Depth of Field, Depth of Focus, and all Circles of Confusion
I've seen this topic discussed in different venues and I have yet to find a source that can clearly show how these things are related without being inconsistent. I see a major issue with how books, articles, and professional photographers refer to these terms almost interchangeably. I think this is part of the reason this is such a confusing (no pun intended) topic.
Now, I will be the first to admit that, being an amateur photgrapher, I lack the experience and intuitive feel for photography that many of the experts discussing this topic have. However, I do have a strong technical background in the area and I think I may have broken the code on this issue.
I'd like to share with you what I understand of this problem in the hope that many of you will provide additional insight and maybe we can come up with terminology that can consistently and clearly articulate this concept.
The best article I've seen on this subject is by Mr. Michael Reichmann (see http://www.luminous-landscape.com/tutorials/understanding-series/dof.shtml). But, even in this article I believe Mr. Reichmann makes the error that causes all the problems as I see it. Depth-of-field is NOT depth-of-focus (I believe Mr. Reichmann knows this and has actually demonstrated it in his examples but the terminology is still too vague). I've seen others state this as well but they don't take it far enough. Depth-of-focus is NOT the focal length at which light comes to focus. Depth-of-focus is the range that the image plane can move in and out of best focus and still be *in* focus. Again, this is not depth-of-field and I will explain why I think it is not. What qualifies as "in-focus"? This is the "circle-of-confusion" which is how blurry a spot can be before we say "aha! it isn't a spot anymore!".
When we talk about depth-of-focus, we are talking about an effect that occurs around the image plane. There are only three factors that affect depth-of-focus (assuming a perfect lens without any aberrations so that the only thing we're dealing with is blur from defocus): 1) aperture stop; 2) focal length; and 3) the wavelength of light. Depth-of-focus increase with a large f-number (small aperture, long focal length) and long wavelength. Indeed, it is proportional to the product of the f-number and the wavelength.
So why do long-lenses give poor "depth-of-field"? If I have a wider range of distances that I can be in focus then I should, by definition, have a wider range of depth....right? Well, the problem is with the word "field". All of the sudden, the angle-of-view of the lens becomes an important parameter. The wider the lens, the wider the perspective and the more of the background gets collected by the lens system and imaged. What happens now is that the various elements making up the background get squeezed into the film format or digital sensor. In other words, subjects that are actually larger than something closer to the camer will actually look smaller. Since they are smaller, they *appear* to be in focus. This is the same effect you get from looking at a thumbnail of an out-of-focus image that looks OK until you see it on the computer screen or on a larger print and you go "ugh!". As a side note, it is interesting to me to note that a wide-angle lens is by necessity a short focal length lens. Again, it seems counterintuitive (again please bear with my techie vs. photographic background) that a short focal length lens has more depth-of-field since it actually has less depth-of-focus.
The reason wider lenses give better depth-of-field is because of the perspective effect I just mentioned. Objects in the foreground and background really stand out from each other because of the relative sizes and how the geometry projects unto the image plane. Coversely, a long focal-length (telephoto) lens will only image objects that are relatively close to each other in depth. Therefore the shallow depth-of-field. In some books I've seen this described as a "compression" of the field. It is true, but only in the direction of the camera to the subject. In the lateral dimension the field is compressed by a wide-angle lens!
Now I see this topic further complicated by the difference in formats between 35 mm film and digital SLR's. We tend to hear about "effective" focal length due to the format factors. It is true that you will get a different field-of-view from a small digital sensor format than a 35 mm film format. [EDIT: This changes the crop for sure and may "force" a different perspective depending on the focal length of the lens]. But it does NOT affect depth-of-focus nor the actual focal length of the lens. Some have said that because of this it doesn't affect depth-of-field. That's wrong. Since depth-of-field depends on perspective, as I have explained above, then digital cameras with a format factor smaller than 35 mm will also have [EDIT: more] depth-of-field.
I'm sorry if I'm rambling on here or if I'm beating a dead horse. I'm really thinking this stuff out as I write. I hope several of you can shed more insight into this or maybe tell me I've got it all wrong and I should stick to book-worm material instead.... Otherwise, I'd like to know how this topic got so convoluted in the photographic literature.
