PDA

View Full Version : Circle of Confusion query



peter brooks
31-Dec-2018, 10:04
I'm looking at hyperfocal distances for a wide angle lens on 10x8 at a fairly small aperture. A 'normal' CoC for 10x8 is (I think) 0.2, whereas a CoC for 5x7 is 0.15.

Would I get a better result (in real life rather than just theoretically) using the smaller CoC of 0.15 (or even smaller)? Would it even be noticeable in contact prints?

By my calculations the difference between a 0.2 and 0.15 CoC would move the hyperfocal distance only about 3ft further away from the film plane.

(I think I understand a little of what the Circle of Confusion is about - but seldom has a technical term been so aptly named :) )

Jac@stafford.net
31-Dec-2018, 11:05
A rule of thumb suggests that you close the aperture (larger numeric ƒ) to gain more DOF. Calculating via CoC just confuses practical expectations. CoC relates more to degree of enlargement which is nil in contact printing. Some people become concerned with diffraction with small numeric apertures, but for contact prints it is not likely critical for any aperture on the lens scale.

Jim Jones
31-Dec-2018, 11:26
The CoC is a personal choice. It should be chosen after considering how the final image will be presented, not on what others might recommend. I suspect many photographers would never notice the difference between CoCs of 0.2 and 0.15, but a few certainly will. Remember, a WA pinhole 8x10 camera may have a pinhole of 0.5mm and a somewhat slightly smaller effective CoC, and still produce photos that many consider pleasing.

Drew Wiley
31-Dec-2018, 11:28
I ignore the whole topic. The only thing that counts is how you want the print to look. Unsharpness as well as sharpness is a tool of composition. That's what so wonderful about a big groundglass. And if you need more critical information, a basic loupe will tell you far more than all the hyperfocal formulas in the world.

peter brooks
31-Dec-2018, 11:32
Thanks for the reply. Yes, I understand the relationship between aperture and DOF. The application is a box camera so it will have a fixed focal length (I have done this with 5x7 and am now going to build a 10x8, using f32 for the calculations). Looking at the hyperfocal distance gives me a range of 'usefullness' (so to speak). As the CoC is part of the formula for calculating the hyperfocal distance I was wondering what practical effect (if any) reducing the CoC would have on the final image.

peter brooks
31-Dec-2018, 11:34
I was replying to Jac@stafford.net and in the meantime two more replies... Maybe my reply makes more sense? Sorry if I wasn't clear in the original query...

peter brooks
31-Dec-2018, 11:48
Hmm. On reflection my reply makes no sense - reducing the CoC in the calculation would have absolutely no effect on the final image.

What it does is moves the hyperfocal distance (slightly) further away (which should be a sharper point) and reducing the band of 'acceptable' sharpness. So is the CoC acting like a kind of resolution tolerance?

Jac@stafford.net
31-Dec-2018, 12:22
So is the CoC acting like a kind of resolution tolerance?

Kinda. For 10x8" contact prints use a CoC of 0.2 mm which approximates the resolving power of the human eye under normal viewing conditions. Simplifies things, no?

peter brooks
31-Dec-2018, 12:58
Kinda. For 10x8" contact prints use a CoC of 0.2 mm which approximates the resolving power of the human eye under normal viewing conditions. Simplifies things, no?

Yeah, that makes sense, thanks. I'll stop considering the CoC now :)

(side note: the box camera is me convincing myself that I need some 10x8 holders etc... before long I'll convince myself that I need a 'proper' 10x8 camera... You know how it is... :) )

ic-racer
31-Dec-2018, 12:58
Only you know. Just focus 3 feet out and see how the prints appear.

Also, with the same scene you can go from f90 to f22 and see if you can tell any difference from Airy disk size. You may find (if you do contact prints) f64 works quite well.

I make 2 power enlargements, (16x20in) and, since I'm nearsighted, I peer at the prints with my nose practically against the paper, so I favor f32 or f22 on my 8x10 camera.

In terms of "hyperfocal distance" I only focus there by coincidence when the main subject is there and I want the background blurred. For a landscape, I usually focus at infinity as outlined in Merklinger's "Ins and Outs of Focus."

Page 69, Chapter 10, Rules of Thumb (author's emphasis):

4. If we want anything at infinity to be critically sharp, focus at infinity.

Tobias Key
31-Dec-2018, 13:34
The CoC is defined as the point where what was a point now becomes a disc - that's the point where an image becomes blurred. Lessening the CoC reduces the effective depth of field, and the bigger the print the shallower the depth of field if you are viewing the print from a near distance. So you could define the CoC as the smallest point an lens could resolve or whatever you need to get away with to print at a certain size. It's not an absolute measure. A real life application is the image we can print up to a certain size but not larger because you *just* missed focus. The missed focus means the optimum CoC hasn't been reached but the number is small enough to yield a good 5x7 for example.

Drew Wiley
31-Dec-2018, 19:11
The whole point of view cameras isn't just the large negative, but how you can control depth of field issues in more ways than just stopping down, hyperfocal etc. Tilts and swings are there for a reason.

Bob Salomon
31-Dec-2018, 19:52
The whole point of view cameras isn't just the large negative, but how you can control depth of field issues in more ways than just stopping down, hyperfocal etc. Tilts and swings are there for a reason.

Plane of focus and subject shape. DOF is still controlled by aperture and focus point and CofC.

Oren Grad
1-Jan-2019, 09:38
Only you can decide what works for you.

