Is the DOF difference between M4/3 and full-frame really 2 stops?

Status
Not open for further replies.

Klorenzo

Mu-43 All-Pro
Joined
Mar 10, 2014
Messages
1,905
Real Name
Lorenzo
Thanks for the POTN link, I found it very useful, in particular the last example on the mountains: "It's because the additional twenty feet that you put between yourself and the people is insignificant relative to the fifteen miles between your viewing spot and the mountains."

And also thanks for taking the time to make and share your test results.
 

EarthQuake

Mu-43 Top Veteran
Joined
Sep 30, 2013
Messages
934
Thanks for the POTN link, I found it very useful, in particular the last example on the mountains: "It's because the additional twenty feet that you put between yourself and the people is insignificant relative to the fifteen miles between your viewing spot and the mountains."

And also thanks for taking the time to make and share your test results.
No problem, glad to help. When I first started shooting with crop DSLRs I had read on the internet that cropped formats could never have the same perspective as a FF DSLR, due to some inherent qualities of the focal length of the lens. This made sense to me at the time and I even went on repeating this information to other people. At some point a more experienced photographer pointed out that this was false and easy to verify for yourself, which I did, and I've been trying to set the record straight whenever I see the same misconceptions ever since.

Its a matter of the physics of light, and it doesn't only apply to cameras and lenses, but to the human eye as well. The same experiment can be conduced without a camera, simply find an interesting foreground object with a distanct. background. As you move closer or farther away from the foreground object it will look bigger or smaller in relation to the background. This is really the basic concept of perspective, and is solely dependent on distance to subject.

In practical use focal length comes in to play, because for a specific framing, lets say a head shot again, you want to maintain a certain distance to your subject to avoid being too close and distorting perspective. Generally, 85mm on FF, or 85mm equivalent on other formats is about the minimum distance for flattering headshots, which is why this focal length, or rather, this angle of view is considered good for portrait work. Now, angle of view varies depending on focal length and sensor size, so you can never really use the actual focal length alone to detriment this, focal length will only have a relative relationship to perspective, never direct control over the phenomena. Angle of view is actually a much more accurate way of determining perspective for a given crop/framing (head shot, head and shoulders, full body, etc) as it is an absolute measurement that does not vary between formats.

People often seem to assume that it works the other way around, that you pick out a focal length, which in term gives you your perspective/distance to subject, I think this is what generally leads to the misconception that focal length controls perspective, because if you think about it that way, or have only ever shot with one format it sort of makes sense in a way, but when you really research the issue it becomes apparent that it is not the case.

In real use I often choose a specific focal length to get a desired perspective, but not because the focal length itself controls perspective, but because I understand how focal length relates to angle of view for a given distance to subject. For example, If I wish to show a very distorted perspective, I will get very close to the subject and use a lens with a wide angle of view to fit it in the frame. On the other hand, to compress perspective, I will move further away and use a lens with a narrower AOV to get a tight crop. Distance to subject determines perspective, angle of view determines what is in the frame, and focal length is a relative measurement that is meaningless without knowing the sensor size.
 

bikerhiker

Mu-43 Hall of Famer
Joined
Dec 24, 2013
Messages
2,005
Location
Canada
Real Name
David
A focal length does not have a "look". The physical focal length, combined with the size of the sensor is what determines angle of view, which in turn determines working distance for a particular subject (ie: a head shot, or a full body shot), and thus perspective (perspective is solely a product of distance to subject and has nothing to do with focal length). A 50mm lens cropped to a head shot on FF doesn't look anything like a 50mm lens for a head shot on M43, as the working distance and perspective are completely different. I don't see any point comparing a specific focal length over different formats, as the photos will be drastically different (different framing if taken from same distance, different perspective if camera is moved to equalize framing).

Ignoring DOF, to get the same framing, same angle of view, to take the essentially same photo, you need a focal length of half on M43 (-/+ a bit for the different aspect ratios).

