Can 24mm cropped be equivalent to 28mm? (24 f/2.8 IS vs 28 f/2.8 IS)

I'm trying to choose between the 24 f/2.8 IS and the 28 f/2.8 IS, both of which are attractively priced at your favorite NYC retailer.  If I get the 24 but want a shot equivalent to 28mm can I just crop it or are there some other considerations native to the 24 that would still make it look different than the 28? (I have a 6D and a 7D.)  One argument against the 28 is it is close to my 35L.

The 28 is tempting at $80 less than the 24 but you can't crop wider.  If my simplistic reasoning holds up I'll probably get the 24 unless someone points out some flaw or other consideration.  If it were that simple, then why did Canon issue two such closely related lenses?

In theory, yes - and other than reducing the resolution due to discarding data (the cropped "out" portion) and introducing what is the equivelent of "digital zoom", I can't think of any other issues.

If you make an exposure with the 24mm at f/2.8 and crop to the field of view of 28mm, you'll have something very similar to the image you'd have gotten with the 28mm at around f/3.3. With the 6D you'll also be down to 15MP from 20MP.

misolo wrote:

If you make an exposure with the 24mm at f/2.8 and crop to the field of view of 28mm, you'll have something very similar to the image you'd have gotten with the 28mm at around f/3.3. With the 6D you'll also be down to 15MP from 20MP.

Actually, though counter-intuitively, it would be slightly the other way around: the depth of field on the cropped version would be slightly less, as if it had been taken with a slightly larger aperture (about f/2.4). The size of the blur circles recorded on the sensor itself would, of course, not be different, since both the uncropped and cropped version were made with a 24 mm lens at f/2.8. But because cropping the 24 mm image to the equivalent FOV of 28 mm makes the image smaller, the cropped version needs to be enlarged more in order to produce the standard print size for which depth of field calculations are calibrated. That enlarges the blur circles more, with the net effect of lower perceived depth of field, for any given viewing distance.

This is exactly the same as using the 24 mm lens with a smaller sensor, with a crop factor of 1.167 compared to the larger "full frame" sensor. So just as the circle of confusion of 0.030 usually used for full frame cameras needs to be adjusted to 0.019 for a 1.6 crop camera, it would need to be adjusted to 0.030 / 1.167 =~0.26 for a full frame image taken at 24 mm but cropped to 28 mm equivalent  field of view. Given the same focal length, f-number and focus distance, that decrease in CoC translates into less depth of field.

It is still true that use of a smaller "cropped" sensor is associated with greater depth field, because in order to produce the same framing, either a shorter focal length or greater subject distance would be needed. Both of those offset the decrease in CoC and result in a net increase in depth of field. But this case is different -- it was stipulated that the lens, aperture and subject distance would be the same, and cropping after the fact would be used to achieve the desired framing. In that case, only the effect of CoC is "in play" and Dof is slightly reduced.

Oh the fun one can have when noodling around with DoF calculators!

Dave

dsjtecserv wrote:

misolo wrote:

If you make an exposure with the 24mm at f/2.8 and crop to the field of view of 28mm, you'll have something very similar to the image you'd have gotten with the 28mm at around f/3.3. With the 6D you'll also be down to 15MP from 20MP.

Actually, though counter-intuitively, it would be slightly the other way around: the depth of field on the cropped version would be slightly less, as if it had been taken with a slightly larger aperture (about f/2.4). The size of the blur circles recorded on the sensor itself would, of course, not be different, since both the uncropped and cropped version were made with a 24 mm lens at f/2.8. But because cropping the 24 mm image to the equivalent FOV of 28 mm makes the image smaller, the cropped version needs to be enlarged more in order to produce the standard print size for which depth of field calculations are calibrated. That enlarges the blur circles more, with the net effect of lower perceived depth of field, for any given viewing distance.

This is exactly the same as using the 24 mm lens with a smaller sensor, with a crop factor of 1.167 compared to the larger "full frame" sensor. So just as the circle of confusion of 0.030 usually used for full frame cameras needs to be adjusted to 0.019 for a 1.6 crop camera, it would need to be adjusted to 0.030 / 1.167 =~0.26 for a full frame image taken at 24 mm but cropped to 28 mm equivalent field of view. Given the same focal length, f-number and focus distance, that decrease in CoC translates into less depth of field.

It is still true that use of a smaller "cropped" sensor is associated with greater depth field, because in order to produce the same framing, either a shorter focal length or greater subject distance would be needed. Both of those offset the decrease in CoC and result in a net increase in depth of field. But this case is different -- it was stipulated that the lens, aperture and subject distance would be the same, and cropping after the fact would be used to achieve the desired framing. In that case, only the effect of CoC is "in play" and Dof is slightly reduced.

Oh the fun one can have when noodling around with DoF calculators!

Dave

Sometimes something is counter-intuitive simply because it is completely wrong...

I suggest you simply try it with your camera.

