Olympus-OM
[Top] [All Lists]

Re: [OM] Re: "Depth of Field being too deep on the 4/3"

Subject: Re: [OM] Re: "Depth of Field being too deep on the 4/3"
From: Thomas Bryhn <thomas.bryhn@xxxxxxxxxx>
Date: Tue, 04 Mar 2003 13:32:00 +0100
At 03:34 04.03.2003, Jan Steinman wrote:
>From: Thomas Bryhn <thomas.bryhn@xxxxxxxxxx>
>
>If you play with one of the DoF calculators and compare DoF for different
>formats you'll discover that for any angle of view DoF only depends on
>absolute aperture.

AND reproduction ratio. That's where people get tripped up. A dime that is life-sized on a 4/3 sensor will have the same DOF as a dime that is life-sized on 8"x10" large format.

Personally I would rather say "OR reproduction ratio", because reproduction ratio makes more sense when you're changing lenses than when you're changing sensor format. For instance, photographing a dime 1:1 with 8x10" sounds like a waste of film to me, but I wouldn't hesitate to do it in 35mm. I would rather compare two pictures with the same angle of view, taken on different film or sensor formats, i.e. you change format *and* lens (and therefor r.r.), distance and framing stays the same.

>f/0.5 is the optical limit for lenses with air on both sides, and
>that would require an infinitly large opening...

I'm not aware of any such limitation. The front element diameter approaches infinity as the focal ratio approaches zero, but that's a long way from 0.5.

Normally we use the approximation f/# = f/D, more correct would be f/# = n1/(n2*2*sin(t)), where n1 and n2 are indexes of refraction and t is the angle spanned by the edge of the front element and the optical axis, seen from the point where the optical axis intersects the film. n1=n2 for our applications with air on both sides of the lens. Sin(t)=tan(t)*cos(t), and for small t you can safely write sin(t)=tan(t) because cos(t)=1. 2*tan(t) is nothing more than D/f, and inverting this gives us our normal approximation.

Now assume a *very* large front element where the normal approximation is no longer valid. Sin(t) will then approach 1 and we're left with another approximation: f/# = 1/2 * n1/n2

Again n1=n2 for normal photography, so 0.5 is the limit.

Note: In microscopy special oils with high indexes of refraction (n2) are used to push this limit, do a search for "oil immersion".

Thomas Bryhn


< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >


<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz