[Tokina]

So the statement "A wide angle lens will stretch a person's face in an unflattering manner, and a telephoto lens will flatten the face in a flattering manner." is not "simply inaccurate", although the statement makes unspoken assumptions and is therefore subjective.

It's still false cause. It's the distance that "distorts" perspective, not the lens, and keep in mind that accurate perspective doesn't exist. There might be distances from where people have been used to see things and those might be called "accurate".

[audiobomber]

I think you should do that test yourself. Ratios in pictures don't change if taken from the same spot.

I will do the measurement test I proposed when I have time.

There's no need for me to do the test. It's demonstrated in this article, complete with photos and a better explanation than I could manage. Perspective distortion (photography) - Wikipedia, the free encyclopedia

[Tokina]

That's incorrect. Perspective is very real and has been understood in art for millenia.

Calling a perspective "flattering" or "weird" is a subjective opinion. But calling a perspective "accurate" is silly. A certain angle of view may simulate human vision, but that's a whole different thing than perspective.

[audiobomber]

Calling a perspective "flattering" or "weird" is a subjective opinion. But calling a perspective "accurate" is silly. A certain angle of view may simulate human vision, but that's a whole different thing than perspective.

We're discussing "normal" lenses and perception distortion. What else would you use as normal if not human vision? An image is accurate if it shows what the human eye sees. If the background appears further away in an image than it does in real life, then it is not normal, it has distorted perspective. Seriously dude, do some reading.

I am not talking about "angle of view" because that has nothing to do with "normal" lenses. It's the perspective that makes it a normal lens, not FOV.

---

Last edited by audiobomber; 12-03-2008 at 12:56 PM.

[Tokina]

If the background appears further away in an image than it does in real life,

It never does! But a wide angle lens makes you walk closer! Walking changes the perspective, not the lens! That's the whole point

I am not talking about "angle of view" because that has nothing to do with "normal" lenses. It's the perspective that makes it a normal lens, not FOV.

Okay, now I will give up on you

[audiobomber]

It never does! But a wide angle lens makes you walk closer! Walking changes the perspective, not the lens! That's the whole point

No. There's no movement involved. The perspective changes due to focal length only.

Here's another good article. Even if you don't read it, at least scroll down and look at the pictures that show wide angle and telephoto perspective distortion. kevinwilley.com - Topic 4 - Telephoto Lenses

[Michael Barker]

It's the perspective that makes it a normal lens, not FOV.

Audiobomber, I actually think both FOV and perspective have something to do with it. When you're really close to someone, you can have trouble looking at their whole face at once, but the wide angle lens has no trouble at all. Which means the wide-angle lens at that distance is giving you an unnatural field of view of someone's face.

As for "flattering" and other descriptions, Tokina, these are empirical fact and should not be thrown away just because they are subjective on a person by person basis. It's what's called a rule of thumb, and very useful. After all, people who take photos of people for a living use the facts about an average person's perception all the time to their advantage.

[Michael Barker]

No. There's no movement involved. The perspective changes due to focal length only.

So, you really are confused, audiobomber.

Here's another good article. Even if you don't read it, at least scroll down and look at the pictures that show wide angle and telephoto perspective distortion. kevinwilley.com - Topic 4 - Telephoto Lenses

The camera was moved in these shots, quote "I couldn't get all the way down to the 12mm focal length, because I was practically on top of the truck as it was at 15mm. At 550mm I was on the neighbor's land, and I was a long way from the truck and the barn."

Perspective changes as a result of focal length only when the focal length causes the photographer to move.

[Michael Barker]

It's still false cause. It's the distance that "distorts" perspective, not the lens, and keep in mind that accurate perspective doesn't exist. There might be distances from where people have been used to see things and those might be called "accurate".

Accurate perspective doesn't exist, that's true, but normal perspective does. There is a range of what is normal, but normal for the "average" person can probably be measured to a tight range. It's not normal to have eyes in the back of your head, for instance, or to see everything with a "God's eye view" in which Shaquille O'Neal and Richard Simmons both look like ants.

[audiobomber]

So, you really are confused, audiobomber.

The camera was moved in these shots, quote "I couldn't get all the way down to the 12mm focal length, because I was practically on top of the truck as it was at 15mm. At 550mm I was on the neighbor's land, and I was a long way from the truck and the barn."

