at the Leitz tables and some others, and they do in fact confirm that these differences do exist for many lenses. But despite some extensive searching and some time spent in NU's Science Engineering Library, I don't understand how these things relate and just which of them you use to calculate which parameters.

    As to Jacobson, I can't find the original source of my statement, but if you look at his Lens Tutorial at

    www.photo.net/learn/optics/lensTutorial

    you will find the statement "It can be shown that zE = f*(1-1/p)" where "zE is the distance the entrance pupil is in front of the front principal point." Here p is the pupil magnification. He then goes on to say "For all symmetrical lenses and most normal lenses the aperture appears the same from front and rear, so p~=1", which means that zE ~= 0. So the main exceptions for large format photography would be long lenses of telephoto design and perhaps some wide angle lenses of reverse telephoto design. I wish I could see that derivation. It might be enlightening.

  7. #67

    Join Date
    May 2006
    Location
    Northern California
    Posts
    29

    Re: Portrait perspective: Quiz and two questions

    Wow Jerry - TeX and pdfTeX on this forum - I thought few outside the scientific/engineering community ever used it.

    Anyway, with all this talk about nodal points and entrance pupils, could we go back somewhat and define "perspective"? What does the statement "The centre of the entrance pupil is the centre of perspective of the lens." mean? I mean really - what is the definition of "centre of perspective" and to take another step backwards, what is "perspective", or at least what does Ray mean by it?

    To point out a few things in your paper - in example 2, the hint is "For the same perspective we need the same u"; however what you then proceed to compute is magnification, not perspective. To get 2x the magnification, there is no reason to assume a constant u, therefore there are infinite solutions to the equation. In other words, I could triple the focal length and increase u and still end up with the same magnification; or I could halve the focal length and decrease u and again end up with the same magnification etcetera.

    If you want a constant u, then you are talking about "real" perspective, i.e., the ratios of the magnifications, and then the focal length does not matter, provided u>>f and u'>>f and u ~ u' as I pointed out in my last post. For close distances and u !~ u', you cannot get the two magnifications M and M' without refocusing (i.e., changing u' or v') thereby making the question of the ratio M/M' invalid. Remember M is defined only when the lens equation is solved (i.e., the lens is focused). M is undefined when the lens is not focused, i.e., when u and v are such that the relation 1/u + 1/v = 1/f is not met.

    I hope I am being clear enough - if not I can write up all this in a TeX document as well ;-)
    Last edited by vijayn; 27-May-2006 at 14:07.

  8. #68

    Join Date
    Apr 2006
    Location
    Hell's Kitchen, New York
    Posts
    525

    Re: Portrait perspective: Quiz and two questions

    I have a sneaking suspicion that the flaw in the apparently bulletproof front-nodal-point-as-the-centre-of-perspective analysis is that it is based on the properties of in-focus objects.

    When you are considering perspective you are considering objects that are brought to perfect focus in the film plane and objects that aren't. A single ray from a point that is in perfect focus can define the location of the image of that point, because it is where that ray intersects the film plane.

    Suppose that there is a coincident ray coming from a point further away - ie from an object that is (almost) in line. This object is not imaged as a point, but as a small circle. The coincident ray from the distant object arrives at the film plane at the same point as the ray from the focussed point, but that does not have to be where the centre of the circle is located. One ray alone cannot define the apparent location of the image of a point that is not in perfect focus. The circle is defined by the whole bundle of rays entering the lens.

    Without going to the lengths of a full explanation (hey, it's 11:45 on a Saturday night and time to go out), is that enough to make it clearer why the entrance pupil is the centre of perspective? Perspective in this case being about how things in three dimensions appear to line up to an observer.

    Best,
    Helen
    Last edited by Helen Bach; 27-May-2006 at 20:53.

  9. #69

    Join Date
    Nov 2003
    Posts
    1,219

    Re: Portrait perspective: Quiz and two questions

    Vijay,

    Remember M is defined only when the lens equation is solved (i.e., the lens is focused). M is undefined when the lens is not focused, i.e., when u and v are such that the relation 1/u + 1/v = 1/f is not met.


    Keep in mind that in principle every point in the subject (object) space is imaged by some point in focus in the image space. The distances of those points are what satisfy the lens equation. Now consider what happens if you consider a subject plane closer than the plane of exact focus. The corresponding image plane is further away than the negative plane. Let u and v be the subject and image distances for the plane of exact focus and u' and v' the corresponding diistances for the closer subject plane. Each pair satisfies the lens equation. The respective magnifications are m v/u and v'/u'. Now consider a distance h' in the closer plane. It exact image (not in the film plane) will be of size v'/u' x h'. But if we pull this back to the film plane, we have to reduce it by a factor of v/v', so the size of the (slightly blurred) image in the film plane will be v/v' x v'/u' x h' = v/u' x h'. u' is smaller than u, so v/u' is larger than v/u. In other words, objects closer than the plane of exact focus will be magnfied in the negative plane more than than objects of the same size in the plane of exact focus, but by not as much as you might expect by comparing magnications for the two image planes. Similarly, objects further away than the plane of exact focus will appear to be magnfied less than the naive calculation would suggest.

    Note, however, that in most situations the correction v'/v for magnification, in either case is going to be pretty close to 1.

    Helen,

    I haven't really absorbed your argument, but I think there is something not quite right in what you say. Any image point is the apex of a cone with base the exit pupil. It intersects the film plane, as you say, in a small disc, and the center of that disc is on a line from the apex of the cone, the image point, to the center of the exit pupil. There is no decentering. (It gets somewhat more complicated iin case of tilsts or swings.)

    I don't see why your argument tells us anything about the center of perspective in any case, but let me think about it.

  10. #70

    Join Date
    Apr 2006
    Location
    Hell's Kitchen, New York
    Posts
    525

    Re: Portrait perspective: Quiz and two questions

    Leonard,

    It was a mistake for me to write half an explanation in a rush on a Saturday night. I'll re-write it with a clearer head.

    Best,
    Helen

Similar Threads

  1. Perspective Correction in Photoshop CS
    By Jonathan Lee in forum Digital Hardware
    Replies: 1
    Last Post: 14-Jul-2004, 10:10
  2. Correction of perspective
    By pancho pistolas in forum Style & Technique
    Replies: 29
    Last Post: 9-Nov-2001, 12:58
  3. Portrait lenses rec. Perspective in 4x5
    By Wayne Crider in forum Lenses & Lens Accessories
    Replies: 11
    Last Post: 6-Nov-2001, 14:22
  4. Perspective Manipulations with View Cameras
    By Patrick Chase in forum Style & Technique
    Replies: 7
    Last Post: 11-Jan-2000, 13:56
  5. A simple perspective question
    By Simon_443 in forum Style & Technique
    Replies: 5
    Last Post: 26-May-1999, 08:30

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •