Focusing in The Digital Era Part Two

[Tim Gray]

I'm not sure what setting the COC to pixel pitch does for you in actual practice, other than re-inforce the notion "in focus" is a plane - not a depth.  It seems that we are moving closer to equating the plane of focus with depth of field.  

Unless I've calculated incorrectly, a coc of .007, f8 200mm @ 10 feet is about 4 inches, less than 1 ft at f22 expanding all the way out to a whopping 8.5 feet at f8 and 100'.    

I'm not arguing that 4" isn't correct in these circumstances with dof being the  "range that is 'acceptably' sharp".  Just wondering at the real world implications of working with dof's that are essentially flat (or diffraction limited).

[BJL]

I'm not sure what setting the COC to pixel pitch does for you in actual practice
[a href=\"index.php?act=findpost&pid=101108\"][{POST_SNAPBACK}][/a]

It means that the image of any subject counted as in-focus by that calculation is not significantly less sharp that what could be achieved by focusing on that subject. In the case of focusing at the hyperfocal distance given by such a calculation, it means that everything in the far distance is as sharp as it can get with that sensor, whereas focusing at the traditional hyperfocal distance based on traditional CoC of about 1/1000th of frame diagonal leads to the images of far distant subjects being detectably less sharp than is possible with a sensor of over about 4 or 5MP.

How important this is in practice, I do not know!

[Jonathan Wienke]

How important this is in practice, I do not know!

With a 1Ds, it makes a big difference.

[howiesmith]

I'm not sure what setting the COC to pixel pitch does for you in actual practice, other than re-inforce the notion "in focus" is a plane - not a depth.  It seems that we are moving closer to equating the plane of focus with depth of field. 

[a href=\"index.php?act=findpost&pid=101108\"][{POST_SNAPBACK}][/a]

Setting CoC to pixel pitch mazimizes DoF.  It does no good to make a CoC smaller than can be resolved.

Howevver, DoF does not have to be maximized.  There are some creative folks who favor making DoF work for them.

[howiesmith]

In the case of focusing at the hyperfocal distance given by such a calculation, it means that everything in the far distance is as sharp as it can get with that sensor, whereas focusing at the traditional hyperfocal distance based on traditional CoC of about 1/1000th of frame diagonal leads to the images of far distant subjects being detectably less sharp than is possible with a sensor of over about 4 or 5MP.
[a href=\"index.php?act=findpost&pid=101121\"][{POST_SNAPBACK}][/a]

"Detectably less sharp" perhaps.  But is that "unacceptably less sharp?"  I think most definitions of DoF use "acceptably in focus."

This is exactly the problrm caused by setting single values of CoC fo all conditions.  I think if you use the pixel pitch as the CoC, and then calclate the hyperfocal distance, the "digital" and "traditional" DoF calculations are the same.  Hence, no difference for film or digital.  Just the choice of CoC.

It makes no more sense to select a CoC smaller than can be resolved by film than it does to select a CoC smaller than than can be resolved by digital.  DoF is an optical phenomenon, not "sensor."

[howiesmith]

In the case of focusing at the hyperfocal distance given by such a calculation, it means that everything in the far distance is as sharp as it can get with that sensor, whereas focusing at the traditional hyperfocal distance based on traditional CoC of about 1/1000th of frame diagonal leads to the images of far distant subjects being detectably less sharp than is possible with a sensor of over about 4 or 5MP.
[a href=\"index.php?act=findpost&pid=101121\"][{POST_SNAPBACK}][/a]

Deleted by author.  A dupelicate post of the one above.

[howiesmith]

Actually, it makes no sense to use a CoC smaller than the lens/sensor system can resolve.  The overall resolution O is:

1/O = 1/L +1/S

where L and S are the resolutions of the lens and sensor, respectively.

So, a CoC of the pixel pirch would be the limit for maximum DoF if and only if the lens were perfect - and not even Canon's L lenses are that.

Then throw onto that the printing process and the human factors of viewing, and pixel pitch is starting to look too small.

[Jonathan Wienke]

So, a CoC of the pixel pirch would be the limit for maximum DoF if and only if the lens were perfect - and not even Canon's L lenses are that.

