What is telephoto compression (perspective compression)? Tom Axford 2023

The fact that you can shoot a scene with an ultra wide then enlarge a portion of the image to create "telephoto" compression is exactly why the term irritates me. But it doesn't actually matter in practice because anyone who wanted a compression effect would use a telephoto, not an enlarged wide angle, so pragmatically we might as well call it "telephoto compression". Still annoys for pedantic reasons.

It is both, because if you use a wide-angle lens and go further away from the subject, the depth in the scene will not feel compressed.

If you use a wide-angle lens and view the image normally, there is no apparent depth compression.

The camera position is irrelevant. Try it for yourself!

I have and there is. Off course there is - things that are further away look smaller than they do close up, that is what perspective is. Put you finger in front of your face - that little digit can block out a distant mountain. It doesn't matter what lens you use, far away things always look smaller than when they are closer.

The reason that there is no apparent compression in wide lenses is because the compression is present in objects the lens rendered so small you can't see it. But it is still there if you crop the frame to 1%. Tele lenses of course magnify the distant objects so the compression is easy to see.

Shoot a portrait with a 17mm lens but make the subject stand 30 metres away. There is no sign of any of the so called characteristic wide angle "Big nose" effect. The face is flattened even more than it would be with a traditional short tele portrait lens shot from a more traditional portrait distance. That's tele compression demonstrated by a wide angle. The downside is your subject will be quite small in the frame.

Exactly!

It's the magnification of the image that leads to the illusion of compression. If you magnify any image then the impression of depth in the image is compressed. It's the same when you look through binoculars. They magnify the image seen by your eye and that causes the illusion that everything is much closer to the viewer, the scene appears compressed towards the viewer.

I'm not sure if we are differing here on principles or semantics.

Perspective is the angular size of things diminishing with distance (things look smaller if they are far away). Most obvious in the railway lines converging to a vanishing point illusion. The tracks are in reality always parallel.

The compression illusion occurs when all the things in the field of view are far away. The fact that some things are marginally further away than others doesn't result in a visible reduction in size of the furthest objects because the additional shrinkage in angular size is small compared to the overall reduction in angular size of all the objects. In other words far away things appear to shrink, even further away things shrink more but only a little bit more and this surprises us giving rise to the compression illusion.

The magnification helps us perceive the compression more easily by making everything large enough to see the compression illusion clearly and removing distracting nearby objects from the field of view that would otherwise disturb the illusion. But it is not the magnification that is the primary cause, it is the relative distance between far off objects. The magnification is a kind of secondary helper than makes it clearer. That compression is still in the background in wide angle shots without zooming in, it's just too small to notice.

It works in reverse with conventional wide angle shots. You get really close to a foreground subject and it looks unusually large in the frame. Everything else in the frame is relatively further away than you think making secondary objects only slightly further away look surprisingly small compared to the huge foreground. You have created an expansion effect. But again, this effect happens because the relative distance between ultra close foreground and moderately distance objects is larger than we intuitively expect, not because of focal length. The wide focal length is a secondary support to the illusion because it broadens the field of view enough that we can still see objects further away than our dominant foreground object. Without the wide angle view, the slightly further away objects are cropped from the frame and the illusion collapses.

One thing I haven't tried is to use a standard lens to get really close to the foreground then expand the field of view using stitching. I would guess that would still produce the wide angle big nose effect even though you are using a standard lens. I'll have to test this.

I don't know what this means.
How far away? One foot, one yard, one mile, ten miles? How far?

An object ten miles away is one tenth the angular size of a similar object one mile away. That is exactly the same ratio as an object ten metres away compared to a similar object one metre away.

I really do not understand what you are saying.

At "big nose" distances, there will be too much parallax effect with a pupil of any size, and even with a pinhole, it will be difficult to maintain the exact same perspective for a quality stitch. A good stitch of a 3D scene requires that everything in the frames is many times farther away than the pupil size or the size of any perspective differences between shots.

The effect should be the same though, despite the Frankenstein stitching, because it is really about the ratio of distances of objects in the scene, at the empirical level.

