Depth-of-Field Hell

I might have lost my cool a bit in the the "P.S." to the "What's a 'Fast Lens'?" post below. I apologize for that.

My problem is that I've been sensitized. I was the editor of a technical photography magazine for six years. Depth-of-field (DoF) is a subject that is deeply beloved of wannabe photo-tech writers, and few things could create instantaneous dismay deep in my editor's heart like opening a submission and finding yet another beautifully prepared, extensive, arcane article on DoF. DoF is perhaps the ultimate hell of photo-tech—graspable enough so many people think they can understand it (or think they should!), complex enough so that very few really can. Whenever the subject really gets going, great reams of material are spewed forth "explaining" it, always—always—strewn with considerable content that happens not to be true. Assuming any given reader engages with said great mass of verbal matter, he comes out the other side usually no more enlightened, and possibly considerably more confused. Finally, add to those conditions the simple fact that a technical understanding of DoF is not required for the successful practice of photography, and you begin to appreciate the extent of my exasperation with the subject.

(By the way, the other day I finally figured out why "grammar police" types can seem to be so fussy and cross. It's because they keep correcting the World, and the World keeps making the same mistakes anyway. If you tell your child three times to hang up the wet towel in the bathroom and he doesn't, you're not that irritated. When you've told him 700 times and you still come into the bathroom and find the damp towel in a moldy pile on the counter, you might well be incensed. So do you see my point? No matter how many times I remind people that "loose" is the opposite of "tight" and "lose" is the opposite of "win," people are still going to misspell it—because it's different people all the time, and there are an inexhaustible supply of them. It's like having 40,000 children, each one of whom needs to be trained to hang up the towel. The first three people I correct, I'm perfectly calm. By the 700th, I'm about to "loose" my mind. Anyway, you see the problem: educate one photographer about DoF, or ten, or a hundred—at however great a cost of time and effort—and you still have numberless millions to go. It's Sisyphean.)

In any event, I've just added more material to the end of the original post. Ctein chimed in. It's now longer. And hopefully a bit more clear. But it's beginning to flirt with becoming yet another of those great masses of verbal matter on DoF with which the literature of photography, and now the internet, is littered.

End-around
Anyway, I figured I might try to do a quick, painless end-around on the subject, and try to tell you what actually matters about DoF. I'm going to ignore the technical explanation: because That Way Lie Dragons; Abandon Hope All Ye Who Enter there; etc. I'm doing that on purpose, so please don't write to fill me in.

1. By far—by far—the most important factor in how much DoF you'll get is your distance setting. The farther the distance between the lens and whatever it's focused on, the more you'll get in the DoF.

2. Next most important: the focal length of the lens relative to the size of the film/sensor. Again, on principle, let's forget all the tech-talk about what's "really" going on and the scientific explanations and all that. All you need to remember is that the longer the lens, the less apparent DoF there will be in your shots. And, big jumps matter, little ones not so much; there's not that much difference between an 85mm and a 100mm, or between a 24mm and a 20mm. Very roughly speaking, when there's a 2x or 1/2x difference in focal length, you're going to have noticeably different DoF characteristics to learn.

3. Last: aperture setting. When I was teaching, all my kids had to memorize the following phrase: "The higher the number the smaller the hole the greater the depth of field." Can you see differences? Of course. But again, it takes relatively bigger jumps to really matter all that much. The other two factors matter more.

Quick'n'dirty
Photography is a visual medium, so let's take a look at a couple of examples of point 1, above. I'm saying that your distance setting is the biggest determinant of DoF. Take a look at these two shots:

28mm lens on 24x36mm, ƒ/5.6

Same camera, same lens, same focus distance setting, ƒ/2.8

In both of the above examples, I've focused on the tree trunk in the middle of the frame. I was standing about thirty feet away from it. There are some differences in the two pictures, mainly because most lenses are at their crappiest wide open and this lens is a fairly crappy little lens, so the lower picture looks a little worse. And the much closer branch-tips in the upper right are a little crisper in the ƒ/5.6 shot. But when your distance setting is 30 feet with a 28mm lens, the aperture isn't going to have very much effect on apparent DoF. No matter what the maximum aperture of the lens, you're going to have a hard time getting anything out of focus with this focal length lens at this distance setting.

Now look at the same two-aperture difference at a much closer distance setting, wider aperture first this time:

ƒ/2.8

 
ƒ/5.6

In this case, you can actually see a little difference in the blur, because you actually have some blur. In both of these, the focus was placed on the flower in the middle of the foreground, at the bottom of the frame. (I should have been a little more careful with exposure and framing. Sorry. Then again, if I made everything perfect, I'd never get anything done.) Now, obviously, there's some difference here. There should be: the aperture has been stopped down two stops. The flower is indeed a bit more "isolated" in the top, wide-open shot, and the blurry parts are indeed a bit blurrier. But just look back and forth between the house details at the top of the frame. There's some difference, all right....but it's not that much difference.

So in the top pair, you'd have a hard time opening up the aperture enough to introduce any blur; in the bottom shot, you'd have a hard time stopping down enough to eliminate all the blur. The point is that the top shot is pretty much going to have no blur, and the bottom shot is going to have some blur, with this camera and lens. And that's because of focus distance, mainly.

In any case, where DoF is concerned, don't sweat the small changes too much—whether this or that sensor is slightly larger or slightly smaller, whether this or that lens is ƒ/1.7 or ƒ/2.4, whether you're focused at 10 feet or 12 feet. When learning to manage sharpness and blur, it's the big swings that make more noticeable differences.

Hope this helps. But if it's just more verbiage added on to the great whacking pile of internet blather about DoF for you, well...at least you've got my sympathies!

Mike

Send this post to a friend

If you need it, you need it.

Featured Comment by FS: "I am surprised that nobody has referred so far to this little gem: The Ins and Outs of Focus by H. M. Merklinger (Warning: PDF download):

http://www.trenholm.org/hmmerk/TIAOOFe.pdf

"This is about infinity focus versus using DOF scales. The idea ist to focus on infinity, and then stop down so that the actual opening of the aperture is about the same size as the smallest object to b resolved."

Mike replies: I was looking all over for a current source for Harold's book yesterday, and was stymied. I had no idea the entire book was a free PDF download. Thanks!

Harold's methods are a bit quirky, but his two books are a fine brief introduction to the understanding of focus and DoF.

Comments

[Paul Ewins]

I have seen a photo of a tattoo saying "Born to Loose" which I'm fairly sure wasn't meant to be quite so profoundly truthful.

[Jammy]

I one up your DOF discussion and raise you a discussion on the relationship between diffraction, aperture, sensor size, and resolution.

Now there's a topic that makes me want to grind my teeth with all the miss information I see being spouted. :-)

[Mike Hess]

Once I started doing photography seriously as a hobby (before doing it professionally), I obsessed over DOF. I had to have shallow DOF. Then I started to realize that unless you are very good at focusing, or had a very good autofocus system, shallow DOF was a PITA. When you start using longer lenses, the difference in focus is so severe in a wide open shot, that most photos were throw-aways.

That being said, I have come to love the shallow DOF look for certain shots. One thing I do like about shallow DOF is that it can make a crappy lens seem sharp. Even if a lens is technically horrible wide open (bust out your focus charts or shoot brick walls). The 'contrast' between the in focus, compared to a nicely blurred out-of-focus can make the soft wide open shot look sharp. I use that technique when possible. I have noticed that my Canon 50mm 1.4 is no where near as sharp as my Pentax 50mm 1.4 when both are shot wide open. But, once to print a picture (yes some people do that other than just pixel peep) the difference is small with proper technique.

In other words, I agree with you PS from before, with just a slight (long winded) editorial added.

[Dale]

Ok Mike, this post actually twisted my arm into reading ALL of the others, even though I agree with you completely.
I thought "this" was about the image.
About a hundred years ago (well ok, 40 years ago)I read my "Amateur Photographer's Handbook" (Aaron Sussman if my memory still works)there were a number of memorable statements: 1. Always get the sharpest possible negative you can given the circumstances, you can always kick the enlarger later if you want to show "motion".
2.THERE IS ONLY ONE CRITICAL PLANE OF FOCUS, EVERYTHING ELSE IN DEPTH OF FIELD IS A COMPROMISE RELEGATED TO "USABLE" FOCUS.

You stated this clearly about 2/3 of the way down. Oh ya, about that diffraction stuff. I don't worry about it much and I am very cavalier about my use of ridiculously small apertures when I find them useful to maximize "usable" focus. I do have to admit I miss those way cool colored lines that helped with that hyper-focal stuff.Now I have to think and these days one more item to think about is a challenge.

I will say that I stopped worrying about correcting grammar/spelling on the NET and decided that if I convinced myself that it all just people from "away" that were making valiant attempts at English. It was much easier than thinking that Jay Leno's Jay-walking contestants really are the standard these days.
Can we really blame/thank Jack Kerouac for enabling writers to abuse the written word?

Have some fun will ya?
dale

[georgec]

Thank you for another clear explanation of how the world turns. This will be a nice post to point people to when they have questions.
I have one small question / piece of input that may very well be the internet drivel that makes your fingers tremble when you eat cereal, but please regard it as a self-check with constructive purposes.

What I make of the endless "small sensor, too much DoF" complaints, which ultimately results in "slow-or-not-fast-enough lens for a 4/3 sensor" ramblings, is simply as follows.
The amateur photographer sees pictures of scenes with buttery bokeh where the main subject is at about 10 feet into the picture, something that was done with a "fast lens" on a 35mm sensor, for example. They go to their APS-C or 4/3 or even their mini-sensor pocket cam and find that, at the same distance-to-subject and the same aperture, the DoF becomes larger than they want. They want distance-to-subject to be the same, so you way: go closer. But the point is, they want the same perspective as well. So the factor that appears to need change for the internet masses, is aperture.

Translate this to the 17/2.8 issue (perhaps a reference to Pulp Fiction would be in place here by calling it "the Pen Situation"), and you get the following. The masses want the bokeh of a 35mm sensor/film camera, but the sensor is smaller so the DoF appears larger than wished for. Going closer to the subject does not solve the issue, because they want to have a perspective that resembles what the big boys use. So they want a bigger hole in their lens to compensate for the small sensor to "throw the background out of focus". And BAM! there you have low 'n' dirty internet-weaned amateur photography logic. Or not?

Wow, upon reading back I am starting to feel like a kid with a wet towel. Your blog is greatly appreciated.

[Kosma]

This post seems to be made entirely of frustration.

[John Camp]

Thank you. I may write the essence of this on a wallet card -- it's the best, clearest explanation I've ever encountered, and explains every one of the situations I posed in my post yesterday. I don't think I need to know more details. My DOF *problems* were purely experiential [is that a word?] as opposed to theoretical, and this explains them, or at least tells me where my problems lie.

Uh, but...if I'm trying to take a photo of a Canaanite seal (diameter of a dime, thickness of an M&M) from a few inches, would I be better off using the Nikon 60mm Micro Nikkor (shorter focal length), as opposed to the 105mm Micro Nikkor (from twice as far away)? Or wouldn't it make any difference in this particular case, since I'm trying to hold the image size constant (filling the frame as much as possible) while focus length and distance cancel each other out? Both lenses are f2.8, not that it matters, since I would be shooting at about f11.

And now for the final important question. In the picture of the two trees, why is the sky deep blue in the holes in the leaves on the bottom tree, but on the edges of the frame, and in the top picture, it's light blue? I have a feeling it has to do with the effective aperture of the tree-holes, but I'm not sure.

[David]

Mike, great summation of the practical aspects of DoF. I shoot most of my shots at f/8 and for normal landscape shots, it's plenty of DoF. Using the same lens and at f/8, I can still get nice shallow DoF when shooting from a few inches away, and I'm not working with a tiny slice of in-focus area.

[Amin]

Mike,

This is a crystal clear, practical demonstration. I get the sense that the entry itself is akin to writing "The comments on this entry are closed", but I'm going to try and make a comment or two anyway :-p.

The issue I was getting at with the E-P1 and 17mm lens is that a lot of folks will plan to use it mostly for photography of people. The people I photograph don't tend to be as far as that tree or as close as those flowers. Thus I most often use a "35mm equivalent" lens to photograph people at the sort of distances where I might expect to create more blur than you've shown in the first (tree) example and less than you've shown in the second (flower). I am guessing that most others who use a 35mm lens to photograph people have this in common with me.

If you can agree with what I've said about distance, then we come to the second most important determinant of DOF, focal length. 17mm is a short focal length compared to what many of us are accustomed with regards to a "35mm equivalent" lens.

The long and the short of it (bad Amin) is that the majority of people photographs made with the Olympus 17mm lens are going to have nearly everything in sharp focus. To ye olde photo dawgs, that probably sounds like a great thing. To that great majority of today's photographers who use "point and shoot" [sic] cameras and are wondering how to get "that look" they've seen before, it's not so great. As usual, I'm somewhere in between.

Amin

[Mark Sperry]

Maybe a dumb question...but:

Do you get less depth of field when you go up in formats because the lenses are longer or because the film size is bigger?

Like there are lots possibilities for shallow depth of field with a 300mm lens on 8x10, but it's less apparent on 4x5 with a 150mm. Is this simply because the required focal length for 'normal' is longer?

[Mike Johnston]

"Do you get less depth of field when you go up in formats because the lenses are longer or because the film size is bigger?"

Because the lenses are longer. The format actually makes no difference, except that it changes the angle of view of the lens/film or lens/sensor pair.

Mike

[Mike Johnston]

"And now for the final important question. In the picture of the two trees, why is the sky deep blue in the holes in the leaves on the bottom tree, but on the edges of the frame, and in the top picture, it's light blue?"

I don't know! Fascinating, isn't it? I need to look into that....

Mike

[Robert Howell]

Ask yourself thish: are you taking a photo of the subject or the lens?

And if that doesn't make any sense you should taste this Alsatian riesling I'm tippling.

Happy Patriarch's Day!

[Steve]

http://luminous-landscape.com/tutorials/dof2.shtml

[Dave Fultz]

Sorry to bumble into this glass shop like a buffalo with dermatitis but isn't this (fun) discussion a bit too much?

Long, narrow lens: shallow DOF
Short, wide lens: broad DOF
Teeny tiny sensor with either of the above: broader DOF

Photograph of buffalo from close: nose in focus, tail not so much, very nervous photographer.

Photograph of buffalo from comfy high ridge above the plains: nose and tail in focus but buffalo seems rather small.

Dave

[Rana]

It's at moments like this that I'm glad my understanding of photography comes from a mixture of lessons from my dad when I was a kid, self-education, and lots of trial and error, and not from some formal education. Less BS that way, I believe. Either something works and is helpful, or it doesn't and isn't.

One of the things I really appreciate about your posts is that there is an admirable lack of arcana and techie ****-waving. Photography is a medium; cameras and lenses are machines used to transform light into photographs. Some do it better/differently/with less/more user error than others, but that's what it all boils down to in the end, whether you're talking pinhole camera or the latest techno-whizbang. Your ability to balance camera geekery with friendly explanations of basic principles is unfortunately rare.

I've never really understood the need to turn it all into a big competition for coolness points or some sort of mysterious Photographer Street Cred. Just go out and take the pictures, you know?

[ctein]

Dear Steve,

That Luminous Landscape page is completely wrong. I've explained to Michael it's wrong, and he understands it's wrong. I don't know why it's stayed up... but it's wrong.

Depth of field is not dependent only on on-film magnification and independent of focal length. At low image magnifications (e.g., outdoor photography) focal length is a DOMINANT determinant of DoF.

Plain and simple, the text on that page is false. The photos on the page don't even support the premise.

No, this is not arguable. It is incontrovertibly provable, both by correct experiment and by calculation.

Ignore the page. Delete its URL from your bookmarks. Expunge it from your memory.

Do not refer other people to it.

pax / Ctein

[Rex Kersley]

In the OLD days lenses used to have little marks showing the depth of field for a given aperture. Made it all a lot easier.

Rex

[Johnny R]

In this case, you can actually see a little difference in the blur, because you actually have some blur.

Next time I´d like to see the Oly 17/2.8 vs a 35/2.8 on FX. Can you do it?

[Mike Johnston]

"Next time I´d like to see the Oly 17/2.8 vs a 35/2.8 on FX. Can you do it?"

Not yet, but if/when I get an E-P1 to fiddle around with, I'll be happy to.

Mike

[ctein]

Dear Mike,

I have found that building up a collection of columns have been very helpful in dealing with my 'tech police' irritations.

When I find a topic being consistently misaddressed, I realize it's good fodder for a column.

There, I can lay down "Da Woid!"

When the same misconception resurfaces, I just have to write "Dear X, I believe this column addresses what you said. Please read it, and afterwards we can discuss it further, should you wish."

It's great for my sanity and blood pressure!

pax / Ctein

[Zach]

Preaching to the converted, Mike :)
I'm going to start bookmarking these for referring to others.

[Lucas]

Thanks Mike! :)

[Speed]

I learned photography using an SLR that closed the aperture when the exposure meter was activated -- what I saw was what I got. It's the "most missed feature" of my modern DSLR.

[hookstrapped]

Thanks to georgec and Amin for laying out the DoF issue as it relates to sensor size and what people might want out of their new Olympus.

Isn't that what the interchangeable lenses are for? If you want to take portraits with nice bokeh, you shouldn't be using the 35mm equivalent lens, but instead slap on (with adpter) a portrait-length lens.

[Iñaki]

I know where all this is coming from. The Autocorrect feature of Word and other text processing software is going to ruin languages and get grammar nerds a nice headache. Then, left alone in front of the web browser, there´s no help as for correcting misspelled words.

And it is very, very irritating how this software tends to correct "centre" into "center".

About depth of field. I´m still not able to control the depth of field. Years after years of trying it, I remember I gave up just the moment I got a picture which was surprisingly out of focus, with the wrong depth of field, and with the wrong exposure; but then, it just works [and I actually am rather fond of it].

By the way. You can not even imagine how irritating is the spanglish mockery I come across with native english speakers. "El cheapo", "cohones" [which, incidentally, is misspelled big time] are too very often misplaced and misused.

[hkki]

nice post

[Iñaki]

Hye Dave:
No, it is not that simple.
In fact, the first element of the equation is ALWAYS the distance of the sensitive element to the subject [be it the sensor, be it the film].

That is the reason why an ultra wide angle [such as the Pentax DA 14 2.8, to put an example I´ve used], which can come very very close to the subject and isolate it.

[Steve Smith]

can't comment on the DOF stuff but I hate the 'loose' - 'lose' substitution as much as you.

Add that to errors in 'there', 'their' and 'they're' and my favourite so far (from an ebay listing) of 'Paypal excepted' - the reverse of what was intended.

[Miserere]

Don't apologise, Mike! I now have the following verses printed out on my wall:

Mike's Letter to the Depthoffieldians, 14:2
And then came the theoreticians of the shared armchair and their picayune calculations of sensor size and diffraction and all else that's much talked of in numberless fora but wrong...

...And the users of large DoF shall inherit the Earth. I'm quoting this last bit from memory, so maybe it's not quite correct, but it was something like that.

[B.J.Scharp]

"And now for the final important question. In the picture of the two trees, why is the sky deep blue in the holes in the leaves on the bottom tree, but on the edges of the frame, and in the top picture, it's light blue?"

Isn't that just chromatic abberations of the lens showing up at the light-dark transitions? The dreaded `purple fringing' that everyone who isn't worrying about sharpness or DOF is worrying about. (the reason why you should only buy L/ED/ASPH/APO/EX/Limited/etc-glass*)

That would also explain why the effect is less on the stopped-down version (as stopping down minimizes all bad habits lenses might have, I'm always told)

*: Sarcasm Alert

[expiring_frog]

@John Camp: Maybe because the clouds, um, moved between the shots?

[Lars Clausen]

I totally agree with your P.S., losing your cool or not. I always got the impression that the "I want to be able to limit my DOF" argument for going full-frame was bogus, simply a way to justify the extra expense over APS-C. It has been trotted out in so many now obviously false prophesies about the upcoming demise of APS-C and other smaller formats. There's nothing magical about the 35mm format in particular, but it is close to a sweet spot where you get the choice between subject isolation (difficult on compacts) and wide DOF (difficult on large format). I just don't buy that a difference of 1.5 or 1.6 or even 2 in the thin end of DOF is worth that extra price & size.

[fiddlergene]

I don't understand what all the bouhaha is about DoF. It's so simple and easy to understand. Really, it's just common sense. OOPS! If sense were so common, more people would have it, huh?

[David Bostedo]

"I don't know! Fascinating, isn't it? I need to look into that...."

It's hard to tell for sure in the pictures, Mike, but maybe some clouds moved behind the tree? Either that, or you accidentally left your "show blue sky through trees" art filter on...

[beuler]

«I might have lost my cool a bit...»
Thank god for that! I have always suspected that, in practice, DoF is just as you described here. But the usual internet forum discussions based on brutally technical articles that I have gone on to read (and understand btw, in the academic sense), have made me lust for lenses I can only dream of affording. You have killed my lust by reminding me, ONCE AGAIN, how one can throw a background out of focus while keeping the desired subject in apparent focus. No fast lens needed, thanks.

[Alex Vesey]

Nice article on an often over-wrought subject.

Just to enjoy the spring foliage--what is the name of the purple flower? I have been trying to figure it for some time, but alas, I am no gardener.

Alex

[Eamon Hickey]

>John Camp wrote:

Uh, but...if I'm trying to take a photo of a Canaanite seal (diameter of a dime, thickness of an M&M) from a few inches, would I be better off using the Nikon 60mm Micro Nikkor (shorter focal length), as opposed to the 105mm Micro Nikkor (from twice as far away)? Or wouldn't it make any difference in this particular case, since I'm trying to hold the image size constant (filling the frame as much as possible) while focus length and distance cancel each other out?

It does not make any difference in this case. In practical terms, you can indeed explain this phenomenon as the greater focus distance canceling out the effect of the longer focal length.

Because I don't want to turn this into yet another arcane DOF discussion (I share Mike's opinion of them as a horrible scourge), I won't go too deeply into the "why" of this. But the fact that it is true is why I, personally, am slightly uncomfortable with telling people that longer focal length lenses have less DOF than shorter focal length lenses. In practice--i.e. in about 97% of real-world situations--it works that way. But when you must hold image magnification constant--i.e. you need the subject to be one particular size within the frame--changing focal lengths will not change DOF (although it can sometimes change, very subtly, the appearance of distant out-of-focus backgrounds).

[Richard]

All this makes sense to me and have known this for a while, however I do have a question in regards to Zoom lenses mostly because I was told this statement which was contradictory to what I had thought. That is, a Zoom lens has a fixed DOF characteristic, you will have the same DOF at f5.6 @ 70mm 50 feet away as you would at f5.6 @ 200mm 50 feet away. True or False. If DOF is something you want to utilize to its best benefit is it better to use a number of fixed focal lengths over one zoom

[Max]

Quite good explanation about the thick issues, which is what most people miss when they start delving into this.
I remember going crazy over landscape shots DOF on 6x9 film. One big thing about such formats is the dof distribution (what is sharp and what is REALLY sharp). But that's just overkill with today's equipment. Mainly what I got from all that is, even if the scales on the lens tell you it can be done otherwise, the centre of your focus range SHOULD be on what you want in focus.
Or go to f/32. But there's nothing like a photo of subjects at different distances were the obvious sharpest point falls right in the middle, just nowhere, a strip of sharper dirt on the floor.

[Charles Laird]

OK, as an unskilled amateur here is what I conclude from the article. The solution to this DoF problem is to carefully shoot a lot of pictures with each lens and study the results many times over.

[Eric Ford]

Nice explanation. Actually, since I started to shoot in 35mm format, I've run into more trouble due to the thinner DoF compared to DX than I expected. On a cropped sensor f/5.6 seemed like enough DoF for most scenes, but on full frame, even with wide angle lenses, I find 5.6 isn't enough.

[Bill Stickney]

I've never been inspired to post on any of the photo sites or blogs until spending time on TOP. The whiplash experience of switching between aggravation and amusement is ample incentive to just close the browser. Mike's philosophy (and Rana's more specific comment above) help to maintain a 40+ year passion to "Just go out and take the pictures....", whether personal or professional.

[Martin Doonan]

For general photography with normal focal lengths I use a four point scale of DoF:
f/2.8 - not a lot
f/4 - some
f/8 - a lot
f/16 unlimited
Works pretty well.
(Yes, I know all the science, numbers, maths etc.)

[John Taylor]

Hyperfocal distance… (sorry i couldn't resist! i use to routinely shoot with a 24 mm ƒ 2.8 lens (35 mm) lol!)

