# Jumping ships - DoF

Started Sep 29, 2012 | Discussions thread
 Like? 1
 Re: Disk of Confusion diameter is related to ratio of Focal Length divided by F-Number In reply to Detail Man, Oct 5, 2012

Detail Man wrote:

Detail Man wrote:

Sf = ( L / F ) x ( ( Df / D ) - 1 )

The form of the identity for Sf in object-space (directly above) is equivalent to your identity above:

b = ((f * ms) / N ) * xd / (s + xd)

... except - the identity for Sf (in "object-space") is not scaled by a magnification-factor ("ms", above), as is your identity for b which appers to be referenced to "image-space".

The Background Distance divided by the Plane of Focus Distance ( Df / D ) remains constant, therefore ( ( Df / D ) - 1 ) remains constant. The diameter of the Disk of Confusion (in "object-space", and as referred back to "image-space") would change by the ratio of the Focal Length divided by the F-Number ( L / F ) - just as is the case in your (stated to be approximate) identity:

b = ((f * ms) / N )

If my thinking is correct, then I agree with your assessment, and it seems precise (not approximate) ...

... except, I think that it's necessary to scale the ratio of Focal length divided by F-Number as stated:

Sf = ( L / F ) x ( ( Df / D ) - 1 )

... and to then multiply the derived ("object-space") Disk of Confusion diameter by the Image Magnification (the ratio of Sensor Size divided by Object Size) in order to refer it back to the image-sensor (in "image-space") if you want to relate blur-diameter to the image-sensor frame dimension(s).

... which is you ask me is exactly formula I used, as b is in in image space and ms which is the Image Magnification.

Anyway ... thanks a lot for your thoughts.

BR
Carsten

Complain
Post ()
Keyboard shortcuts: