Lens Compression Doesn't Exist

Recently I've concluded that most professional photographers actually believe that telephoto lenses compress the subject and background in an image and that simply is not true. 

 
I got a call from a local photographer last week. He knew that I had recently filmed a tutorial on architectural photography with Mike Kelley and he had a question about lens distortion. While he was photographing a home interior, another well known and successful architectural photographer in the area approached him and offered to give him his "secret technique" to produce distortion free images. Instead of shooting one vertical 17mm image (which has significant distortion in the edges because it's so wide), he suggested that my friend shoot three horizontal shots with a 50mm lens while tilting the camera from top to bottom and stitching the files into one photograph. My friend was shocked by the simplicity and potential of this suggestion and called me to confirm it. "Is it really possible to get distortion free images this way?" He asked. No it's not. In fact the images will have an identical field of view. 
 
This was impossible for my buddy to believe so he had to put it to the test, and after comparing the two shots he found that I was right. The wide angle shot looked identical to a stitched shot created with multiple images from a longer lens taken from the same location.
 
After the test he called me to tell that the distortion was indeed the same but he was still convinced a longer lens would produce more "lens compression" meaning that the foreground and background objects would appear closer. I then blew his mind for the second time; "You realize that if you take a portrait of someone with a telephoto lens and then take the exact same shot with a wide lens from the same spot but crop into that image, the 'compression' will look identical," I said. My buddy simply could not believe this to be true. He had always been taught that the lens was doing the "compressing." After quizzing a few other photographers I realized that the majority of professional photographers I know didn't understand this either. 

Lens compression has nothing to do with the lens and has everything to do with your distance from the subject. 

 

 

I explained to my buddy that lens compression doesn't happen in the lens, it happens when you physically move the camera back. No matter how I explained it he wouldn't believe me until he did the test himself. 
 
 

Why do we use the term "lens compression?" 

 
Most photographers believe that the lens is "compressing" the scene because they will take a test shot with a wide lens, and then back up to capture a similar composition with a telephoto lens. It's easy to see that the telephoto shot is far more "compressed" and it's also easy to completely overlook the fact that the camera has been moved and the field of view has changed. 
 

Why does compression happen?

 
Imagine you are photographing a person and behind that person is a mountain. Let's say your subject is one foot away from the camera and the mountain is one mile behind your subject. From this distance the person's face would look huge compared to the mountain because the person is exponentially closer to the camera than the mountain. 
 
To make the person half the size that they currently are in the frame, you would have to move back only one foot, doubling the distance of the camera from the subject. But when we half the size of the person by moving back one foot, the size of the mountain remains almost identical because to make the mountain half the size you would need to move the camera back one mile. This ratio between the distance of the subject and background to the camera is what causes compression. 
 
As you continue to move backward your subject will get smaller and smaller and the mountain will remain about the same size. To keep the subject the same size in the frame, you will need to zoom in using a lens or "crop in" in post and as you do, you are enlarging both the person, who has been shrinking as you back up, and the mountain, that has remained basically the same size. 
 
When we look at the final images it may appear that the person is the same size in each picture while the mountain is growing but in reality the opposite is happening. The mountain is staying the same while the person is shrinking and we are zooming in to compensate. 

 

Here's Proof

 
In this gif below you can see two images that Elia Locardi took during the filming of his landscape photography tutorial. In both shots, the camera was in the exact same place on a tripod. The image on the left was taken with a 16mm lens and the image on the right was taken at 145mm. Notice that when we "crop in" or "digitally zoom" the wide shot, we can get an almost identical looking image to the shot taken at 145mm with the exact same "compression."
 
 

Conclusion

 
Compression is a real thing but it really doesn't have anything to do with a lens. It has everything to do with your distance from the subject/background. You could get a shot with identical compression if you stand in the same location and take a photograph with a telephoto lens, a wide angle lens but crop in, or a camera with a "digital zoom." The magic isn't in the lens, it's in simply backing up. 
 
And all of this leads back to the very beginning of the article. Is it possible to create a distortion free image by stitching multiple telephoto images rather than using a wide angle lens? No, because the distortion/compression is caused by the location of the camera in relation to the subject and not by the lens. If you stitch multiple images you'll end up with a higher resolution shot that has identical compression. 
 
