Header banner

<< Previous Thread DoF tables Next Thread >>

Subject: DoF tables
Date: 2006-11-10 04:06:10
From: Don Lopp
Thanks to both John Thurston and David Kesner, my clumsy question,
concerning, what the distance numbers pertained to, was clearly answered.

I do not understand why the distance measurements, suggested, were
listed so precisely, such as focus at 31.4 ft,allowing a near point at
15.7 feet, when shooting at f/11. I do not know of any MF camera, or
any color slide film that would be capable of producing such precision,

I agree with John T. when he mentioned having doubts as to using f/stops
such as f/2.8 or f/4.0, as I would not, intentionally, produce, 'On Film
Deviations', of 0.4mm, or less.

For many, the DoF tables may be OK. For my MF stereo slides, I would
close down the lens by one f/stop. If shooting at f/16, I would use
the f/11 DoF table. I prefer my MF stereo images to appear
reasonably sharp. The latest DoF table is, apparently satisfied to
use a, 'theoretical', on film resolution of about 15 lp/mm, which is
less than what I am satisfied with. I prefer to have an, 'actual',
on film, resolution of 20+ lp/mm.

In the next MF folio, I will enclose a slide which will show an example.

I do consider the new MF camera to be a real boon to the future of MF
stereo photography. I am jealous of those that will be able to use a
camera with an accurate focus and an accurate viewfinder

I believe that Sam Smith should be given much credit, for getting the MF
camera off of the drawing boards.


Best regards,

DON
Subject: Re: DoF tables
Date: 2006-11-10 11:47:03
From: Michael K. Davis
Don,

At 03:58 AM 11/10/2006, you wrote:

>I do not understand why the distance measurements, suggested, were
>listed so precisely, such as focus at 31.4 ft,allowing a near point at
>15.7 feet, when shooting at f/11. I do not know of any MF camera, or
>any color slide film that would be capable of producing such precision,

Here is the table reproduced with less precision. Again, it's preceded by
the original paragraph explaining its purpose:

Assuming a desired resolution of 5 lp/mm as seen in a 75mm viewer that's
focused at Infinity, using a Maximum Permissible CoC Diameter of 0.059 mm
and a camera focal length of 80mm, here are the Near and Hyperfocal
distances (in feet) at various f/stops, for scenes that include Infinity
subjects:

Stop____Near____Focus____Far

f/2.8____60____130____Infi
f/4______40____90____Infi
f/5.6____30____60____Infi
f/8______20____45____Infi
f/11_____16____30____Infi
f/16_____11____22____Infi
f/22______8____16____Infi


>I agree with John T. when he mentioned having doubts as to using f/stops
>such as f/2.8 or f/4.0, as I would not, intentionally, produce, 'On Film
>Deviations', of 0.4mm, or less.

I included the range of apertures shown in the table only because those are
the apertures available on the TL-120's lenses. I, too, would find no
appeal in producing stereo images of subject spaces that require only the
DoF yielded at the wider apertures.

Another formula - the DiMarzio-Davis Equation for % of MAOFD - can be used
to calculate the On-Film Deviation (expressed as a percentage of MAOFD)
that will be delivered at a specified f-Number using a fixed-base stereo
camera, when that f-Number provides nominal DoF for the subject space:

% of MAOFD = B / (Fc * Di * Fv / (6000 * Ca * N * (Fv + Di)))

Where B is the stereo base (lens separation) of your stereo camera, Fc is
the camera focal length, Di is the virtual image distance at which the
viewer is focused (not the physical lens-to-film distance), Fv is the
viewer focal length, N is the f/number indicated on your Depth of Field
scale or table, and Ca is the permissible Circle of Confusion diameter that
was used to calculate DoF.

I've produced the following tables of On-Film Deviation (expressed as a
percentage of MAOFD) vs. f-Number, for various Maximum Permissible CoC
diameters one might have deployed for DoF calculations. These tables
assume you are using 80mm taking lenses with a fixed 65mm stereo base and a
viewer with 75mm lenses with focus fixed at Infinity.

CoC______%M@f/2.8__%M@f/4__%M@f/5.6__%M@f/8__%M@f/11__%M@f/16__%M@F22

0.148mm_____27.2______38.5______54.4_______77.0______108.8_____153.9_____217.7
0.098mm_____18.0______25.5______36.1_______51.0_______72.0_____101.9_____144.2
0.074mm_____13.6______19.2______27.2_______38.5_______54.4______77.0_____108.9
0.059mm_____10.9______15.3______21.7_______30.7_______43.4______61.4______86.8
0.049mm______9.0______12.7______18.0_______25.5_______36.0______51.0______72.1
0.042mm______7.7______10.9______15.5_______21.8_______30.9______43.7______61.8
0.037mm______6.8_______9.6______13.6_______19.2_______27.2______38.5______54.4

If the depth of the subject space requires an f-Number where the resulting
% of MAOFD would exceed 100%, as shown in the table below, you'll run the
risk of window violations (if the Fars are mounted at Infinity) -or- the
risk of forcing the eyes to diverge (if the Nears are mounted to the
window). Subject spaces having a depth that yields less than 30% (a
subjective threshold) will appear "flat" in the viewer.

Where did I get the CoC values shown in the first column? From my earlier
post, where I had calculated the Maximum Permissible CoC Diameters required
to achieve desired resolutions, as seen in a viewer that imposes a 3.39x
magnification (a 75mm viewer focused at Infinity):

2 lp/mm - 0.148 mm
3 lp/mm - 0.098 mm
4 lp/mm - 0.074 mm
5 lp/mm - 0.059 mm
6 lp/mm - 0.049 mm
7 lp/mm - 0.042 mm
8 lp/mm - 0.037 mm

So, if you wanted your DoF calculations to support a resolution of 8 lp/mm
as seen in a 3.39x viewer, use a Maximum Permissible CoC Diameter of
0.037mm to do your DoF calculations, then refer to the 0.037mm table of
OFD's vs. f-Numbers.


>For many, the DoF tables may be OK. For my MF stereo slides, I would
>close down the lens by one f/stop. If shooting at f/16, I would use
>the f/11 DoF table. I prefer my MF stereo images to appear
>reasonably sharp. The latest DoF table is, apparently satisfied to
>use a, 'theoretical', on film resolution of about 15 lp/mm, which is
>less than what I am satisfied with. I prefer to have an, 'actual',
>on film, resolution of 20+ lp/mm.

The "latest DoF table" you're referring to, was an arbitrary example. The
0.059mm CoC diameter used to calculate the table will support a resolution
of 5 lp/mm after 3.39x magnification, but I included the aforementioned
table of CoC's to encourage people to decide for themselves how small they
want their CoC's to be. From what you've written, above, I suspect you
would be happy with a CoC that supports no less 7 lp/mm after 3.39x
magnification. As stated in an earlier post, I use a CoC of 0.03115 to
support a resolution of 7.625 lp/mm after 4.21x magnification in my
adjustable-focus 78mm SaturnSlide viewer. Your venerable one-of-a-kind
viewer, with it's multi-element lenses, is capable of even higher
magnification than the SaturnSlide - all the more reason you would be
unsatisfied with larger CoC's.

Mike Davis