Now, DOF is the only real area where different focal length lenses (or rather, actual focal length on FF side and equivalent FL on M43 side) vary in any way, expect of course for the typical differences in rendering that every lens has. Of course I agree with your second paragraph, as I have verified the behavior myself on both formats. The examples posted in this thread are faulty due to the differences in distance to subject/framing altering the DOF/background blur, which I assume are due to quickly taking handheld shots instead of using a tripod to maintain the framing.
The focal length of a lens is the distance from the optical center of a lens to the point where light rays from a distant object converge. In order to calculate a given lens' focal length, the lens is focused at an infinity distance and the distance from the lens' optical center to the focal plane is measured, hence the focal length. But this measurement applies only to a single element design. All these tests that people showed here and in DPREVIEW are NOT taken at infinity.
Go to this link.

https://support.nikonusa.com/app/answers/detail/a_id/16644/~/lens-focus-distance-and-angle-of-view

As you can see here.

3 Nikkor zoom FX lenses were tested with the same camera, same distance, same aperture and same settings. They are tested at infinity and then at close focus.

At infinity with all the lenses set to 200mm, they all LOOK at the same with the same angle of view. Perhaps with some slight variations of blur on the front if you have a Retina display to nitpick! When the camera is moved closer to 5 feet from the subject matter, then observe the different angle of view!
If you want the same angle of view of the 70-300VR @ 5 feet, then you HAVE to use the 70-300VR. You can not use the 70-200VR II because it gives a slightly wider view. In order to match the angle of view of the 70-200VR II to the 70-300VR @ 200mm, you need to move closer or crop the image to match. But by moving closer, you changed the shooting distance and hence the DOF coverage the resulting blur; the look of the image. Also cropping the image to match the same angle of view meant a slight loss of resolution.

As you see here. Despite the same FX format, and the same FX focal length, the field of view or angle of view is NOT always the same when shot at closer distances. Why then we worry about a Panasonic Leica 25mm not being the same as a Zeiss 50mm?!? Just shoot and adjust when necessary.
 

bikerhiker

Mu-43 Hall of Famer
Joined
Dec 24, 2013
Messages
2,005
Location
Canada
Real Name
David
http://photography-on-the.net/forum/showthread.php?t=672913 - Very good explanation of how distance to subject determines perspective, 100mm, 200mm, these numbers have no relevance without a sensor to attach them to. On 645, a 50mm is a wide angle, on FF its a normal, on M43 its a portrait lens. Do these lenses transform when attached to different cameras? No, because there is nothing magical about a focal length.

Its actually very easy to get the same look of a 200mm lens on FF with a 100mm lens on M43, I've done this myself, I suggest anyone else actually try it out, with a fixed subject, on a tripod, if you still have doubts.

EM5, 100mm lens
Subscribe to see EXIF info for this image (if available)


A900, 200mm lens
Subscribe to see EXIF info for this image (if available)


EM5 25mm lens
Subscribe to see EXIF info for this image (if available)


A900 50mm lens
Subscribe to see EXIF info for this image (if available)


Above shots: Apertures have been equalized to provide similar DOF (FF lenses are stopped down two stops extra). Minor differences due to different aspect ratio can be seen as well.

Its also easy to make a 50mm lens "look" like a 200mm lens, at least in terms of the perspective compression that is usually associated with a 200mm tele, only caveat is that by cropping significantly you lose quite a bit of resolution. Just check the POTN link above for that comparison. Again, this is because perspective is determined by distance to subject, and has nothing at all to do with focal length. If this was not true, the very short focal length lenses in point and shoots and cell phones would be incapable of taking anything but photos with fish-eye-esq perspective distortion.
A focal length of the lens is calculated and determined when the subject is shot at "INFINITY" taken into account all corrective lens distortions and aberrations are corrected. All these tests are invalid because they are not shot at infinity. They are shot closer; but we all know that angle of view is NOT always the same with the same focal length of different lenses (Nikon USA had proven this and so have all makers like Canon, Olympus, Sony, Zeiss). So the look you provided is just that; "Your" personal interpretation of what the correct angle of view it should be and you perform the crop necessary to reflect that.
 

Amin Sabet

Administrator
Joined
Apr 10, 2009
Messages
10,899
Location
Boston, MA (USA)
http://photography-on-the.net/forum/showthread.php?t=672913 - Very good explanation of how distance to subject determines perspective, 100mm, 200mm, these numbers have no relevance without a sensor to attach them to. On 645, a 50mm is a wide angle, on FF its a normal, on M43 its a portrait lens. Do these lenses transform when attached to different cameras? No, because there is nothing magical about a focal length.