[Matthew Miller]

I think the obsession comes from this: gadget-oriented buyers purchase a dSLR because it's _clearly_ the better, serious, "pro" way to go. You don't even need to look at the marketing material -- you can tell from the price tag!

When they show their new gadget to their buddies with compact cameras, they want something amazing to tell about that their camera can do that the compact just can't. Noise and high ISO performance is a matter of degrees, and the ability to interchange lenses seems like a *downside* to the buddies. The compact has a better video mode, even, and you don't have to peer through that viewfinder. The scene modes probably aren't as good, and no face detection! But there's one thing: the big sensor allows a much narrower DOF than the compact, and so:

"Ah, but I really need the DSLR Look for my Art. You know, that really dreamy bokeh blur. That's why people spend the big bucks for these things. What you've got is good for snapshots, but this is what the pros use, because they need that control over depth of field."

[Mike Johnston]

"what is the name of the purple flower?"

Alex,
No clue. I bought 'em, planted 'em, and they're growing in front of my house. But ask me what they are with a gun to my head, and it'd be sayonara Mikey.

I used to work in a garden center, too.

Mike

[Mike Johnston]

"OK, as an unskilled amateur here is what I conclude from the article. The solution to this DoF problem is to carefully shoot a lot of pictures with each lens and study the results many times over."

That's exactly it. That's both how you learn your equipment and materials, and a pretty good description of how you become a better photographer. I used to settle down for bed at night with a stack of prints, and look through them, just for pleasure.

Mike

[David Bostedo]

Ctein (or Mike) - I would love to see an article explaining what is wrong about the Luminous Landscape article referenced above. Even if it involved equations! Any pointers?

[John Camp]

Further up this comments column, I asked why, in Mike's two tree examples, the sky seen through holes in the bottom tree is a darker blue than the sky in the broader parts of the photo, or in the top photo. I have formulated a simple possibility (and no, the clouds didn't move.)

I think the answer is that in the majority of the sky showing, light is coming from all directions to hit the sensor, overloads it, and whites-out the sky. In the holes in the trees, light is coming from only a small part of the sky -- that directly visible as a hole, or, more accurately a tunnel -- and is surrounded by dark foliage which effective acts as a *flag.*

The effect now seems to me to be visible in the top photo as well, but not nearly as strongly, which suggests to me that the phenomenon is time-related, rather than aperture related. That is, in the bottom photo, the shutter was not open as long (because the overall exposure is roughly the same, but the bottom photo was shot at f2.8 and the top at f5.6.) But because the leaves were effectively acting as an irises, of sorts, they normalized the amount of light coming through, so for those holes, it didn't matter too much what f-stop was used. The only thing that mattered was shutter speed, and the faster shutter speed reduced the light striking the sensor through the holes, thus allowing it to appear as blue rather than as a washed-out white.

That's my story, and I'm sticking to it.

I understood the entire DOF conversation until Eolake stuck his oar in. Since it seems that a very large number (and proportion) of photographs are taken with the idea of a constant size in the frame -- that is, the human face, singly or in groups, looking at the camera, shot from the waist up, with an inch or so of space above the head -- his point seems relevant. I just don't know what it means, except that, for me, I should probably shoot portraits and groups at 75mm or more (so noses don't spread out) and walk backwards until they fit in the frame. Does that solve the Eolake conjecture?

[Iñaki]

"Isn't that just chromatic abberations of the lens showing up at the light-dark transitions? The dreaded `purple fringing' that everyone who isn't worrying about sharpness or DOF is worrying about. (the reason why you should only buy L/ED/ASPH/APO/EX/Limited/etc-glass*) "

Are we talking about lenses or a medical prescription for a very rare and exotic unknown condition?

[erlik]

But when you must hold image magnification constant--i.e. you need the subject to be one particular size within the frame--changing focal lengths will not change DOF

Eamon, that is true. But when you want to keep the same size, you have to change the distance to the subject with different focal lengths.

[Oskar Ojala]

I don't care about DOF in this case; I would have preferred 17/2 over 17/2.8 due to low-light performance alone. Sometimes it is just very dark! Of course, then a focus scale could help too. But in any case I use 6x6 for DOF control, 35 mm is not entirely satisfactory for that, which is also why I consider the DOF-argument against u4/3 as bogus. Many times DOF just doesn't matter and as photographers we need to make compromises with the equipment we are carrying.

[Andreas]

You know, when I was learning the basics of photography, I read several of those DOF articles, and not one of them mentioned your first point about distance to subject. I managed to figure it out myself experimentally, but I find it odd that the "recipe" for background blur is often given simply as "shoot wide open".

[Miserere]

David Bostedo wrote: "Ctein (or Mike) - I would love to see an article explaining what is wrong about the Luminous Landscape article referenced above. Even if it involved equations! Any pointers?"

I second this motion, and in fact demand equations for the ubergeeks among us (raise your hands, don't be embarrassed).

Thanks Ctein!

[Philip Storry]

DOF is a subject that tore me apart.

I have some cracking lenses. I'm lucky in that regard, and I know it.

But when I'm using these hulking great masterpieces of optical joy, I'm always a little scared that I'm going to get a result with too little DOF. Which can just ruin a picture.

Learning to use a lens takes time, especially if it can do a thinner DOF than your other lenses. In the process I've lost some shots because I was too eager to open it all the way up and "use what I've bought".

It's like owning a very fast car. Just because she'll do 180mph, doesn't mean you can control it. Maybe there's a reason most people are happy with around a hundred miles an hour, eh?

But what if you needed that speed for an assignment?

And think of the prestige of the fast car!

The occasional use and the small ego-stroke make lenses capable of DOF very tempting. So I was torn.

I've come to realise that there's a time and a place for razor-thin DOF, and it's nowhere near as frequent as I'd first thought.

Funnily enough, my favourite lenses top out at f/2.0, or f/2.8. I have an f/1.4, and it's good. But it's not my favourite. Getting to f/1.4 has cost it some sharpness, and it lacks that mysterious sweetness.

So my faster lenses aren't actually about speed for me - it's the higher build quality and the better optical performance that they bring which I now really value. I like using them, but don't tend to use them wide open that often.

And looking through my pictures, it seems that most of the shots I'm happiest with are out towards f/3.5 and higher. So this isn't theoretical, this is practical observation based on what I shoot.

(Or am I just not using my lenses to their fullest capabilities? Oh, no! Now I'm torn again as to whether I need thin DOF!)

Now, all of this reasoning against a razor-thin DOF makes perfect sense. Many people will agree with me that thin DOF isn't needed much. Right up until I mention I shoot with Four Thirds.

Then, apparently, I'm crazy and don't have enough DOF. Because although I'm afraid to use a very thin DOF for losing the shot, and although I rarely use it anyway... I don't have the option, so I'm evidently missing something.

*shrugs*

This is why I don't bother with internet forums anymore. (Or even some apparently reputable magazines...)

[Paul Mc Camm]

........and then there is the myth about stopping down the lens to check depth of field. Never worked for me when I was young and using full frame 35mm slrs so what hope is there today ?

[Mike Johnston]

"I don't care about DOF in this case; I would have preferred 17/2 over 17/2.8 due to low-light performance alone. Sometimes it is just very dark! Of course, then a focus scale could help too. But in any case I use 6x6 for DOF control, 35 mm is not entirely satisfactory for that, which is also why I consider the DOF-argument against u4/3 as bogus. Many times DOF just doesn't matter and as photographers we need to make compromises with the equipment we are carrying."

Ahhh, the voice of reason.

Mike

[01af]

All articles, essays, and comments about depth-of-field contain errors, misconceptions, or inaccuracies---including Mike J.'s (sorry Mike!), and including Michael R.'s poor so-called DOF tutorial on Luminous Landscape ... with one noteworthy exception: See Paul van Walree's excellent and error-free (as far as I can tell) articles at http://toothwalker.org/optics/dof.html and http://toothwalker.org/optics/dofderivation.html. Very recommended reading for the technically inclined!

-- Olaf

[ctein]

Dear Eamon,

You said, …"when you must hold image magnification constant--i.e. you need the subject to be one particular size within the frame--changing focal lengths will not change DOF …"

That is only correct for close working distances (e.g., John C's situation). Focal length matters hugely for small image magnifications.

There should be a column up tomorrow (right, Mike?) which lays this out in more excruciating detail (which should satisfy your request, David).

And, anticipating the question, no I don't know why Eamon's incorrect assertion is universally repeated. The perversity of the universe.

pax / Ctein
==========================================
-- Ctein's Online Gallery http://ctein.com
-- Digital Restorations http://photo-repair.com
==========================================

[Stephen S.]

I love how clear and concise this is. If you want shallow depth of field, move closer. Clever folks like Cinematographer David Mullen ASC have the same understanding:

"However, I just finished a TV show on a 35mm-sensor digital camera, when we had contemplated using a 2/3" camera... and I was glad we went with the 35mm-sensor camera in the end because we didn't shoot many close-ups on the show, lots of medium shots in small houses, at an f/2.8 generally, usually on a 35mm prime lens. So it was helpful to have the fall-off in focus since I couldn't get the walls farther away from the subject and we weren't pushing in tighter either to make the focus drop off faster."

(Italic emphasis is mine)

http://reduser.net/forum/showpost.php?p=339839&postcount=10

[ctein]

Dear Olaf,

Thank you so much for those links. To the best of my cursory read, the pages are 100% correct.

The author even goes to the trouble of including the effects of pupil magnification, something I never bothered with (I always used the P=1 approximation). Sweet!

I've bookmarked these pages for future reference.

And I imagine that Miserere is in ecstasy!

pax / Ctein

[Damon Schreiber]

Well, I must say that this whole discussion has me scratching my head. I thought I understood DOF, and I still think I do, but I hadn't realised the extent to which people enjoy *arguing* about it.