Thanks,
Erich :scratch
Now, I will be the first to admit that, being an amateur photgrapher, I lack the experience and intuitive feel for photography that many of the experts discussing this topic have. However, I do have a strong technical background in the area and I think I may have broken the code on this issue.
I'd like to share with you what I understand of this problem in the hope that many of you will provide additional insight and maybe we can come up with terminology that can consistently and clearly articulate this concept.
The best article I've seen on this subject is by Mr. Michael Reichmann (see http://www.luminous-landscape.com/tutorials/understanding-series/dof.shtml). But, even in this article I believe Mr. Reichmann makes the error that causes all the problems as I see it. Depth-of-field is NOT depth-of-focus (I believe Mr. Reichmann knows this and has actually demonstrated it in his examples but the terminology is still too vague). I've seen others state this as well but they don't take it far enough. Depth-of-focus is NOT the focal length at which light comes to focus. Depth-of-focus is the range that the image plane can move in and out of best focus and still be *in* focus. Again, this is not depth-of-field and I will explain why I think it is not. What qualifies as "in-focus"? This is the "circle-of-confusion" which is how blurry a spot can be before we say "aha! it isn't a spot anymore!".
When we talk about depth-of-focus, we are talking about an effect that occurs around the image plane. There are only three factors that affect depth-of-focus (assuming a perfect lens without any aberrations so that the only thing we're dealing with is blur from defocus): 1) aperture stop; 2) focal length; and 3) the wavelength of light. Depth-of-focus increase with a large f-number (small aperture, long focal length) and long wavelength. Indeed, it is proportional to the product of the f-number and the wavelength.
So why do long-lenses give poor "depth-of-field"? If I have a wider range of distances that I can be in focus then I should, by definition, have a wider range of depth....right? Well, the problem is with the word "field". All of the sudden, the angle-of-view of the lens becomes an important parameter. The wider the lens, the wider the perspective and the more of the background gets collected by the lens system and imaged. What happens now is that the various elements making up the background get squeezed into the film format or digital sensor. In other words, subjects that are actually larger than something closer to the camer will actually look smaller. Since they are smaller, they *appear* to be in focus. This is the same effect you get from looking at a thumbnail of an out-of-focus image that looks OK until you see it on the computer screen or on a larger print and you go "ugh!". As a side note, it is interesting to me to note that a wide-angle lens is by necessity a short focal length lens. Again, it seems counterintuitive (again please bear with my techie vs. photographic background) that a short focal length lens has more depth-of-field since it actually has less depth-of-focus.
The reason wider lenses give better depth-of-field is because of the perspective effect I just mentioned. Objects in the foreground and background really stand out from each other because of the relative sizes and how the geometry projects unto the image plane. Coversely, a long focal-length (telephoto) lens will only image objects that are relatively close to each other in depth. Therefore the shallow depth-of-field. In some books I've seen this described as a "compression" of the field. It is true, but only in the direction of the camera to the subject. In the lateral dimension the field is compressed by a wide-angle lens!
Now I see this topic further complicated by the difference in formats between 35 mm film and digital SLR's. We tend to hear about "effective" focal length due to the format factors. It is true that you will get a different field-of-view from a small digital sensor format than a 35 mm film format. [EDIT: This changes the crop for sure and may "force" a different perspective depending on the focal length of the lens]. But it does NOT affect depth-of-focus nor the actual focal length of the lens. Some have said that because of this it doesn't affect depth-of-field. That's wrong. Since depth-of-field depends on perspective, as I have explained above, then digital cameras with a format factor smaller than 35 mm will also have [EDIT: more] depth-of-field.
I'm sorry if I'm rambling on here or if I'm beating a dead horse. I'm really thinking this stuff out as I write. I hope several of you can shed more insight into this or maybe tell me I've got it all wrong and I should stick to book-worm material instead.... Otherwise, I'd like to know how this topic got so convoluted in the photographic literature.