Speaking of viewing distances and long lenses, did you know the Earth-Moon distance is only 133.35 meters when viewed through a Nikon 200-500 f/5.6 lens @ 500mm? A lot less than when viewed with the naked eye, which I'd guess about some 400,000,000.00 meters.

;-)

What I'm saying it that if your primary subject is 10 miles away and your secondary subject is 11 miles away and your tertiary subject is 12 miles away, they will all look roughly like they are all 10 miles away and you will get that cardboard cut out stacked compressed layered mountains look. The reason being that the 2 mile difference between the subjects is relatively small compared to the average 11 mile difference between them and the camera and at that distance the brain doesn't have a way to figure how big they are anyway. Hence the illusion. I think for short distances, the brain typically knows what size the objects are and the illusion fails. Of course, this ought to be demonstrated through testing.

Talking of illusions, is it just me, or does this picture provide the illusion that the statue on the left is wearing headphones (it isn't)?

I find this terminology extremely confusing. The viewer sees the world through the camera, so the camera's perspective is the viewer's perspective. Do you conflate "viewer's perspective" with angle of view/focal length? Do you believe that changing focal length moves the viewer through space?

'Perspective' is 'from the point of view of the viewer'. It doesn't matter if you use any kind of equipment in front of your eyes or not.

The viewer's perspective is how the viewer sees the image.

Try this test. Print an image at 12" x 8". View the print from around 14" distance with your eyes in line with the centre of the image. That is viewing it normally. If the shot was taken with a normal focal length (40 to 50mm FFe), then you see the image with the natural perspective. This means that the image you see is the same size and shape as the image you would see if you viewed the original scene from the camera position.

If you move closer to the print, the image you see is enlarged. If you move further away from the print, the image you see is reduced in size.

So changing your viewing distance (from the print) changes the viewer's perspective. Of course, what the viewer sees is a combination of the camera's perspective (as captured in the print) and the viewer's perspective (viewing distance relative to the size of the print).

Suppose you take a shot with a wide-angle lens, say 18mm FFe. If you view that image normally (from a distance equal to the length of the diagonal), then it may show obvious wide-angle distortion (depending on the subject matter).

However, if you view that image from a distance equal to about 40% of the length of the diagonal, then the image will be seen with its natural perspective and show no wide-angle distortion. When doing this test, you should make sure that your eye is directly in line with the centre of the image.

If you have never done this test, I urge you to do so. It is very instructive and quite surprising to see the wide-angle distortion completely disappear when your eye is in the right position.

Ahh, this I understand. Moving the viewer's eyeball closer to the print/screen is what you are referring to as "viewer's perspective". In that case, your earlier assertion that the dolly zoom changes both the camera's perspective and the viewer's perspective is wrong. A dolly zoom does not grab the viewer and force their eyes closer to or further from the screen.

Sorry, but in my haste to give a quick and brief answer, I omitted to say that the viewer's perspective is their viewing distance from the print/screen relative to the size of the image and the focal length of the lens used.

To see the print/screen with the natural perspective the viewer's distance must be the same as the focal length of the lens multiplied by the print size to sensor size ratio. In other words, the viewing distance relative to the print size must be the same as the focal length relative to the sensor size.

An alternative (and perhaps easier) way of thinking about this is that the angular size of the image seen by the viewer must be the same as the angular size of the image captured by the camera.

If you view an image from a fixed position and zoom the focal length of the lens, that effectively zooms the viewer's perspective as well. It magnifies the image they see.

Let me quote a few excerpts from " The Manual of Photography" (7th Edition, 1978):

The Manual of Photography, from which these excerpts are quoted, was first published in 1890 as The Ilford Manual of Photography. Early photographers had a very good understanding of perspective. It is sad that perspective is so poorly explained by many modern tutorials on photography.

Wouldn't it all just come down to x y and z with perspective, angles, depth and being able to just control z as a floating point.

Danny.

I don't understand the point you are making. Please elaborate.