Its actually very easy to get the same look of a 200mm lens on FF with a 100mm lens on M43, I've done this myself, I suggest anyone else actually try it out, with a fixed subject, on a tripod, if you still have doubts.

EM5, 100mm lens
Subscribe to see EXIF info for this image (if available)


A900, 200mm lens
Subscribe to see EXIF info for this image (if available)


EM5 25mm lens
Subscribe to see EXIF info for this image (if available)


A900 50mm lens
Subscribe to see EXIF info for this image (if available)


Above shots: Apertures have been equalized to provide similar DOF (FF lenses are stopped down two stops extra). Minor differences due to different aspect ratio can be seen as well.

Its also easy to make a 50mm lens "look" like a 200mm lens, at least in terms of the perspective compression that is usually associated with a 200mm tele, only caveat is that by cropping significantly you lose quite a bit of resolution. Just check the POTN link above for that comparison. Again, this is because perspective is determined by distance to subject, and has nothing at all to do with focal length. If this was not true, the very short focal length lenses in point and shoots and cell phones would be incapable of taking anything but photos with fish-eye-esq perspective distortion.

Great explanation, great examples.
 

EarthQuake

Mu-43 Top Veteran
Joined
Sep 30, 2013
Messages
934
The focal length of a lens is the distance from the optical center of a lens to the point where light rays from a distant object converge. In order to calculate a given lens' focal length, the lens is focused at an infinity distance and the distance from the lens' optical center to the focal plane is measured, hence the focal length. But this measurement applies only to a single element design. All these tests that people showed here and in DPREVIEW are NOT taken at infinity.
Go to this link.

https://support.nikonusa.com/app/answers/detail/a_id/16644/~/lens-focus-distance-and-angle-of-view

As you can see here.

3 Nikkor zoom FX lenses were tested with the same camera, same distance, same aperture and same settings. They are tested at infinity and then at close focus.

At infinity with all the lenses set to 200mm, they all LOOK at the same with the same angle of view. Perhaps with some slight variations of blur on the front if you have a Retina display to nitpick! When the camera is moved closer to 5 feet from the subject matter, then observe the different angle of view!
If you want the same angle of view of the 70-300VR @ 5 feet, then you HAVE to use the 70-300VR. You can not use the 70-200VR II because it gives a slightly wider view. In order to match the angle of view of the 70-200VR II to the 70-300VR @ 200mm, you need to move closer or crop the image to match. But by moving closer, you changed the shooting distance and hence the DOF coverage the resulting blur; the look of the image. Also cropping the image to match the same angle of view meant a slight loss of resolution.

As you see here. Despite the same FX format, and the same FX focal length, the field of view or angle of view is NOT always the same when shot at closer distances. Why then we worry about a Panasonic Leica 25mm not being the same as a Zeiss 50mm?!? Just shoot and adjust when necessary.
Right, AOV/actual focal length will vary a bit depending on focus distance and specific lens design, however, these variances also tend to remain fairly consistent when comparing similar types of lenses (there is still some variance but to a small degree, small enough to be negligible in my opinion), and I don't think it's relevant to my main point. My main point is that perspective, and thus the "look" a particular lens/focal length/what have you creates, is a product of distance to subject, not focal length, special glass elements, the rendering of the lens, or anything else. People often suggest that you can't get the same perspective compression on an 42.5mmm M43 lens that you can with an 85mm FF lens, which is completely false. My secondary point would be that comparing lenses that have completely different AOVs is pointless (I'm talking 1.5x, 2x here, not 5% off), as you're comparing completely different images no matter how you compose them.

In no way am I trying to give a tutorial on how to get exactly the same image across different formats and lenses, as I don't see the point in that either. I do think it is valuable to understand how to get very similar results though, and why the differences exist. I think more knowledge is always a good thing. Of course in the field, you just have to make it work.

If you would like to repeat the tests that I've done, but at a variety of focus distances, by all means go right ahead. Unfortunately I no longer have the Sigma 70-200/2.8 lens used in the A900/100mm shot, nor the Sig 50, and the A900 and 85 will be gone soon too.
 
Status
Not open for further replies.
Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Mu-43 is a fan site and not associated with Olympus, Panasonic, or other manufacturers mentioned on this site.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Copyright © 2009-2019 Amin Forums, LLC
Top Bottom