6

Suppose I’ve taken a photo and then cropped it. If the crop is extreme enough then I might want to go back and take the same photo with a more telephoto lens so that I can take full advantage of my camera’s sensor.

If I’ve taken a photo at a given focal length and then cropped it by a certain amount, how do I calculate the “effective focal length” of the cropped photo?

12

It's pretty much strictly linear unless you're talking about very close focusing distances or macro distances. For everything else, what little you may be off is probably less than the rounding error between the actual focal length and the marketed focal length of the lenses in question. For example, a lens with a focal length of 192mm will probably be sold as a 200mm lens. So will a lens with a 197mm or 203mm focal length.

Assuming you are shooting digital, all you have to do is find the ratio of your total sensor width (or height) and divide it by the pixel width (or height) you have left after cropping, then multiply the result by the focal length of the lens with which you shot the photo. If you crop to a different aspect ratio, use whichever side of the image, width or height, you reduced by the lower ratio.

Suppose you used a 200mm lens on a camera with a 6000x4000 pixel sensor. You then cropped the photo to only 3000x2000 pixels. 6000 divided by 3000 is 2.0. Multiply 2.0 times 200mm and you would have needed a 400mm lens to fill the frame with the same field of view you got after cropping the original image.

Suppose you used an 85mm lens with your 24MP camera with the 6000x4000 pixel sensor. You then cropped the image to 1250x1000 pixels (going from a 3:2 to 5:4 aspect ratio). You reduced the long side by a factor of 4.8. You reduced the short side by a factor of 4.0. 4.0 times 85mm is 340mm, so it would have taken a 340mm lens to fill the short side of the frame with what you had left after you cropped. The long side of your photo with the 340mm lens would contain a slightly wider field of view than your crop of the original image, but when you crop the second image from 6000x4000 pixels to 5000x4000 pixels to get the same aspect ratio as your 1250x1000 pixel crop of the original image you'd have the same field of view in both directions.

In the case of Macro lenses focal lengths aren't directly applicable, since the Minimum Focusing Distance and Reproduction ratio are what matters. Focal length is generally expressed in terms of when the lens is focused at infinity.

Personally, I would divide diagonal by diagonal and then aspect doesn't matter.

Diagonals don't work that way when you are comparing two different crops of images from the same sensor.

The sensor itself doesn't change aspect ratio. If the lens is made for that camera, the width of the image circle will be enough to cover the sensor's diagonal at the camera's original aspect ratio, not the reduced diagonal of the cropped aspect ratio.

Say you crop a 6000x4000 pixel image to 5000x4000 pixels to make an 8x10. You haven't changed the magnification ratio at all, but the diagonal will be shorter. You still need the same focal length lens made for that camera that you started with to produce that cropped image. You can't use a slightly longer focal length and get the same picture, even though the diagonal of the 5000x4000 pixel image is shorter than the diagonal of the 6000x4000 pixel image. If you used a longer focal length on the same camera you'd cut off some of the field of view you left in the original crop along with some of what you cropped off on the long ends.

The same concept scales when you change the magnification ratio as well as the aspect ratio. When using the same sensor, as this question presupposes, you must always base it on the linear measurement of the side you reduced by the lowest ratio.

  • Great answer, thank you! One more thing to point out is that if you start with a landscape photo and crop it to portrait orientation, or vice versa, you should still compare the long side of the cropped photo to the long side of the original, and likewise for the short sides. (Unless, I suppose, there’s some constraint forcing you to only shoot in landscape or only in portrait.) – bdesham Aug 1 '17 at 17:20
  • 1
    Personally, I would divide diagonal by diagonal and then aspect doesn't matter. – Octopus Aug 2 '17 at 0:54
  • An you would cut off part of what was included in the crop in the process... – Michael C Aug 2 '17 at 7:41
  • When you're using the same sensor the diagonal of the sensor is always the same. If you change the aspect ratio the side you reduced by the least ratio is the one that matters. – Michael C Aug 2 '17 at 7:55
7

Short answer: if you halve the width, you have (effectively) doubled the focal length.

  • and if you reduced the with to say 75%, then the effective focal length is what it was / .75 – WayneF Aug 1 '17 at 14:17
2

Display the initial image on your monitor. Using a millimeter ruler, measure the height or width of an object as depicted. Next display the cropped image. Now take a measurement of the same but now enlarged object. Divide the measurement of the larger object by the smaller. As an example: 12mm vs. 18mm = 1.5. This is the magnification factor (delta). Now multiply this delta by the focal length used to take the initial image. Suppose 55mm was uses X 1.5 = 82.5 ---- This is revised focal length needed to duplicate the cropped image.

