# CoC and depth of field

Discussion in 'Exposure Discussion' started by michael_r, Sep 9, 2011.

1. ### michael_rSubscriber

Messages:
6,846
Joined:
Feb 2, 2010
Location:
Shooter:
4x5 Format
A question about balancing depth of field and resolution. Bear with me this might be confusing if I don't write it out clearly enough. Assume 35mm negatives to be enlarged to around 8x10".

In my photographs I want everything sharp (read no selective depth of field). At the same time, whenever possible I prefer not to use apertures smaller than f11 as f16 seems to be where diffraction effects become readily noticeable. So there is always a balance that must be struck. I've started using my brother's nifty golf rangefinder to measure the distances in a scene before focusing. So I'll measure distances to the far away objects, the near objects, and/or the most important objects.

Now, most depth of field tables, including the scales on most lenses (which are not necessarily even accurate) assume a CoC of 0.03 or 0.032 as far as I can tell. But for film that needs to be enlarged a CoC this large is really pushing the limits of what I would consider acceptable sharpness. So I usually try to use a smaller CoC, arbitrarily something like 0.02 or lower if possible (I use the DoFMaster website to calculate depth of field for different CoC sizes using my distance measurements).

What I'm never sure about is whether to favour a smaller CoC with less depth of field or a larger CoC with more depth of field. Which will result in the higher perceived sharpness (all other factors remaining equal)?

Here's an example. Suppose for a scene I've determined f11 gives sufficient depth of field for a CoC of 0.025. Would it be better or not, to stop down to f16 (ie more diffraction) and theoretically get a CoC of 0.015 (ie more depth of field)?

And how much smaller does a CoC have to get from the standard 0.03 for there to be a visible improvement in resolution?

Is 0.02 significantly better than 0.03?

Is 0.01 significantly better than 0.03?

What is the lower cutoff point beyond which we can't really see an improvement (assuming a constant optimum aperture)?

Obviously the enlargement factor plays a major role...

Michael

2. ### Bill BurkSubscriber

Messages:
5,352
Joined:
Feb 9, 2010
Shooter:
4x5 Format
I don't know if this is oversimplifying things, but I just set the aperture at f/16 and use the scales for f/11.

3. ### David A. GoldfarbModeratorStaff MemberModerator

Messages:
18,165
Joined:
Sep 7, 2002
Location:
Honolulu, Ha
Shooter:
Large Format
There are various ways to calculate the value for acceptable CoC, but the standard values vary by format. Bigger format, smaller enlargement factor for the same print size, so a larger CoC is going to be visually acceptable.

Now, if you're like me, you might want things a hair sharper than the standard values, and rather than enter into a complex calculation, I do exactly what Bill Burk recommends above, with the same result: I stop down one or two stops from what the scale or table says to use for the DoF I want, and I don't really worry about diffraction, unless I'm shooting macro at fairly high magnification, where the effective aperture is going to be significantly smaller than the value engraved on the lens. Inadequate DoF when you need it is a bigger visual distraction than diffraction at non-macro subject distances.

4. ### michael_rSubscriber

Messages:
6,846
Joined:
Feb 2, 2010
Location:
Shooter:
4x5 Format
I guess this is really what I'm getting at. Is that statement always true? There is likely a crossover point beyond which diffraction obliterrates whatever additional detail you gain by increasing depth of field. On the other hand, if any amount of blur resulting from defocus always looks worse than any amount of fuzziness from diffraction, then I guess the statement would indeed always be true - meaning for sharper prints you are always better off decreasing the size of the CoC than maximizing resolution. Is this true?

5. ### ic-racerMember

Messages:
7,812
Joined:
Feb 25, 2007
Location:
Midwest USA
Shooter:
Multi Format
Well informed question. You have to take pictures at the various apertures with your system (or look back at at your records) and calculate backwards your acceptable CoC. Again, you have asked a good question and not many people seem to understand that 'acceptable CoC' is a personal value.

6. ### michael_rSubscriber

Messages:
6,846
Joined:
Feb 2, 2010
Location:
Shooter:
4x5 Format
Agreed. I was actually just thinking about trying a few real world tests. It could make for an informative discussion.

Michael

7. ### MonitoMember

Messages:
345
Joined:
May 16, 2011
Location:
Nova Scotia,
Shooter:
Multi Format
Depth of field depends on seven factors, not just the four that most people think of. In addition to focal length, f/stop, and distance, DoF depends on film size, print size, viewing distance, and visual acuity.

This is the only calculator that I know of that applies all seven: http://eosdoc.com/jlcalc/ Explore it and vary all 7 variables.

You can't just go with a given size for the CoC.

Then go with the results of the DoF calculator and/or the DoF markings on the camera lens if it has any.

Last edited by a moderator: Sep 9, 2011
8. ### MattKingSubscriber

Messages:
19,434
Joined:
Apr 24, 2005
Location:
Shooter:
Multi Format
This thread would make a great "Sticky".

9. ### David A. GoldfarbModeratorStaff MemberModerator

Messages:
18,165
Joined:
Sep 7, 2002
Location:
Honolulu, Ha
Shooter:
Large Format
Good idea. I just stuck it on my iPhone using Tapatalk. Did it work?

Messages:
19,434
Joined:
Apr 24, 2005
Location:
Shooter:
Multi Format
It sure did.

11. ### David A. GoldfarbModeratorStaff MemberModerator

Messages:
18,165
Joined:
Sep 7, 2002
Location:
Honolulu, Ha
Shooter:
Large Format
Thanks. The mod functions in this interface can be hit or miss sometimes.

12. ### michael_rSubscriber

Messages:
6,846
Joined:
Feb 2, 2010
Location:
Shooter:
4x5 Format
Wow, I started a sticky thread!

Messages:
6,846
Joined:
Feb 2, 2010
Location:
Shooter:
4x5 Format

15. ### JamesDeanMember

Messages:
31
Joined:
Nov 14, 2010
Shooter:
Large Format
I think this is covered well by Ken Rockwell. http://www.kenrockwell.com/tech/focus.htm

He argues that for any given scene/camera arrangement there will be an optimum aperture when the effects of depth of field blur and diffraction blur are equal. Deviating from this aperture will mean increasing one effect while decreasing the other. Since the sharpness of the image will be limited by whichever effect is greater, you can't do better than when they are equal.

Ken explains it much better than me. More importantly he goes on to show how to determine it. His formula is fairly theoretical, but it seems to give me good real world results.

16. ### HikariMember

Messages:
188
Joined:
Dec 12, 2010
Shooter:
Medium Format
This thread ended on the first response and was confirmed on the one following that.

BTW, I would not trust Ken Rockwell's subjective definition of permissible circle of confusion. There have been better minds than his working on this for over a century.

17. ### JamesDeanMember

Messages:
31
Joined:
Nov 14, 2010
Shooter:
Large Format
If you think this thread is over then stop reading and go somewhere else. Grown-ups are talking now.

18. ### keithwmsMember

Messages:
6,070
Joined:
Oct 14, 2006
Location:
Charlottesvi
Shooter:
Multi Format
Rockwell's writeup on this particular issue actually isn't half bad. I landed on it many years ago when I got my mamiya 6, and found it useful. As I recall, his writeup is quite similar to a Popular Photography article that you can look up. So... try not to dismiss Rockwell's writeup on this just because he comes across as a complete ass on some other things; he probably can't help it

The answer will depend on many things- not only pertaining to shooting but also to printing. E.g. the CoC becomes especially important at large print enlargement factors.

To me, it's important to remember something really fundamental about perception: we seldom miss what isn't there. If you have minor losses of fine detail here and there in a composition, in favour of more consistent focus throughout, I think you will find that the result is perceived favourably. But if you have regions of sharp detail near regions of blur, you will really notice that very quickly. So OOF transitions are an important thing. Think about a superfast lens in 35mm format- you can get disturbing bokeh lines where you least want them, if you're not careful.

But what you want of course depends on your composition and your intentions.

What I just wrote may sound vague so I will give a concrete example. Suppose you are taking a headshot of your dear uncle Bertie. He's got a pleasant, friendly face but (God bless 'im) he also has quite a lot of nose hair. Anyway, he won't be around much longer, so you want to take a nicely detailed shot for future generations to enjoy. You considered some petzvalish work with eyes in sharp focus and everything else creamy, but let's suppose you really want good detail throughout. Do you (a) give more value to critical sharpness and go for say f/8; or do you (b) go for consistent level of detail and go for f/11?

The answer, of course, is (c)... you do both, and you take multiples, because this is a valuable shot and you want to get it just right

The point is that you can debate this and come to many different but reasonable conclusions. And you can easily make a mistake (I think) of putting that plane of superduper sharpness right on the nose hair, and then what? At some almost subliminal level, the viewer will see that. Wouldn't you rather have consistent focus throughout?

I can't answer for you, you have to decide for yourself, maybe the nose hair is such a special thing about uncle Bertie...

It all depends on your subject and your intentions.

So I'd say the bottom line is not to forget that the familiar definition of CoC is based in perception. If you really want to get the largest net amount of detail, the answer won't necessarily be appealing.... it depends on your artistic intentions. And the fact that we photographers concern ourselves with perceptual issues is what makes us artists. If we just followed some formula for maximum detail then, well...

19. ### MaximusM3Member

Messages:
756
Joined:
Jul 11, 2010
Location:
NY
Shooter:
35mm RF
Yep..Ken actually is A LOT smarter than everyone thinks he is. Not only that, but ample credit must be given to him for actually talking a language that most photographers can actually understand and put into practice. That's an art in itself, especially today, when we need less complications to keep film alive.

Messages:
8,729
Joined:
Sep 19, 2003
Location:
fl,USA
Shooter:
Multi Format

21. ### premorthoMember

Messages:
47
Joined:
Mar 24, 2012
Location:
Bombay, NY
Shooter:
Large Format
I've wondered about this as well. I decided to use 100 (or 50) asa films in order to be able to use the lens in 5.6 to 11 mode. Most 35's have a plethora of shutter speeds and a shortage of usable f stops. Diametrically opposed to LF which has as few as one (!) instantaeous shutter speed, but all kinds of usable aperatures. When I was a news photographer (60 years ago!!), on a 4X5 I used f32 and whatever it took in shutter speed or flash to get what I wanted (and more importantly what the art department wanted---they signed the paychecks!):munch:

22. ### RalphLambrechtSubscriber

Messages:
8,729
Joined:
Sep 19, 2003
Location:
fl,USA
Shooter:
Multi Format
the nice thing about working with cocs s that the enlargement factor does not play a role any longer.

23. ### RalphLambrechtSubscriber

Messages:
8,729
Joined:
Sep 19, 2003
Location:
fl,USA
Shooter:
Multi Format
viewing distance and print size cancel each other out and are no longer varibles if a coc is uded.

24. ### andrew.roosMember

Messages:
567
Joined:
Jan 4, 2011
Location:
Durban, Sout
Shooter:
35mm
In his book Post Exposure Ctein suggests that "adequately sharp" (meaning, it appear sharp when viewed alone) requires about 5 lp/mm on an 8x10 print, but that "perfectly sharp" (meaning no difference is visible in side-by-side comparisons if sharpness is increased further) requires about 30 lp/mm! On an uncropped 35mm negative (enlargement factor 7) this corresponds to CoC diameters of about 0.014 mm and 0.002 mm respectively, but the latter is not achievable due to film, lens sharpness and DOF limitations. For "adequately sharp", an aperture of f/16 should give an airy (diffraction) disk diameter of around 0.014mm.

My own inclination is to retain diffraction sharpness and compromise DOF if necessary since I can adjust the focus distance to ensure that the aspects of the scene that I consider most important - typically those closest to the camera - are as sharp as possible, while allowing some loss of sharpness for (due to DOF limitations) in other areas. If you start to lose sharpness due to diffraction, then this affects everything in the photograph, which I find the worse of the two evils.

Last edited by a moderator: Apr 4, 2012
25. ### paul ronMember

Messages:
2,207
Joined:
Jan 22, 2004
Location:
NYC
Shooter:
Medium Format
26. ### RalphLambrechtSubscriber

Messages:
8,729
Joined:
Sep 19, 2003
Location:
fl,USA
Shooter:
Multi Format
Michael,I have a must-read for you on this topicbut it's too big for APUG.please send me a n email to rlambrec@ymail.com.