Thanks,
Erich :scratch
0
Comments
Erich
Catapultam habeo. Nisi pecuniam omnem mihi dabis, ad caput tuum saxum immane mittam
http://www.mcneel.com/users/jb/foghorn/ill_shut_up.au
by Barbara London and John Upton that could explain things better than I ever could..You can get it on Amazon.
Depth of Field - The area beyond and in front of the object/subject focused on that is in [font="]acceptable focus.
Depth of Focus - Inside the camera. The amount that the recording medium can be off plain and still yeild acceptable focus.
[/font]Circle-of-Confusion - Your's is as good an answer as any. "which is how blurry a spot can be before we say "aha! it isn't a spot anymore."
I beleive this is correct.:D
Mitch
I'm guessing you are still going after this depth of field vs. focus argument. I think circle of confusion is pretty well defined and used correctly, but you are totally right in that these two are interchanged very often.
Lets get down and dirty then, we need some variables and equations. I'm not an optics expert, but I am an engineer. And I'm going to get my optics book right now and see what I can find...
this'll be fun
moderator of: The Flea Market [ guidelines ]
What About Digital?
moderator of: The Flea Market [ guidelines ]
Catapultam habeo. Nisi pecuniam omnem mihi dabis, ad caput tuum saxum immane mittam
http://www.mcneel.com/users/jb/foghorn/ill_shut_up.au
Lattely I have been reading about these subjects and here I send you an excellent link on this matter.
http://www.dofmaster.com/
I am "building" a Depth of Field Calculator for my two lenses. IN PAPER !!!...
I will show it to you all, as soon as it is finished. I think I'll remember to do so.
Meanwhile here is a site which gives differents formulas from the one above:
http://www.nikonians.org/html/resources/guides/dof/hyperfocal2.html
I think here they are wrong because I tried the formulas in Excel. With dofmaster Things worked fine.
I studied the topic of circles of confusion and related info but have found very little practical application for this knowledge and see it as a needless complication in the photographic process. The fact that you can go your whole life without having to know this info and yet succeed in photography tells me all I need to know ;-)
Some of the calculator types might need this for camera design and such, but the average Joe doesn't. It can all be boiled down to a few rules of thumb that are easy to remember and implement.
Shay (putting my slide rule away now) Stephens
"Failure is feedback. And feedback is the breakfast of champions." - fortune cookie
Just LOOK through the viewfinder. You can also evaluate DOF after you press the DOF preview button to stop your lens down if needed. But I bet the pros rarely use this feature, just like Shay said.
Moderator of the Technique Forum and Finishing School on Dgrin
A former sports shooter
Follow me at: https://www.flickr.com/photos/bjurasz/
My Etsy store: https://www.etsy.com/shop/mercphoto?ref=hdr_shop_menu
http://www.dofmaster.com/dofjs.html
Also has software available from the menu on the left side of that page that will print out DOF charts for your own camera and lenses you use.
Good tutorial on hyperfocal distance calculations, charts, etc.
"You miss 100% of the shots you don't take" - Wayne Gretzky
Shay.
I agree with you but if you study a little of this it is no harm at all... It is knowledge.
Finally, it gives you another way to look at the picture. It gives, shall I say culture ? and that is good.
Let me join a little example.
The photo you will see at http://www.antoniocorreia.com/album/album50/Brasil_Iguasu_2005_01?full=1, was shot in Brasil Iguasu, a place to visit.
The lens - no question - is a supperb L USM 16 35 mm f/2.8 Canon.The body is a 20D what does not matter here.
Shutter Speed 1/320; Normal Program f/9 ISO 100 Focal lenght 16 mm.
It was focused to the water. Because the depth of field was not enought my wife and I are a little bit out of focus.
This can only be seen when you blow up the photo.
We are not so sharp as we shoul be if I had rotated a little bit the focus ring after having switched off the auto-focus.
This picture made me to search for information I knew the existence of.
I'm sure you have seen on some old - say 30 years - zoom lens, the graphics on the cylinder. In blue for the f/5.6 as a can remember.
This was quite usefull (for the connoisseur).
Thank you for your point of view and forgive ye bad english for I have studied it only for 3 years.
