lens perspective question?

[Bob Salomon]

Sorry, Bob, that is a misconception that many very knowledgeable photography instructors spend their lives trying to correct. With so many photographers at differing levels of experience these days, the term perspective is used to mean whatever any person is trying to say. What YOU are doing is changing the composition. True perspective, the relationship among elements within a scene, is a factor of camera to subject distance.

And in photo school in the early 60's that was the explanation of perspective at that time. We were not only explained it, we were assigned to illustrate it.

But that was in the dark ages compared to most experiences today. Like the definition of Pluto not being a planet. Anything can change over time! In sixth grade Mrs. Vetter took the better part of a morning one spring day in 1954 to make sure that we all understood why man could never go to the moon!

[Ken Lee]

Bigger film/sensor requires longer lenses. Longer lenses - at the same aperture - have less depth of field than shorter lenses.

How much less ? When we double the focal length, we have to stop the lens down by 2 f-stops, to get the same depth of field.

For a 100mm lens to match the depth of field of a 50mm lens at f/16, we need to shoot the 100mm lens at f/32.

[SergeiR]

And in photo school in the early 60's that was the explanation of perspective at that time. We were not only explained it, we were assigned to illustrate it.

But that was in the dark ages compared to most experiences today.

Meh. This is exactly how i still illustrate it to me students any time i see them shooting full body or 3/4 portrait from the shoulder level.
That said - i cant for the life of me remember how we been taught perspective in art school and later in school. Merciful nature wiped most of that stuff out.

[rdenney]

Bob, I learned perspective as a draftsman, not as a photographer. Perspective is the relationship in size between near objects and far objects. That is affected only by the distance between the observer and those objects.

Perspective exists in all three spatial dimensions, of course. When you raise your position, you have changed your observation point in at least one of those dimensions, and that will change perspective.

Now, when we project that three-dimensional space onto two-dimensional film, what we interpret as perspectives shows up as converging parallel lines.

We can distort those converging lines by tilting the film with respect to the scene, and if we are clever, our distortion can cancel out the convergence. When we turn the camera, we are tilting the film with respect to the scene, and it alters the way those lines appear to converge. We call that a change in perspective, but it is really just a trick that has the appearance of a change in perspective.

Convergence happens only when three-dimensional space is projected onto a surface. It does not affect the apparent size of near and far objects from an observer's location.

Now, on the depth-of-field thing.

If we take a 300mm lens and put it on an 8x10 camera, and make a photo of a person's face such that it nearly fills the frame, we have a full-face portrait with a given depth of field. If we take that same lens (along with the same aperture setting) and put it on a 35mm camera (using, say, a bellows), we will get exactly the same depth of field and perspective, if both are printed to the same magnification.

But the picture from the 35mm camera will only include the subject's nose, not the whole face. And if we print to the same magnification, the print from the 8x10 camera will be over 8 times bigger in width and height than the print from the 35mm camera. If we enlarge the 35mm picture to the same print size, the depth of field will actually be less. But that nose will be a lot bigger.

When we change the lens on the 35mm camera to provide a similar view as we got using the 300 on the 8x10 camera, we might end up using something like a 45mm lens. That will give us that full-face framing. Since we put the camera in the same location, the perspective will be the same. But the 45mm lens has greater depth of field than the 300mm lens, when both are printed to the same size.

Since we usually base comparisons on similar subject framing shot from similar locations, the focal lengths will be different. Thus, for convenience, I approximate equivalent apertures (for general depth-of-field effects) as being two stops per major format change. Going from 35mm to 6x9 requires two stops smaller to achieve the same depth of field. Going to 4x5 requires two more, and going from there to 8x10 requires two more. Like all such approximations, it's not precise. But it provides a starting point.

Rick "noting that a 300mm f/5.6 lens on 8x10 can produce an effect when used wide open not possible on 35mm with any available lens" Denney

[rdenney]

Meh. This is exactly how i still illustrate it to me students any time i see them shooting full body or 3/4 portrait from the shoulder level.
That said - i cant for the life of me remember how we been taught perspective in art school and later in school. Merciful nature wiped most of that stuff out.

You are actually changing the camera position with respect to the vertical dimension when you go from sitting to standing. So, you are changing the perspective relationships, but only in the vertical dimension.

This is one of the key changes in event photography between 50 years ago, when Rolleiflexes and Hasselblads--used at waist-level--ruled such photography, and 25 years before that, when press cameras were mostly used at eye level, and now, when DSLR's are used at eye level.

Rick "everyone is right" Denney

[Heroique]

...Now, when we project that three-dimensional space onto two-dimensional film, what we interpret as perspectives shows up as converging parallel lines. We can distort those converging lines by tilting the film with respect to the scene, and if we are clever, our distortion can cancel out the convergence. When we turn the camera, we are tilting the film with respect to the scene, and it alters the way those lines appear to converge. We call that a change in perspective, but it is really just a trick that has the appearance of a change in perspective...

Can everybody please note this important point – it will eliminate a lot of pain in future discussions.

Rick is describing a “change in perspective” that really isn’t a change in perspective – that is, rendering converging lines straight, as one does w/ back tilt (for vertical lines) or back swing (for horizontal lines).

I think one of our longest-ever threads – about perspective, of course – suggested that we call this type of perspective “geometric perspective,” to distinguish it from “viewpoint perspective,” the other kind that changes with camera/lens position.

I think Doremus came up w/ those terms, a good idea.

[Steve Smith]

Not true if the final prints are the same size.

I disagree with that statement.


Steve.

[SergeiR]

You are actually changing the camera position with respect to the vertical dimension when you go from sitting to standing. So, you are changing the perspective relationships, but only in the vertical dimension.

Projection control. Nowdays people like to call it "perspective's distortion"

[Mark Sawyer]

Not true if the final prints are the same size.

I think you're off on this one, Dan. One can't change the depth of field in a negative by printing it larger/smaller.

[rdenney]

I think you're off on this one, Dan. One can't change the depth of field in a negative by printing it larger/smaller.

You don't change what's on the negative, but you sure do change the size of the maximum circle of confusion that defines what is acceptable depth of field.

Rick "but more concerned about gross effects than fine effects when comparing aperture effects across formats" Denney