This does not mean that telephoto lenses are obsolete. Long lenses allow us to back up and zoom in without losing resolution; everyone knows that. But this post simply explains the phenomenon of compression in general which many photographers do not fully understand. 

Take the poll

In talking with my photography buddies about this I've discovered that most photographers do not seem to know this. I'm curious how many of you reading this also didn't know that lens compression doesn't really exist. Take the poll below and let's find out. 

Log in or register to post comments

108 Comments

Asutosh Bansal's picture

Third Option for the poll is the best one.

Richard Munro's picture

Good read

Ingemar Kenyatta's picture

Indeed.

Alex Stokic's picture

Now this is a very nice post! Have been following this site, but now I had to register because of this one!
From my perspective, anyone who plans going into phorography more seriously should at least learn basic physics and optics as well, not to mention electronics and how every component of your gear works.
Nice one Lee!

Peter Blaise's picture

"... from my perspective ..."

I see what you did there.

Thanks for that!
.

Jeff P.'s picture

[glass shatters]...

So...all of those that say that a 35mm on a crop sensor gives less compression than a 50mm on FF are wrong?
I wouldn't have believed it if I hadn't seen your test!
Great read!

Lee Morris's picture

If the cameras are in the same spot then the compression will be the same. The depth of field will not be though. In the example I used above everything is in focus so it's a non issue but telephoto lenses will produce shallower dof compared to a wide shot cropped in.

ALEXANDER TARDIF's picture

Unless, of course, you are using a faster WA on your crop sensor that offers a similar DOF as the longer FL glass on a FF. Such as, say, 50 1.8 on FF next to 35mm 1.4 on aps-c... the "look" you get will be similar enough that you could reasonably call DOF equivalent. Not the angle of view, but the DOF.

Good article, Lee, I got into this very same discussion last weekend at an air show that I was shooting with a 200mm next to guys with 600mm who claimed to have better "grouping" of Blue Angels because of "compression" their longer teles offered. We were shooting from the same spot ;)

Lee Morris's picture

"I got into this very same discussion last weekend at an air show that I was shooting with a 200mm next to guys with 600mm who claimed to have better "grouping" of Blue Angels because of "compression" their longer teles offered. We were shooting from the same spot ;)"

That is a fantastic example of a real world situation where this information is useful.

Jeff P.'s picture

yeah I knew the depth of field would be different, glad you didn't shatter that glass too ;) haha

It's nice to know the theory! Still, on a crop, you need to get further away to get the same shot as a FF camera but at least we know it's the distance, not the lens, that affect compression.

Peter Blaise's picture

In response to [ Jeff P. ] who suggests that we have to move to get the same shot when using different size sensors.

No, we cannot get the same shot by moving.

We can only get the same shot from the same position, by not moving.

If we move, we get a different shot.

There are reason to move, and there are reasons to stay.

Moving gets different shots.

Staying gets the same shots.

The only subject that will not change ( much ) when moving ( on axis ) is flat copywork.

For all three-dimensional subjects in three-dimensional surroundings, moving will get a different shot of the supposed main subject, a different shot of their surroundings, and definitely a different relationship between them and their surroundings

Thanks for exploring this and sharing.
.

Usman Dawood's picture

The angry photographer on YouTube is trying to dispute your article (not doing a great job to be fair).

Lee Morris's picture

Link?

Usman Dawood's picture


I think he's misunderstanding the fact that although compression does exist LENS compression does not.

Lee Morris's picture

I responded on his video. Not sure it matters.

Usman Dawood's picture

I agree it doesn't matter I just found it funny lol.

Peter Blaise's picture

Paraphrasing [ Lee Morris ] "... the depth of subject field focus is lens-focal-length dependent ..."

Oh please do not go there.

All that different lenses do is change magnification in-camera, essentially cropping or stitching in camera instead of in post-capture-processing.

A different focal length lens cannot do anything else but change the magnification of the existent view of the subject scene.

If a lens changed anything other than magnification, no one would manufacture it, because no one would buy it, because it would be distorting.

Thanks for exploring this and sharing.
.

T Dillon's picture

Totally true. This is the article everyone needs to read. Its why a FF 35mm can be a great choice for portraiture. Just make sure your distance is the appropriate.

ALEXANDER TARDIF's picture

Since I'm the mood to "add" to others' comments tonight... I'd say make sure your distance is appropriate BUT ALSO your subject placement is too: you place your subject towards corners and you've got other things in perspective distortion to worry about more with a wider angle focal lengths.

Michael Aubrey's picture

I understand what you're saying, but you should really include more discussion of magnification and it's relation to both subject distance & FL.

Sure, you can create a compressed image with a wide lens, certainly, but you can't do the reverse. A long focal length constrains a subjects magnification (and the magnification of the background) in a way a wide focal length does not.

The lens' focal length still comes into play. If it didn't, you'd be able to go the other direction--create the minimal compression look of a wide with a long one. But, obviously, you can't.

Lee Morris's picture

I'm not sure I totally understand your point but you can create minimal compression look of a wide lens with a long one by shooting multiple images and stitching like the first example in this article.

Michael Aubrey's picture

Fair point. That's not something I can easily do with film. I can certainly crop film. And I can make panoramas with separate frames and hang them side by side.

Michael Aubrey's picture

I should add though:

You've chosen to make FL the constant and you've written around that by saying that FL doesn't matter because you can change image format/sensor size (cropping/stitching).

You could take this very same article and just flip the script and say "image format" doesn't matter and then just phrase things in terms of focal length and point out for that for a given sensor size, your only option to change compression is by changing FL in relation to subject magnification. You're still talking about the same things, but now you're framing the issue by keeping a different variable constant.

The question you need to ask is:

"Why should image format/sensor size get priority over FL when either option affects angle of view relative to magnification?"

Surely choosing between them is arbitrary!

Karim Hosein's picture

Not at all. When learning photography, we are taught perspective choices, (where you stand, relative to your subject & surrounding objects), and framing, (what focal length lens is chosen, or what is cropped latter in the darkroom).

Perspective is decided on strictly by where one stands, and has no bearing whatsoever on focal length nor film/sensor size/aspect ratio. film size, aspect ratio and focal length affect framing, but not perspective. Subject/object compression is all about relative sizes, and is only dependent on where one stands.

Michael Aubrey's picture

...

Peter Blaise's picture

Thank you, [ Karim Hosein ] for this superlative quotable [ with edits ]:

"... When learning photography, we learn
perspective choices,
-- where we stand, relative to our subject & surrounding objects,
and framing,
-- what focal length lens we use, or what we crop[ or stitch ] latter in the darkroom [ or lightroom ].

Perspective is decided on strictly by where we stand,
and has no bearing whatsoever on focal length nor film/sensor size/aspect ratio.

Flm size, aspect ratio and focal length affect framing,
but not perspective.

Subject/object compression is all about relative sizes,
and is only dependent on
-- where we stand [ when taking a picture
-- versus where we stand when viewing a presentation of that picture ] ...".

- - - - -

Yes, you get it, and you give it:

- perspective and framing are unrelated,

and each is accomplished by different, unrelated means.

Thank you.

It's so nice to know that some people actually do understand this, and are willing to share, in spite of the thoughtless naysayers.
.

Patrick Hall's picture

I'm pretty sure this is the exact scenario that started the whole conversation. You can make a wide angle shot by stitching multiple images shot at a more telephoto focal length. And when you do, you should have the same distortion around the edges if you can get your software to stitch them in an orderly fashion.

Michael Aubrey's picture

Ah...my bad.

Not something us film shooters would have thought of!

Peter Blaise's picture

In response to [ Michael Aubrey ] writing "... Not something us film shooters would have thought of! ..."

It has nothing to do with film.

It has only to do with remembering that
all a different focal length lens does
is change magnification.

If anything else changes,
then the lens is bad, broken, defective.

Thanks for exploring this and sharing.
.

Michael Aubrey's picture

Hi Peter, that wasn't the (four year old) point.

Peter Blaise's picture

Ah, THIS is "Dental Photography", in that getting to the point is like pulling teeth:

[ Michael Aubrey ] wrote "... Hi Peter, that wasn't the (four year old) point ..."

So ... the (four year old) point was ... ?

As I re-read https://fstoppers.com/comment/281035 ( and notice dates, oh my, but it could be 1890 and still be equivalently meaningful ), I see the lack of understanding of what perspective distortion is, the lack of understanding that all that changing magnification in-camera does is crop in-camera ... or call for stitching in post-capture processing if the photographer wants more rather than less ... but it's all pieces and parts of the exact same picture nonetheless, the participation of different focal length lenses does not change the capture qualities of each piece and part of the exact same picture.

Perceived perspective compression ( or expansion ) is a system observation, an observation made by the observer of the presentation image, not an observation made by the photographer during capture, it's a product of the total capture-and-reproduction system, where if there is any difference between the photographer's angle of capture versus the angle of reception / perception by the viewer of the presentation image, then there may be the perception of compression or expansion, but if the angles are the same, there is no compression or expansion.

In other words, the only reason folks think a telephoto lens compresses perspective is because they are looking at a presentation image from a different perspective than it was taken from, and if they look at the presentation image from the same perspective that the image was taken from, there would be no perception of perspective distortion.

Perspective distortion is in the eye of the beholder, not in any lens.

My point is that the "... (four year old) point ..." was in error.

Does that seem to address this and move it forward?

Thanks for exploring this and sharing.
.

Mr Blah's picture

Soooo how do you explain this massive stitching giving a wide angle feel yet clearly taken by many "zoomed in" lenses?

http://gigapan.com/gigapans/15374

Rob Mynard's picture

What you're noticing in stitched telephoto images (often referred to as the Brenizer method) over a wide angle lens is the difference in depth of field, not compression.

Mr Blah's picture

I was refering to his comment: "The lens' focal length still comes into play. If it didn't, you'd be able to go the other direction--create the minimal compression look of a wide with a long one. But, obviously, you can't."

you clearly can create the "wide" look from a telephoto and a "tight" look from a wide one.

Michael Aubrey's picture

Ah...my bad.
Not something us film shooters would have thought of!

Steve Seeley's picture

Regardless of the semantics, the results are the same. This knowledge isn't going to replace the need for a wide or long lens to get a specific look.

That said, the more you understand the physics of photography and how your equipment works, the better.

Karim Hosein's picture

« This knowledge isn't going to replace the need for a wide or long lens to get a specific look.»

But that is the point of this article, that people are being told to use a different lens and stitch for a different look, when they end up with the same look, had they used the correct lens. Changing lenses and standing in the same spot DOES NOT change the look, no matter how much one stitches or crops. The look is dependent on position only. Lens choice affects framing (until one decides to stitch or crop).

Deleted Account's picture

Lee, you're right as long as the camera position and lens axis doesn't change. Here's an old diagram related to Alberti's window:

http://media.tumblr.com/tumblr_l34tqtKrO61qafpe7.jpg

In the diagram, the "center of projection" is the same as the camera position and the "principal ray" is similar to the lens axis plane. As long as the camera position and lens axis plane stay the same, then it doesn't matter what focal length lens is being in terms of maintaining perspective. However, if a photographer changes camera position (even without changing the lens axis plane) then he alters the distance to subject and the angle of the "projecting ray" at the bottom of the diagram. The angle of the projecting ray is what causes the sense of compression when changing the distance to the subject.

Basically, what I'm saying is that you're right and the part that confuses most photographers is related to the angle of the projecting-ray/camera-position rather than the principal-ray/camera-position.

Peter Timmer's picture

You are right.. if you stay in the same spot relative to your subject there will be no more compressing at lets say 70mm than 24mm in this case it's only about framing. Only when you move yourself or the subject to have an equal size relative to the frame will you see more compression at 70mm than 24mm.

Basically this means if we'd have unlimited megapixels there would be no need for telephoto lens. We could just crop everything and it would have the same effect. However the other way around is a bit more difficult because it's pretty difficult to shoot a 4 by 5 room with lets say 200mm lens. It's a lot of work stitching all those files together and there'd be a problem with getting your lens to focus close enough.

