What makes you believe your statement that "an object at a certain distance beyond DOF will appear ... more blurred in a large-format shot"?[a href=\"index.php?act=findpost&pid=178944\"][{POST_SNAPBACK}][/a]
Images. Unfortunately, they're not mine, and they are not on the Internet, so no URL. They were shot with several different format digital cameras at short distance (approx. 1 - 2 ft or so). The differences in the DOF
and blur characteristics were obvious.
I contend that this statement is wrong, based on a mathematical formula derived from a geometric model of image blur.[a href=\"index.php?act=findpost&pid=178944\"][{POST_SNAPBACK}][/a]
After some more thinking about it, I don't insist in the words 'more blurred.' But I do insist in 'blurred in a different (smoother) way.'
Since you do not believe in formulas ...[a href=\"index.php?act=findpost&pid=178944\"][{POST_SNAPBACK}][/a]
Actually, I am an engineer, and I do believe in formulas as much as you do. However I am aware that formulas usually don't reflect the real world but the (often simplified and/or idealized) models that we make to describe, and to grasp, the world ... or part thereof. So formulas don't always tell the full truth---the COC formula discussed above, for example, does not take lens aberrations or diffraction effects into account. And currently I am wondering if it is accurate at all distances or only at distances much longer than focal length.
If that formula was accurate at all distances (which, to me, still is subject to further investigation) then it would suggest that 'equivalent f-stop' (with regard to DOF), for the larger frame format, is not simply small format's f-stop multiplied by form factor but somewhat greater (greater aperture number, that is, not wider aperture), depending on focus distance.
Be q the quotient of the linear sizes of the formats (also known as 'crop factor'), d the focus distance, f_small and f_large the equivalent focal lengths of the small and the large formats, and k_small the f-stop number for the smaller format, then the equivalent f-stop number of the larger format, k_large, according to that COC formula, was:
k_large = q * [(d - f_small) / (d - f_large)] * k_small(With d approaching infinity, the factor in brackets [] approaches 1, and thus it turns into the well-known formula
k_large = q * k_small.)
With this modified equivalent f-stop, the COC curves of both formats, according to the COC formula given further up this thread, will match exactly across all object distances. This suggests that the blur beyond the DOF range, at equivalent focal lengths and equivalent f-stops, was
exactly the same for all frame formats.
But as a matter of fact, the blur is not equal for all formats. So the COC formula obviously is over-simplified. I suspect it is meant for long focus distances but becomes increasingly inaccurate at shorter focus distances. Or maybe it does describe the
sizes of the circles of confusion accurately but not the characteristic, or appearance, of the blur. I don't know.
An interesting question ...
-- Olaf