Apparent Perspective and Focal Length

© Mike Spinak


Many photographers think focal length affects perspective. They think wide angle lenses exaggerate perspective, and telescopic lenses flatten perspective.

.

Technically speaking, that’s not true. All focal lengths show perspective identically.

Some of you might be thinking, “But I’ve seen exaggerated perspective with wide angle lenses and flattened perspective with telescopic lenses”. So, what’s going on?

Here’s the way it’s actually working:

In order for a wide angle lens and a telephoto lens to exactly fit a subject of the same size (such as a yardstick fitting exactly across the length of the frame), the wide angle lens needs to be closer to the subject than the telephoto lens. If we wanted to exactly fill the frame with a yardstick, using a 24 mm lens on a 35 mm format camera, we’d need to be about 2 feet away. To similarly fill the frame with a yardstick, using a 600 mm lens, we’d need to be a little over 49 feet away.

Now imagine there’s a second yardstick, a blue one, two feet behind the first one, which we’ll say is yellow. As you photograph the yellow yardstick with the 24 millimeter lens from 2 feet away, the blue yardstick will be 4 feet away – twice the distance from your lens – which means the blue yardstick will look half as large as the yellow one in your picture. By comparison, when photographed with the 600 millimeter lens, the blue yardstick is 51 feet away – about 4% farther away than the yellow yardstick 49 feet away, which means it will look 2% smaller in your picture.

Thus, while all lenses actually show perspective the same way, the way we use lenses, due to their focal lengths, gives us the appearance of different perspectives. The way wide angle lenses usually get used (comparatively closer to subjects) appears to “exaggerate” perspective; the way telephoto lenses usually get used (comparatively farther from subjects) appears to “compress” perspective.

If you used both the 24 mm lens and the 600 mm lens from 49 feet away, and then cropped the 24 mm picture down to where it covered the same area as the 600 mm picture, then compared the two shots side by side, you’d see that the perspective in the two pictures is identical.


View East of Tioga Pass, Near Yosemite National Park, California


All pictures and text are © Mike Spinak, unless otherwise noted. All pictures shown are available for purchase as fine art prints, and are available for licensed stock use. Telephone: (831) 325-6917.

  • Tim Parkin - Of course this is true but if took a shot of the stick filling the frame with a 600mm lens and then took a shot of the stick with the 24mm lens but placed the stick in the corner of the picture and then cropped to fill the frame, the stick would be massively distorted in comparison with the 600mm shot.

    *and* you wouldn’t be able to reproduce that distortion with the 600mm lens no matter where you pointed it.

    This difference in distortion can be used to change the balance of a picture if you are willing to lose a few pixels (or have a tilt shift lens).

    For instance, if you want to take a picture of a large landscape view with someone standing in the bottom left hand corner, you would usually end up with their head stretched horribly. If you use a wider lens and shoot directly at the person then you will have an undistorted person with the rest of the view stretched. This will probably end up a lot ‘natural’ looking as we know people aren’t meant to look squished and we will forgive stretched landscapes.

    The same effect can be acheived with a tilt shift lens by pointing at the person and then using rise and shift to recompose your picture. This will also help because the centre of your lens, the bit with the highest acuity, will be used on the person.

    This can also be used to rebalance a picture where you may have too much mountain at the top of the picture. If you shoot directly at the mountain then your foreground will be stretched even more and your mountain stretched less.

    In large format terms this effect is referred to as ‘looming’ (and can be easily acheived by just using rear tilt to focus on the foreground- a movement which is very similar to drop front and front tilt).

    Which begs the question – if we can do this with a lens, i.e. distort the picture in an almost arbitrary way, why shouldn’t we do similar things in photoshop? i.e. rebalancing a picture using the transform tool is equivalent to the visual distortions we apply by using wide angle lenses or by using perspective adjustments with tilt shift lenses or by cropping. This could give more creative freedom in ‘interpreting’ the landscape artistically.ReplyCancel

  • naturography - Hi, Tim,

    Thank you for your thoughtful reply.
    I agree with everything you say about this optical characteristic, and how it can be used. However, I think the word “massively” may be a bit of an overstatement. Look at the barrel distortion measurements on this site, to see what I mean. http://www.photozone.de/reviews
    For example, the Canon 24 mm f/1.4 lens shows 0.74% barrel distortion. http://www.photozone.de/canon-eos/152-canon-ef-24mm-f14-usm-l-test-report–review?start=1
    The Canon 16-35 f/2.8 shows 0.81% barrel distortion at 24 mm. http://www.photozone.de/canon-eos/175-canon-ef-16-35mm-f28-usm-l-lab-test-report–review?start=1
    The Canon 24-70 f/2.8 lens shows 0.75% barrel distortion at 24 mm. http://www.photozone.de/canon-eos/184-canon-ef-24-70mm-f28-usm-l-test-report–review?start=1
    Nikon’s 24mm f/1.4 mm lens shows 0.77% barrel distortion. http://www.photozone.de/nikon–nikkor-aps-c-lens-tests/557-nikkorafs2414dx?start=1
    And so on.
    Of course, there are some lenses with greater distortion than these; but these examples I gave are popular lenses, displaying common amounts of barrel distortion, which are fairly minimal – well under one percent. In my case, I’d need to specifically buy a lens with far more extreme distortion (unless I’m using the 15 mm fisheye lens, which really does distort massively), in order to do as you suggest in a markedly noticeable way.
    So, using the lenses I commonly use, “massive distortion” and “stretched horribly” are, in my opinion, overstated.
    As to your question about using the transform tool in Photoshop to distort the picture in an arbitrary way:
    In my opinion, there is no intrinsic reason not to. It’s a matter of your objectives, and your taste. When we are talking strictly about making art, then I think we should take full license to interpret the landscape any way we want.
    It’s not especially to my taste, so I don’t tend to do that sort of stuff, much. But I welcome anyone to do so, unfettered.
    I should add that, in some circumstances, I find such manipulations fake looking and distracting, in a way that doesn’t work well for me. For example, you often see this in architectural photos, where movements have been used to counter the “keystone effect”. While this may work insofar as making a tall building’s edges closer to parallel, it has a visually confused look when you can see the bottoms of window sills on the upper stories of the buildings, and yet you see the window’s shape as though you are looking at them straight-on. This works poorly, to my eye. The same also sometimes applies to landscape photography – though the nature of the subject matter makes that issue painfully visible less frequently than neatly geometric buildings.ReplyCancel

  • Tim Parkin - ah sorry to not be clear – I wasn’t talking about barrel distortions, I was talking about the natural distortion caused by wide angle lenses. It’s not really distortion in the true sense of the word. In any recti-linear lens, the edges of the lens will show a distortion because each additional degree of coverage gets spread over a large and larger area.

    There is a great shot in Stroebels book looking down on a table covered in tennis balls and all the balls in the centre looks natural but the ones on the edges are elongated quite dramatically. I think it’s called “volume anamorphosis”.

    I’ll try to dig out the picture if possibleReplyCancel

  • Maria - Is it really THIS simple?

    So it’s just a matter of comparative perspective, so to speak. Like why the moon looks larger on the horizon — when it can be compared with familiar objects such as hills and trees.

    Or why a year seems a lot longer when you’re 10 years old (1/10 of your life) than when you’re 50 years old (1/50 of your life).

    Thanks.ReplyCancel

  • Interesting Links, September 27, 2011 | An Eclectic Mind - [...] Apparent Perspective and Focal Length – Interesting explanation of why focal length affects apparent perspective. [...]ReplyCancel