Depth of focus: This has nothing to do with depth of field. This has to do with ones ability to see past the surface and focus more deeply to find the burried meaning. If say you drink a glass of fine wine, and all you can think of is "a couple more of these and I can get a good buzz on" you have a very limited depth of focus....Or a very bad first date.
Circle of confusion: Liberals trying to discuss facts.
Sam
Please share these rules with us
Thierry
Exactly
You already know them:
- Wide aperture = small DOF...............Small aperture = wide DOF
- Telephoto = small DOF.....................Wide angle = wide DOF
- Close focus on subject = small DOF....Far away focus = wide DOF
Do you want really wide DOF? Use a small aperture (e.g. f/32), wide angle lens (e.g. 28mm), and focus far away from the lens (e.g. try infinity)Want crazy small DOF instead? Use wide aperture (e.g. f/2.8), telephoto lens (e.g. 200mm), and focus up close to the lens (e.g. closest focus distance).
Want something in between? Modify one or more of the above variables to be somewhere between the extremes. You can add salt and pepper to taste, bake for 20 minutes, and enjoy
"Failure is feedback. And feedback is the breakfast of champions." - fortune cookie
Good point Shay !
People are interesting !
Good point Thiery
But who is right
The latter (nikonians.org, shared for us by J. Ramón Palacios - who I don't know, but he deserves to be acknowledged for his efforts) uses a value that is close to Carl Zeiss's recommendation (1/1730th of the length of the "negative" diagonal ). The former is based on a more "generous" value that is the one used by Nikon for the markings on its lenses - when there used to be markings :cry (funny, you'd think it would be the other way, with the Nikon CoC definition used by nikonians.org
Thierry
Thierry
Thank you.
Shay, I agree with you on the practical application of these concepts. That's why I added the caveat that I was trying to reconcile my techie perspective with a practical one.
Since I posted this I've learned a lot from just shooting. I find myself worrying less and less on the specific settings I'm applying and more and more on naturally using the rules of thumb Shay laid out.
That said, there is a venue for working through these technical details. I can see it in terms of comparing lenses in a quantitative way for the purposes of design evaluation, etc. Useful for the lens makers, maybe not so much for the shooter. On the other hand, having insight into both should give you a, shall we say, "more complete picture"?
One item I didn't mention in my original post was the effect of distance to subject. What happens here is that if you are close to the subject, it doesnt take much distance in depth for the image to be out of focus on the focal plane. On the other hand, if you are far way, the same depth doesn't span as large a range at the image plane and therefore is in focus (within the circle of confusion).
Erich
Anyway, my question is this:
I understand how to get a shallow DOF, by focusing in close to the end of the lens. But what if I want a shallow DOF, but in the middle of my shot? i.e. I want a very narrow band, in the middle of my shot, to be in focus (and everything before and after it, out of focus).
like this (OOF = out of focus):
Andrew,
The problem here is, the technique that you would need to adopt to do this in a point and shoot camera is unneccessary with a dSLR.
With a P&S, say you want a certain shot with a narrow DOF at about 1-2m from a subject. You would need to adjust the zoom and move backwards away from the subject until you achieved the desired framing and narrow enough DOF. This is due to the small sensor size, that even at large apertures everything is in focus.
Hope this helps,
David
SmugMug API Developer
My Photos
I was actually wondering how it would be done with a dslr
set a large aperture (eg f2) then starting with the closest focal point slowly adjust the focus and the field of focus will slowly away from you... like in your diagram.
SmugMug API Developer
My Photos
Hope to be usefull.
http://www.night-ray.com/photores/canon16_dof_10-50mm.png
the previous:
http://www.night-ray.com/photography.php?id=6
The lens focus ring allows you to adjust what object distance gets to be at "best focus". Some of the better lenses will actually show you what distance you are focused at. If you use a large aperture you will limit the range that is in focus. This affects both the objects before and after the location you focused at.
The longer the focal length the shallower this in-focus region will be. So, if you want to have a very specific and limited range in-focus you should be using a telephoto lens. Shorter focal length lenses will give your more DOF which means more will be in-focus and it's harder to limit the range where objects look sharp.
Erich