Archived posting to the Leica Users Group, 2005/06/28
[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]At 1:16 PM -0700 6/28/05, Adam Bridge wrote:
>I don't understand this, Henning.
>
>It seems to me that the depth of field is solely a fuction of the
>geometry of the lens/film system. That the print and enlargement has
>nothing to do with it.
>
>If I took an R8 with a 50 shot wide open at f1.4 I'd get some depth of
>field. I replace the back with the DMR. Same shot. Same depth of field
>but cropped.
>
>So then I open the images on photoshop and look at them. How does the
>depth of field change?
>
>So I'm confused but I could certainly be enlightened! :)
>
>Adam
Adam,
Imagine an image formed by a given 50mm lens at f/8 on some
sensor/film. By definition the highest resolution with a well
corrected lens occurs at only one specific focus distance. Now we
want to have not just one plane in focus, but some depth. Therefore
we allow a point to be imaged as a smudge, allowing it to be a
certain size, say 1/10mm. If that is imaged onto a sensor that is 1
inch long, then we have to magnify it by 10x to produce a 10in long
picture. That will result in the point that is out of focus, but
within the depth of focus to be printed to be 1mm in diameter. If
this lens is now imaged onto a sensor that is 1/2in long, it has to
be magnified by 20x to produce a 10in long print, so that the same
point is now 2mm in diameter.
Obviously, if our quality criterium states that a point can be
represented by a smudge up to 1mm but no more on the final print,
then the second print is unacceptable. Therefore, we can accept no
more than 1/20mm diameter smudge on the sensor plane for the image
formed by the 50mm lens at f/8 with the 1/2in long sensor. Our Circle
of Confusion criterium at the film plane for the 1/2in long sensor is
therefore only 1/20mm, but 1/10mm for the 1in long sensor, to give
the same effective circle of confusion at our final result, the 10in
print.
If you intend to stand the same distance from the final print, and
want it to look just as sharp, and it is 20in long, your CoC criteria
at the sensor plane can only be 1/2 as large again, so 1/20mm for the
1in sensor and 1/40mm for the 1/2in sensor. Normally we don't stand
as close to larger prints, but if we do detailed landscapes where we
feel we might want to look closely at fine detail, our CoC criteria
have to be more severe, so we stop down more, or back up, or apply
Scheimpflug's ideas or whatever.
If you are talking about the same angle of view, as opposed to the
same lens and f/stop combination, then you only need a 25mm lens to
give the same angle of view for the 1/2in sensor, and your same
f-stop will give you _more_ effective depth of field.
So, overall for the same angle of view you get more depth of field in
the end, for the same print/image size, but not as much as you get by
using the shorter lens on the larger film area; ie, the 25mm lens on
a 35mm film camera will give you more depth of field when the final
print criteria are considered than when the 25mm lens is used on a
piece of film 18x12mm.
You can check this by looking at the depth of field scales on MF
camera lenses compared with 35mm lenses. The depth of field indicated
at f/8 on a 100mm/3.5 Planar for Hasselblad is much greater than the
depth of field indicated at the samer aperture on a 100/2.8 Canon
lens, and it's not because Hasselblad is less concerned about quality
than Canon. :-)
>On 6/27/05, Henning Wulff <henningw@archiphoto.com> wrote:
>> Since the image of the 50 on APS has to be enlarged more, the final
>> CoC on the print will be larger, so that, conversely, means that the
>> DOF of a 50 at any given f-stop will be less for a same size print
>> than the DOF on a full size sensor/film.
>>
>> --
> > * Henning J. Wulff
--
* Henning J. Wulff
/|\ Wulff Photography & Design
/###\ mailto:henningw@archiphoto.com
|[ ]| http://www.archiphoto.com