Portrait perspective: Quiz and two questions

[David Hempenstall]

Sorry, but reading all this I'm getting to the point of pouring gasoline all over myself and striking a match.

[Walter Calahan]

"Sorry, but reading all this I'm getting to the point of pouring gasoline all over myself and striking a match." LOL

Who cares what the math says, it's what feels correct. I've made portraits with 12mm to 600mm lenses in 35mm format. It doesn't matter which lens I use, as long as the lenses works with the situation.

For my Carroll County project (http://www.walterpcalahan.com) I use a Nikkor 240mm on an 10-8 camera to make full length portraits of farmers. Seems right to me. But what do I know?

Simply never been good at math in the field. Never been good a pop quizzes either. Grin.

Now where's my match?

[Frank Petronio]

Those Carroll County farmers sure have big feet! ;-) Nice shots

[Henry Ambrose]

Jerry,
I answered your question number 1. Go back and read what I wrote.

And again, if you want the same perspective, you have to have the same distance from the camera to the object being photographed.

[Jerry Fusselman]

Yes, you came the closest to answering my original boldfaced question in your post (post #7). If you view this topic in threaded mode, then you will see that I already responded to your post. This is what I wrote (post #20):

Henry and David,

One quick question. Did you two give an answer to my question #1? Is it u, u+v, or something else? That is, are you two asserting that perspective is determined by the subject to front-nodal-point distance in agreement with my guess, or is it the subject-to-film-plane distance in disagreement with my guess, or is something else entirely?

I asked a question directly to you and the poster who agreed with you about perspective. I can rephrase my question: When you say "camera" do you mean "film plane" or "lens" or some other point? Probably if you say "lens," you mean the front nodal point---is this right? It makes a difference to me, because I am shooting near 1:1. When I first read what you wrote, I thought you meant film plane, but now I am less sure as to which location you mean.

[Jerry Fusselman]

Jim, do you recall some details of how you shot these? Any movements? Do you recall the taking apertures approximately?

[Henry Ambrose]

Jerry,

I wrote:
............... perspective is determined by:
1) where you stand
2) how far you set up the camera from the subject
3) the distance from the object being photographed to the film plane
4) object distance to the rear principal point of the lens

All being esentially the same location stated differently and in increasing levels of precision. Number 4 becomes difficult because the values change for different lens types and anyway are way beyond the precision needed for a portrait.

So, its:

4) object distance to the rear principal point of the lens

[Jerry Fusselman]

I see which post you mean now, and I see what you mean now, thanks! You are saying the exact answer to my question #1 is "object distance to the rear principal point". Thanks so much for being specific. But why the rear nodal point? That cannot be right, can it?

Let us look at an example. My 360mm Telomar has its rear nodal point about 130mm in front of its front nodal point, due to its strong telephoto design. There is just air there. A person inserting his eye at that point would see a quite different perspective than what is embodied in the rays of light hitting the front of the lens. It would be easy to construct an example with some object visible from the rear nodal point in front of the lens that is invisible to anyone standing at the lens. Do you see what I mean? This is why I do not think the rear principal (or nodal) point is correct.

[vijayn]

Jerry - let us define perspective first: Take two objects (for simplicity, two lines of the same length suffice) of the same size at two different distances from the image plane. Perspective is simply the ratio of the length of the images of the lines. This has nothing to do with the absolute magnification of each line; but rather is the ratio of those two magnifications.

Now, the important thing to remember is that every lens offers the exact same perspective, regardless of focal length, since the ratio of the magnifications cancels out the 'v' (or is it 'u'?) in the equation; thus the ratio of the images will be the ratio of the distances of the two objects from the lens plane, or more specifically from the front nodal point.

That is counterintuitive, I know, but true nonetheless. Thus, if you have an object at 100m from your lens and another at 200m, and both are of the same height (as in the optical term for linear size) then every lens, regardless of focal length, will form an image (assuming enough covering power etc) wherein the further object is half the height of the nearer.

So now I ask you this: if every lens yields exactly the same perspective, all that matters is the ratios of the distances of different objects from the lens, and the lens has no idea of what format is sitting behind it, why do we have to care about anything other than angle of view?

In other words, when you talk of "perspective" of an image, you are not talking of the mathematical definition of perspective, because that does not depend on either the focal length of the lens or the image format. You are, in reality, talking about angles of view to begin with. When you talk of angles of view, then you do have to double focal lengths as format size doubles.

Your fallacy is that you are trying to calculate perspective for two objects that are very close. However, remember that you cannot focus on two objects at different distances simultaneously; meaning that magnification can be calculated for one of the objects but not the other, out of focus object. This makes the issue of perspective irrelevant unless the distance of the objects is such that they are relatively close together and far away from the lens (mathematically, u>>f and u'>>f and u ~ u' i.e., u is approximately equal to u').

1. When you change to a 600mm lens, you get the same perspective as before provided the subject-to-front-nodal-point distance stays at 1.2 meters. I hope we agree that this is true!

Yes, this is exactly true, but it is true for all focal lengths, not only 600mm, and the idea of perspective is meaningful only when the objects are far away from the lens.

2. You say that when you change to a 600mm lens and use 8x10 film at that distance, you get the same percentage of the face on the film as before. Or stated another way, if you switch to 600mm and keep the same perspective as before, the correct film for capturing the entire image is 8x10.

This is approximately true for front-nodal-point to subject distances much larger than the focal length of the lens, much larger means at least 10 times larger. Thus for all distances above 3m this is more or less true. And you are correct that it is exactly true only when the lens is focused at infinity.

Hopefully this makes it clearer.

[John Berry]

Let's put another shovelful in the wheelbarrow. Lens choice MUST also take into consideration print size and viewing distance. ( If were going to jump in to the Nth degree, might as well go for Nth squared) Perspective envisioned in the shot will only be realized when the print is view at the right distance. Vijayn said much better than I, what I was thinking. Any lens will do when viewed at it's proper distance. I've shot the same head shot with a 24 and a 105 on 35mm. When viewed at 5 ft we can see how bad a choice the 24 was. Now take the 24 shot and put it 1" away from your nose. Perspective will appear much more acceptable. As for 8x10, Jim proves it very well, some of these rules of thumb are made to hit with a hammer once in a while. Other factors are How far in front of the backdrop do you need for the lighting you want and how much too small did you make said backdrop, forcing you into a longer lens.