Can't we all be right? Mike says that Internet sources are 99.75% wrong when discussing DOF. Ctein says that the Luminous Landscape article is wrong. Olaf says that TOP's DOF article as well as LL's one are both wrong, but that Toothwalker's article is right. I read Toothwalker's article and it seems to support the LL article's premise (and so do Eamon and Erlik above). This is what I mean about scratching my head.

Unlike you, Mike (I think), I can't dismiss shallow depth of field as an advantage, or at least a characteristic of 'full frame' sensor cameras. It's demonstrably true that a picture at a given angle of view with the same f number on a cropped sensor camera (say an APS-C camera like my Pentax K10d) will have more depth of field than on a full frame camera like the Nikon D700 I sometimes use. For example, put a 50mm lens on the D700 and a 35mm lens on the Pentax, both at f/5.6, and you'll have similar angles of view. Shoot the same scene focussing on the same point, and the Nikon will have less depth of field than the Pentax. I've done this.

I'd say there's about a stop difference. Compound this with the fact 50mm lenses are commonly f/1.4 or f/1.8 wide open whereas 35mm lenses are more commonly f/2.8 (at least mine is), and the shallow depth of field becomes even more common on the Nikon. Yes, I can mount a 50mm lens on a Pentax, but it's not a 'normal' lens anymore.

Now you may or may not have a bias towards greater depth of field , but that's another issue. When looking at video as opposed to film, huge depth of field is what immediately jumps out even before many of us realise it, and makes us say 'that looks like video; that looks like a home movie.' This is the main reason why videographers are excited about DSLRs with video modes, so that they will be able to achieve shallow depth of field approaching that of film cinema cameras (for a lot less expense). Now whether they will be able to actually achieve focus is another question.

It may be true that not that long ago, people had a hard time getting away from shallow depth of field, mostly due to slow (and physically large) film. But that limitation also became an aesthetic, one that people are now attempting to emulate, trying to achieve that 'film look'. Is that so wrong?

Yes, wrongly judged shallow DOF in a photo can be annoying, but so can too much DOF. Can everyone at least agree on that? I don't argue that full-frame cameras are better because they tend to have shallower DOF, I'm only noting that they do.

Oh, and the purple flower is Spirea.

[David Bostedo]

Olaf - Thanks for those links. I haven't quite read through all the equation derivation, but the main article is great. (Of course, I liked the Luminous Landscape one, too, only to find out it's an incorrect over-simplification... live and learn.)

I'm curious to see what Ctein adds tomorrow!

[ctein]

Dear Damon,

You need to reread Toothwalker's articles more carefully. They don't support what the LL page says. You are misunderstanding the information being presented.

Mike's comments on DoF are simplified rules of thumb that are substantially correct for the circumstances they describe, although there are indeed minor errors. But none, you will note, so egregious that I thought there was a need to call them out; they are on the level of nitpicking.

The assertion on the LL page is simply erroneous. It isn't even supported by the photos presented (not that it could be, since it's a false assertion).

The photos on the LL page create a weird optical illusion. At first glance, it appears they confirm the 'focal length doesn't matter' claim, because the amount of detail in the tower in the background is about the same in all the photos.

BUT... the tower is very much smaller in the WA photos than the telephoto photos! In terms of line pairs per millimeter on film (which is what DoF's about), there's a lot more fine detail in the WA photo-- the far background is much, much sharper, lp/mm-wise.

It creates a visually compelling, but entirely coincidental, illusion.

I really don't know why MR hasn't pulled down the page... other than maybe having about 12,000,000 more important things to do!

pax / Ctein

[JOHN MCMILLIN]

To all the grammar loosers out they're, all of you still on the lose, I have just one thing to say.

I eschew homophones like your word-mangling kind, so just get out of my faze!

[B.J.Scharp]

"Are we talking about lenses or a medical prescription for a very rare and exotic unknown condition?"

Oops, drop one `B', add one `R' there. My excuse is not being a native english speaker, but I should just have looked that one up.

[Miserere]

I've bookmarked these pages for future reference.

And I imagine that Miserere is in ecstasy!

pax / Ctein

I read the articles yesterday evening and woke up today with a hangover. They were a good read :-)

[Juanan Albarral]

Really nice article!
(haven't readed it yet, but everyone in the comments says that. Isn't it enough? It's like when you search for advice on buying "the best of the best", you have to hear the wisdom of the masses, they can't be wrong :)

"When learning to manage sharpness and blur, it's the big swings that make more noticeable differences."
The big swings? Always believed that a big tilt was better...

By the way, what a nice crappy little lens! It's wonderfully crappy in the corners of the frame!
Even stopped down!
Even at web size!!
And the "roundness" of the background blur at 2.8 is... is... well, it's round.
What's the name of that gem? Seriously. Not everyone has to have the same preferences about what a "good" lens is (have you ever tried a Lubitel?). Sometimes a certain amount of crappiness in the form of corner fuzziness and wild light fall-off is a lot better to isolate a subject than the elusive "shallow-DOF-at-near-infinite-focus-distance".

Hmmm... "Appreciating crappiness"... That will make for a good article.

Juanan

P.S. Sorry for writing this comment a week later than the date of the post. Commercial photography schedules can be maddening. And I'm only an assistant!