Perspective changes as a result of focal length only when the focal length causes the photographer to move.

Oops, I just scanned that article quickly. I guess I'll have to do the test now, with no walking, just lens changes.

To recap, I say that the relative size of the foreground and background images will change with focal length, i.e. that perception distortion is solely due to FL, and that that's what defiines a normal lens. For a normal lens depth perception (perspective) is neutral (i.e. same as human vision).

[Michael Barker]

I did that test on the first day I had a zoom lens. I was shocked to find that distance is the only thing that affects the relative sizes of objects that are near vs. objects that are far. Shocked because I'd always heard people say things like "that pic makes his foot look big because it was taken with a wide angle lens", which is a statement that contains only part of the truth.

[Tokina]

Accurate perspective doesn't exist, that's true, but normal perspective does. There is a range of what is normal, but normal for the "average" person can probably be measured to a tight range. It's not normal to have eyes in the back of your head, for instance, or to see everything with a "God's eye view" in which Shaquille O'Neal and Richard Simmons both look like ants.

A wider angle lens just shows more from the same perspective.
Perspective means the relations between things as they appear to be from a certain point. The angle of view limits the appearance if it's not full 360 degrees. I guess an image of 360 degree can correctly be printed only on a ball. Do such cameras exist?

Having two eyes on the neck too would be interesting and then there would be four different images from different perspectives combined by our brain. That would be weird

[Wheatfield]

Well, of course it means something. I don't see how you can say it doesn't. "The field of view of the eye" can mean entirely different things depending on whether one considers peripheral vision or not. And even if you decide it doesn't, it's still an *extremely* subjective determination as to where "normal" vision stops and where "peripheral" vision starts.



Only if you insist on defining "normal" as "equivalent to the FOV of the eye". If you prefer, you can define it as "equivalent to the FOV of the non-peripheral portions of the eye", but that's simply not something you can pin down to a specific number. And while sure, you *could* define "normal" that way, there are several other "definitions" of normal being proposed here that are similarly imprecise. The so-called "normal" field of view is sort of a rough approximation of a lot of things - it "sort of" resembles the non-peripheral field of view, it "sort of" provides a field of view that, when printed at "typical" sizes and viwed at "typical" distances, resembles the view of the print itself, and it "sort of" provides a decent compromise between wider and longer focal lengths if you were forced to use only one focal length. But imagining that all these different possible definitions would all converge on exactly the same exact focal length: in particular, either the diagonal of the film/sensor, or perhaps 50mm/33mm if you prefer - is just ludicrous. The different "definitions" of "normal" I gave are all very inexact and broad.

Well Mark, you are going up against several generations of optical engineers, photographers and every camera manufacturer from Alpa To Zeiss and telling them all that they were wrong.
Carry on if you like, but do read up on the subject of peripheral vision to find out why it is meaningless from a photographic point of view.

[Wheatfield]

The view through my viewfinder at about 55mm (roughly 80mm full-frame equivalent) shows objects in the focal plane at roughly the same size as my naked eye.

This has to do with the viewfinder magnification. If you repeat your "test" with a 35mm film camera, you will find the same thing as you found with your DSLR.
By rights, because the digital format is smaller, viewfinder magnification should be increased to compensate.
The camera companies have chosen no to do this, so now we are saddled with finders that are too small.

[Michael Barker]

To Tokina: I think we agree there is a normal angle of view.

The point I was trying to make is, there is a concept of normal perspective also, it's just not as well defined, in fact it's really hard to define. There are certain distances at which we normally interact with things. If you want to be artistic, it's great to throw out all ideas of "normal", but there has to be a normal to throw out. And there's also something to be said in art for showing people a situation as it would look to them if they were there.

If there weren't such a thing as normal perspective, we wouldn't have the freedom of applying abnormal perspective to obtain an effect, like the humourous effect Barry Sonnenfeld gets from stretching people's faces. Is it funny "weird", or funny "ha ha"? In either case it's not normal, and he has to use wide angle lenses to do it, out of sheer practicality. If he had the freedom to change the size of the actors, or show only part of their face, he wouldn't have to use wide angle lenses, but he does, and so I believe, in practice, that wide angle lenses have a certain perspective attached to them. Feel free to unattach it at any time, but there are consequences.