0

I am by no means an expert, and I'm afraid some might see this response as pedantry, but for completeness I wanted to say that cropping the field of view is not the only effect of a longer focal length.

If I were to invert your question and ask "what focal length would I have to use to get the same image, but cropped by a certain amount?" the answer is, in general, that there is no such focal length. You could restrict your field of view by the same amount, but the content of your cropped image would look different.

In the same way, the answer to your question is "there is no effective focal length that will give the same result as cropping a photo".

A longer focal length foreshortens the perspective in your image such that the effect of distance is rendered differently. If you have a "special case" image where there is little or no relative distance involved, then please ignore my comment.

I understand this might not be in the spirit of what you were asking, but it might prove useful to somebody.

Update - please see my comment below. What I stated above is not correct, although it is quite a common belief so I don't feel too bad :-)

  • 3
    This isn't right; the perspective will be the same if you stand in the same place, no matter what focal length you use or how you crop. Zooming or cropping (assuming sufficient resolution and identical viewing size) should will produce identical "foreshortening" or "flattening". – mattdm Aug 1 '17 at 12:57
  • "A longer focal length foreshortens the perspective in your image such that the effect of distance is rendered differently." This is not true. As long as you are shooting the same scene, cropping is exactly the same as increasing the focal length (ignoring the fact that actual lenses often don't have the exact focal length that is printed on them). This is exactly how the effect of crop sensor cameras is calculated. The sensors are 1.6x (Canon) / 1.5x (Nikon) smaller than 35mm full frame sensors. Lenses on crop cameras have a FoV-equivalent focal length of focal length times the crop factor – scottbb Aug 1 '17 at 13:01
  • 1
    I just found this helpful article which supports both mattdm and scottbb. patricktaylor.com/1988. It seems that your position is an important factor. There can be the impression of foreshortening with a longer focal length, but this is an optical illusion. Thanks for your patience! – Martin Aug 1 '17 at 13:25
  • 2
    Foreshortening is not an illusion — it's a real effect, it only happens if the distance between the camera and subject changes along with focal length. If you keeping the same framing of a human subject (i.e, head-and-shoulders shot), and move the camera back to keep the framing as you increase focal length, you will see background compression, and flattening of facial features. But by moving the camera, you have changed perspective. – scottbb Aug 1 '17 at 13:29
  • 1
-1

Short Answer: Use the geometry of "objects that line up" in the original photo.

Your question should be: How do I find my original position? The same focal length (y) is irrelevant if you don't have x and z.

It is technically impossible or at least very difficult without surveying the actual scene, made easier if there is a third dimension elevation.

You need to triangulate back to the original position, by using objects in the distance lined up in the original photo. The wider the triangulate lines are, the more accurate the position.

You will then be at the accurate focal length and x & z. You now just need to focus the same as the original picture and adjust your zoom to the crop size you now want.

Of course all of this in not possible if the scene does not have fixed objects in a line out into the distance.

Enjoy!

  • I think you are confusing focus distance (the distance from the camera to the plane of focus) with focal length (a property of a lens that determines the lens's field of view). OP is not trying to perform any measurements of objects in the scene, or the camera's position/location. – scottbb Aug 1 '17 at 13:05
  • You are right I maybe confused with this. But his question suggests he does want and need the distance to the object and his left right offset and his up down offset, he says he wants to go back and take the photo again but more zoomed in, focal distance is most likely more important than focal length, otherwise objects will not be located in the same part of the frame. – NZ Dev Aug 1 '17 at 13:11
  • Using your words: "As long as you are shooting the same scene", this is exactly my point but to get the same scene you must be located at the same location as the original, and hence using you words again "cropping is exactly the same as increasing the focal length". – NZ Dev Aug 1 '17 at 13:18
  • The original question didn't indicate that determining the position of the camera was a concern. For example, going back to a scenic overlook and reshooting the same landmark or landscape, but this time using a longer lens. – scottbb Aug 1 '17 at 13:21
  • But it did, he clearly says he "wants to go back and take the same photo again". – NZ Dev Aug 1 '17 at 13:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.