That's ridiculous. There are plenty of Canon lenses capable of resolving single-pixel detail with the 1Ds and other Canon DSLRs. If you really want to pick nits, use a CoC slightly larger(maybe 1.4x) than the pixel pitch to account for the blurring effect of the AA filter. But if you actually intend to use the resolution of the camera, a CoC value close to the pixel pitch is the most accurate way to predict what will appear in focus and out of focus when viewed at 100%on screen and in larger print sizes where the print PPI is <300 or so.

[howiesmith]

If you really want to pick nits, use a CoC slightly larger(maybe 1.4x) than the pixel pitch to account for the blurring effect of the AA filter.
[a href=\"index.php?act=findpost&pid=101404\"][{POST_SNAPBACK}][/a]

40% doesn't seem like a nit to me.  And unless I have missed something along the line (probably have), the AA is part of the camera.

[BJL]

Actually, it makes no sense to use a CoC smaller than the lens/sensor system can resolve.  The overall resolution O is:
1/O = 1/L +1/S
where L and S are the resolutions of the lens and sensor, respectively.
[a href=\"index.php?act=findpost&pid=101394\"][{POST_SNAPBACK}][/a]

I think you make the "brick wall" mistake, of assuming that as son as the CoC is smaller than the resolution length scale 1/O = 1/L + 1/S, it has no effect on overall resolution. This is a version of the false belief that the resolution limit set by lens and sensor is the minimum of L and S.

It seems to me instead that no matter how small the CoC, it can smear some light that "should" fall near the edge of one photo-site over onto an adjoining photo-site, and so to some extent reducing resolution. (Likewise for even the very small diffraction spot size of very large apertures.)

That 1/L + 1/S is an empirical approximation, with 1/L and 1/S both in units of length, so says that the size scale of the smallest resolvable feature is roughly the sum of the size scales due to each factor, lens and sensor. That is, angular resolutions combine additively.

This might generalize to include OOF effects at a particular location in the image as something like

    overall resolution length scale = Ab + Pix + CoC + Dif

where CoC is the diameter of the circle of confusion at that location in the image, Dif is the diffraction spot size, Pix is pixel size and Ab is the resolution size limit set by lens aberrations.

This indicates that overall resolution is improved somewhat by reducing any of the length scales (reducing pixel size, CoC size or diffraction spot size, or increasing lens resolution) even one that is already smaller (better) than the others. Of course the greatest benefit comes from improving the worst factor.

[BJL]

"Detectably less sharp" perhaps.  But is that "unacceptably less sharp?"  I think most definitions of DoF use "acceptably in focus."
[a href=\"index.php?act=findpost&pid=101378\"][{POST_SNAPBACK}][/a]

I was not talking about definitions of DoF. I was responding to a question about "what setting the COC to pixel pitch does for you in actual practice", and neither the post I was responding to or my response ever mentioned "depth of field".

Sometimes people are interested in limiting OOF effects to less than that allowed by "most definitions of DOF", because those definitions refer to certain viewing conditions that differ significantly from the way that large prints of high resolution images are sometimes viewed.

[howiesmith]

I was not talking about definitions of DoF.
[a href=\"index.php?act=findpost&pid=101435\"][{POST_SNAPBACK}][/a]

Didn't say you were.

[BJL]

Didn't say you were.
[a href=\"index.php?act=findpost&pid=101474\"][{POST_SNAPBACK}][/a]

So you agree that your comment was completely irrelevant to my post, and does not add to or correct what I said? I wonder what the point was then.

[howiesmith]

So you agree that your comment was completely irrelevant to my post, and does not add to or correct what I said? I wonder what the point was then.
[a href=\"index.php?act=findpost&pid=101546\"][{POST_SNAPBACK}][/a]

No.  Nor would I ever suggest you were in error.  Take a couple deep nreathes.    I think there was a question in there.  I mY have gotten an answer in your replies, if I cared anymore.

[BJL]

I think there was a question in there.
[a href=\"index.php?act=findpost&pid=101564\"][{POST_SNAPBACK}][/a]

I think I understand now: I answered a question, and you responded with comments and question (which seemed rhetorical) on the slightly related but rather different topic of DOF and traditional definitions thereof. Fair enough I suppose, but an essentially different topic that we have discussed often, so I do not have anything new and worthwhile to say.


Perhaps the Luminous Landscape needs a forum devoted entirely to discussions of DOF.

[Craig Arnold]

Perhaps the Luminous Landscape needs a forum devoted entirely to discussions of DOF.
[a href=\"index.php?act=findpost&pid=101720\"][{POST_SNAPBACK}][/a]

What a brilliant idea!