But FWIW, I bought a Hobo designed for fixed-hyperfocal use with a 120mm Super Angulon, did some tests at f/45, examined the results as I would normally examine contact prints, and concluded that the pictures would be acceptable only when I lucked out and the main point of interest in the picture happened to be at the fixed focus point or very, very close to it. Otherwise, it was a great way to waste 8x10 film at $4-5 per sheet. So I had the Hobo remodeled to include a focusing helical.

To put it differently, I concluded that even with a focal length short enough to give an ultra-wide field of view, the concept of hyperfocal distance was of no practical use to me on 8x10. Even in contact prints, given the way I view them, the effective image magnification is just too large to get away with cheating on the point of focus.

To better grasp the theory and how it relates to the way the pictures come out, I agree with ic-racer that Merklinger is worth a read.

Randy Moe
1-Jan-2019, 10:45
Oren, you just saved me a lot of time.

Thank you

Peter Collins
1-Jan-2019, 11:49
FWIW, I am usually IN a CofC....

Randy Moe
1-Jan-2019, 14:54
Is it the snow?


FWIW, I am usually IN a CofC....

Drew Wiley
1-Jan-2019, 15:06
Don't get fancy telling me something we all know anyway, Bob. I just don't care. I compose in relation to exactly what I want to draw the eye to in the composition by being in precise focus, and what I might want to de-emphasize by being ever so subtly out of focus. I don't give a damn about rules or charts. I do know how to make very very crisp prints. I cut my teeth on big Cibachromes. But that's just half the story.

Pere Casals
1-Jan-2019, 18:57
The application is a box camera so it will have a fixed focal length

If I understand well, you have to decide what hyperfocal settings you are to use, the problem is that when using fixed hyperfocal you are to have distant subjets in the worst case...

You can use this app:

185955

...to understand how the system will behave across subject distance, to decide what f/ and what focus plane distance you want.

Drew Wiley
1-Jan-2019, 19:19
Doesn't Dramamine or Pepto Bismol cure circle of confusion issues?

pgk
2-Jan-2019, 03:24
IMO CofCs are really only a sort of guidance. Calculating them is all very well but in reality few photographers have sufficient control over the final viewing conditions and detailed scrutiny to which prints are put, so they have to be used with a great deal of discretion. When shooting 35mm I tend to use the engraved dof scales for two stops wider aperture these days because people insist of examining images at much higher magnifications than they used to and because digital noise has reduced relative to film noise. Whilst this isn't so relevant to LF, viewer's behaviour has changed and more information is sought even when it can't exist.

Pere Casals
2-Jan-2019, 05:14
Calculating them is all very well but in reality few photographers have sufficient control over the final viewing conditions and detailed scrutiny to which prints are put, so they have to be used with a great deal of discretion...

Please see the context, this is about making a box camera with fixed "bellows draw" and no movement, not about managing CoC in the field. It is an interesting question to find what hyperfocal settings would perform optimally with what limitations.

Probably, depending on what we mainly shot, different fixed parameters would be desired for the box...

pgk
2-Jan-2019, 05:25
As I said, it is guidance. The camera type is somewhat irrelevant unless you also know end usage parameters. If you want an 'infinity' landscape to be in focus and want to print 'large' then focus at infinity regardless. If you want to contact print and view at 'normal' 10"x8" viewing distances then a very different, old fashioned and defined dof can be used to calculate the accepted hyperfocal distance and the camera set accordingly. The 'problem' here is not defining output requirements, not the camera type. DoF and hyperfocal settings are part of a workflow which requires parameters to be specified throughout in oredr for any calculation to be meaningful. If you can't specify output requirements then the CoC can't be properly calculated.

Jac@stafford.net
2-Jan-2019, 07:14
As I said, it is guidance. The camera type is somewhat irrelevant unless you also know end usage parameters.

The OP asked if it mattered for an 8x10 contact and I suggested that a CoC of 0.2 mm approximates the typical resolution limit of the human eye at normal viewing distance. So we have the degree of enlargement (none) and viewing distance so if one must use CoC then 0.2 mm is natural.

To reiterate what ic-racer wrote: It best to have your important subject in-focus. DoF might work, might not.

pgk
2-Jan-2019, 07:37
It wasn't entirely clear to me whether contact prints were the only option for final output. I fully agree though that the point of focus remains critical and DoF (however calculated) should not replace precise focussing .....

Pere Casals
2-Jan-2019, 07:48
To reiterate what ic-racer wrote: It best to have your important subject in-focus. DoF might work, might not.

Perhaps it would be useful if the hyperfocal settings could be adjusted with a variable shim, perhaps in the film holder side... let's say that with just adding some rectangular frames in back (shimming the holder) we would have an optimal job for infinite, for say 20m and for say 3m. Just an idea, perhaps a weird one.

Peter Collins
2-Jan-2019, 11:47
The recent posts tend to reinforce my feeling that I am in a circle of confusion.

peter brooks
2-Jan-2019, 13:59
Thanks everybody for your contributions. I think I understand CoC better than I did before, and what place it has in the formula to calculate a hyperfocal distance.

It tickles me how replies amble gently off topic - my query wasn't about whether using hyperfocal focussing was worthwhile (although the practical experience from Oren is much appreciated), or the benefits of using a view camera and movements.

For the record, I enjoy using my view camera as much as my 5x7 box camera. They are different, the box camera is primitive and delightfully simple (maybe just a step more complex than a pinhole), and quick to use.

In the 10x8 box I'll be using paper negatives (oh no! Cue more hands up in horror!), so a relatively cheap experiment. If it doesn't work out I'll still have the lens and film holders to use elsewhere...

I do love the breadth and depth of photography. :)