Ryan Cooper's picture

I've always felt like compression was a bad way of defining it too. Just like how I hate when people say sensor size impact depth of field. It doesn't. Distance to subject does and the bigger sensor just lets you stand closer at the same focal length.

That said, in my opinion, the benefit of stitching a telephoto to create a wider "look" is that it helps reduce wide angle distortion and also lets you shoot at a much narrower depth of field than a wide angle would allow.

Deleted Account's picture

I maybe 100% off-base with this comment but it sounds like you are telling how to compensate for compression (by altering distance from subject) rather than saying there is no compression. Changing the variables of distance to subject and distance to background and saying you get the same result is like saying I can make 10+10=32 by changing one of the 10's to a 22 instead.

Karim Hosein's picture

« Changing the variables of distance to subject and distance to background and saying you get the same result ….»

But he never said that. What he said is if the distance does not change, the result does not change, (regardless of lens, sensor, etc.), but if you change lenses and also change distance, then, because the distance changed, the compression has also changed.

The apparent compression between subject and object is a matter of the ratio of the distance from observer to subject, and observer to object. If the subject is 10m away and the object is 20m away, it would have the same amount of “compression” as when the subject is 20m away and the object is 40m away, (because 10:20 = 20:40 = 1:2 same ratio). That is, the object would appear to be twice the distance away.

In the case of the man and the mountain, one foot from the man, a mile from the mountain, and the mountain appears to be 1,609÷0.294=3,636 times the distance away. (One foot =0.294m, one mile = 1,609m). Moving back one more foot, we now have the mountain appearing 1,609.294÷.588 = 2737 times the distance away.

To get a real change in perspective, by moving 100m from the man, we now have 1709÷100 = 17 times the distance away.

To put it differently, by moving from one foot to two feet from the man, he now appears twice as small, (2÷1=2), while the mountain appears about the same size, (1609.294÷1609=1.0001872 times smaller, a change of less than 0.019%). moving 100m away, the man now appears 340 times smaller, (100÷0.294=340.1), while the mountain now appears 1.062 times smaller, (1709÷1609=1.062), or still about the same size, (at about 6.2% smaller).

All of this is regardless of the lens/sensor/post-crop, et al.

Anonymous's picture

Or, in other words, if you frame the subject the same (moving closer for a short lens or further with a long) and your DOF matches then a photo taken with, say, an 85mm and an 11mm will look identical.

Karim Hosein's picture

Nope, you got that wrong, unless your subject was flat, and there are no foreground/background objects). Once you move, perspective has changed, (compression has changed, etc.).

What he is saying is if you contact print a 10×8 inch negative, shot with a 300mm f/28 lens, and stayed in the same spot and shot the same scene with an APS-C camera with a 30mm f/2.8 lens, and enlarged the image to 10×8, then not only will you get the same perspective, (because you did not move), but the same framing (because you had the same FoV), and the same DoF, (because you had the same aperture diameter).

Peter Blaise's picture

OMG we think as one!
.

Craig Staples's picture

In your animated gif of the guy in front of the tree, when using the 200mm lens the tree is covering the entire frame behind your subject as opposed to the 24mm where it is barely visible behind him.
Because you had to use the long lens to achieve this look, even if it is because you also had to move the camera. This effect is caused because you are using a long lens creating a "compressed" distance between subject and background.
So while it may not be the actual lens doing the compressing, you can't say the lens has no affect on the equation.

Deleted Account's picture

and just as exaggerated are the facial features and brim of the hat the subject is wearing. That is the practical side of "compression"

Lee Morris's picture

Did you read the whole article?

Craig Staples's picture

Yup, even the part of your article that says "Compression is a real thing but it really doesn't have anything to do with a lens"
And as I said. The lens may not create the compression but it does have something to do with the final effect.

Mark Davidson's picture

Actually the lens only has the effect depending on its position.
It's final effect is cropping.
Zoom from one position and crop to identical size (as in the illustration) and the image has the same "effect".

More comments