Re: [OM] Binary DoF

[Chuck Norcutt]

Rockwell's example <http://www.kenrockwell.com/tech/focus.htm> using the 
Leica 40/2 shows how to set the hyperfocal distance on a prime lens 
using the depth of field scale.  He sets the infinity mark above the 
aperture he wants to use (f/8) at the right side of the scale.  He then 
reads 10 feet above the f/8 mark on the left side of the scale.  That's 
telling you that your depth of field ranges from 10 feet to infinity. 
The definition of hyperfocal distance is that focus distance that gives 
focus from half of the hyperfocal distance to infinity.  Now you can say 
you understand if you've inferred that the hyperfocal distance at work 
here is 20 feet (double the closest distance). If you look at the Leica 
focusing scale again look at the triangle that marks the actual focus 
point on the scale. You have to do a little interpolation since the 
scale is not linear but I think you'll agree that the actual focus 
distance is 20 feet.
Part of the argument that's been going on here purports that the 
parameters on which this Leica scale is based are old and inappropriate 
for more modern films and digital gear.  I can tell you that this Leica 
scale assumes a resolution of approximately 30 lines/mm.  That's often 
sufficient for me.  Want a more stringent parameter?  Shoot at one stop 
smaller (f/11) while continuing to assume the 10 feet to infinity given 
by f/8 on the scale.  That ups the assumed resolution value to 40 
lines/mm.  Still not good enough.  Up it another stop to f/16 while 
still assuming the limits of the f/8 scale.  That assumes about 57 
lines/mm.  But be careful now.  As you increase the f/stop Yyu're 
starting to head into diffraction territory which will destroy the 
resolution you think you're going to get.  Also, unless you're on a rock 
steady tripod your gear is never going to give you 57 lines/mm even 
though the lens be perfect.
Not yet discussed is that some labs limit print resolution to 250 dpi. 
If you plan the perfect parameters for a 300 dpi 10x15 print and your 
lab only prints to 250 dpi you're fooling yourself.  No doubt Moose 
needs 400 dpi.  I'm not sure how it's being done.
Chuck Norcutt

On 10/23/2014 12:08 PM, Sandy Harris wrote:
> Ken Norton <ken@xxxxxxxxxxx> wrote:
>
> > The comment was made that DoF and Hyperfocal calculations are based on
> > a binary decision: It's either sharp (in focus) or it isn't. I highly
> > disagree with that assessment. ...
> >
> > The history behind the DoF calculations (and lens markings) is based
> > on the viewing of a standard size print observed from a standard
> > viewing distance. It's not a binary thing, but extremely variable ...
> I know he is far from a popular man on this list, but Ken Rockwell
> has a page on this topic that seems to me (far from expert) to
> make sense: http://www.kenrockwell.com/tech/focus.htm
>
> Anyone care to critique that? Or point to a better resource?
>
> A method I have used with a reasonably fast prime (f2.0 or
> better) is to just pick something at a good distance (infinity
> focus), open the lens as wide as it will go, shorten focus
> until I can see blur, lengthen it a smidgen to eliminate the
> blur, stop down to f5.6 and shoot. I'm reasonably happy
> with results, but wonder if there is a better way.
>
> Also, that method won't work for a slower lens. Does
> anyone have a good suggestion for those?
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/