But if you really prefer online depth of field calculators try, say,

http://www.dofmaster.com/dofjs.html

with the settings:

7D, 24mm, f/2.8, distance 2m --> depth of field 0.76m

misolo wrote:

dsjtecserv wrote:

misolo wrote:

If you make an exposure with the 24mm at f/2.8 and crop to the field of view of 28mm, you'll have something very similar to the image you'd have gotten with the 28mm at around f/3.3. With the 6D you'll also be down to 15MP from 20MP.

Actually, though counter-intuitively, it would be slightly the other way around: the depth of field on the cropped version would be slightly less, as if it had been taken with a slightly larger aperture (about f/2.4). The size of the blur circles recorded on the sensor itself would, of course, not be different, since both the uncropped and cropped version were made with a 24 mm lens at f/2.8. But because cropping the 24 mm image to the equivalent FOV of 28 mm makes the image smaller, the cropped version needs to be enlarged more in order to produce the standard print size for which depth of field calculations are calibrated. That enlarges the blur circles more, with the net effect of lower perceived depth of field, for any given viewing distance.

This is exactly the same as using the 24 mm lens with a smaller sensor, with a crop factor of 1.167 compared to the larger "full frame" sensor. So just as the circle of confusion of 0.030 usually used for full frame cameras needs to be adjusted to 0.019 for a 1.6 crop camera, it would need to be adjusted to 0.030 / 1.167 =~0.26 for a full frame image taken at 24 mm but cropped to 28 mm equivalent field of view. Given the same focal length, f-number and focus distance, that decrease in CoC translates into less depth of field.

It is still true that use of a smaller "cropped" sensor is associated with greater depth field, because in order to produce the same framing, either a shorter focal length or greater subject distance would be needed. Both of those offset the decrease in CoC and result in a net increase in depth of field. But this case is different -- it was stipulated that the lens, aperture and subject distance would be the same, and cropping after the fact would be used to achieve the desired framing. In that case, only the effect of CoC is "in play" and Dof is slightly reduced.

Oh the fun one can have when noodling around with DoF calculators!

Dave

Sometimes something is counter-intuitive simply because it is completely wrong...

I suggest you simply try it with your camera.

But if you really prefer online depth of field calculators try, say,

http://www.dofmaster.com/dofjs.html

with the settings:

7D, 24mm, f/2.8, distance 2m --> depth of field 0.76m

I'm sorry, you are correct; my original post was misleading and was intended to make a different point. Indeed a 24 mm f/2.8 shot cropped to 28 mm would have the field of view and depth of field similar to one shot with a 28 mm lens at f/3.3. My point was that, counter-intuitively, this is less depth of field, because the only thing that changed as a result of the cropping was the needed degree of enlargement. I shouldn't have implied that your equivalence was wrong, but rather emphasized the more surprising element of that.

So to put numbers to it:

Full frame camera, 24 mm, f/2.8, focus distance 10 ft, Circle of confusion 0.030: DoF = 11.1 ft

Full frame camera, 28 mm f/3.4, focus distance 10 ft, circle of confusion 0.030: Dof = 9.16 ft

Cropped scenario, where only CoC is adjusted to reflect greater enlargement:

Full frame camera, 24 mm, f/2.8, focus distance 10 ft, Circle of confusion 0.026: DoF = about 9.07 ft

And, as noted earlier, one observation is that when a cropped image is used (either by the sensor of after the fact) without changing focal length, f-number or focus distance, depth of field decreases.

Sorry for the confusion.

Dave

go for 24. you already have 35, and I find 24 much more interesting lens than the 28.  with the exception of slight difference in dof and resolution, you will get the same image if you crop the 24 to 28 fov, as if you were shooting with a 28 from the same spot.  however.  what will happen more often is that you would move closer with the 24 to get same framing as you would have gotten with 28, and there you will very much change your perspective.

Thank-you all for your replies and analyses - the jury is in and I'll go for the 24!

However, any reason Canon issued two lenses that are so similar?  I probably would have jumped on either one by itself but having this choice kept me from buying either so far.  Seems like poor marketing strategy unless there is a bona-fide market segmentation issue between the two lenses that I'm unaware of (and I'm sure I'm not the only one with analysis paralysis between them).

rowlandw wrote:

Thank-you all for your replies and analyses - the jury is in and I'll go for the 24!

However, any reason Canon issued two lenses that are so similar? I probably would have jumped on either one by itself but having this choice kept me from buying either so far. Seems like poor marketing strategy unless there is a bona-fide market segmentation issue between the two lenses that I'm unaware of (and I'm sure I'm not the only one with analysis paralysis between them).

Some people have a strong preference for either. Some photographers shooting people and events really want 28mm because that's about the point where if you go any wider anyone near the edges of the image will look weirdly distorted, and they don't want to be cropping hundreds of images (nor having to get the framing right when shooting without visual guidance). On the other hand, 24mm has become very popular because it's where you start getting images with a more 'dramatic' and noticeable perspective effect, but without the greater challenges of more extreme wide angle lenses (both for composition and image quality).