Archived posting to the Leica Users Group, 2004/07/07

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: [Leica] "focus then reframe" (aka fungus in camera)
From: mark at rabinergroup.com (Mark Rabiner)
Date: Wed Jul 7 22:37:41 2004

On 7/7/04 6:14 PM, "Henning Wulff" <henningw@archiphoto.com> typed:

> At 11:25 PM +0200 7/7/04, Jean Louchet wrote:
>> Hi all,
>> 
>> Beware, the calculation maniac is back on line :-)
>> 
>>  Apart the Descartes formula, the second important formula is the "Setala
>> formula" which allows to calculate the depth of field, the diameter of the
>> blur circle etc. (According to the "photo club de Bievres", Mr Setala was
>> a Leica User from Finland, and the inventor of the "depth of focus" scales
>> one now finds on almost all lenses)
>> 
>> 
>>  Its parameters are:
>>    the diameter of the blur circle: Delta
>>    the diameter of the diaphragm:   Diaph
>>    the focal distance: f
>>    the distance on which the lens is focussed: z
>>    the actual object's distance: d
>> 
>> Setala's formula is:
>>    Delta = Diaph * f * abs(1/z - 1/d)
>> 
>> It is only true when z is much greater than f (always the case except in
>> macrophotography)
>> 
>> Let's take as an example a 35mm lens on a leica. As the focal length f
>> (35mm) is about equal to the width of the image (36mm), if we want to
>> photograph a vertical wall at distance d, the part of the wall which
>> projects onto the side of the picture will be at an actual distance z
>> (measured by the rangefinder before turning the camera) such that
>> 
>>   d^2 = z^2 + e^2 where e is half the width of the part of the wall that
>> will be visible on the picture (sorry I can't make a drawing here, but
>> this is Pythagore's theorem), and e=z/2 (because we have a 35mm lens and a
>> 18mm half-width of the picture).  Thus we can calculate 1/z - 1/d,
>> eliminating d (d = z * (sqrt(5)/2 = z*1.118) gives
>>  1/z - 1/d = 1/z *(1 - 2/sqrt(5)) = 0.105/z
>> 
>> At aperture f/2 we have (in millimeters)
>>  Diaph = 17.5mm
>>  and
>>  Delta = 17.5 * 35 * 0.105 * (1/z)  (z also in millimeters!)
>>  Delta = 64.3/z
>> 
>> This is the diameter of the "blurring" on the picture of the wall if we
>> rely on the "focus then reframe" technique.
>> 
>>  At infinity (z = inf) the picture will be (in theory) perfectly sharp.
>>  At z = 10 metres we have Delta = 64.3/10000 = 6.4 microns = perfectly
>> sharp in practice.
>>  At z = 5 metres, Delta = 13 microns = very sharp (would still resolve 80
>> lines/mm).
>>  At z = 2 metres, Delta = 32 microns = usually considered "good" or
>> "acceptable" (30 lines/mm).
>>  At z - 1 metre, Delta = 64 microns :-( it is worth to make the focussing
>> correction.
>> 
>> The correction coefficient on the edges is 1.118 (say 12%). If one uses
>> the rangefinder on an object and wants this object to be sharp when it
>> gets on the side of the picture, he will have to SUBTRACT 12% to the
>> distance given by the rangefinder.
>> 
>> The same formulas with aperture 1.4 would give:
>>  At z = 10 metres Delta = 9 microns "very sharp"
>>  At z = 5 metres Delta = 18 microns "sharp"
>>  At z = 2 metres Delta = 45 microns "fair" :-|
>> 
>> TO SUMMARISE:
>>  with a 35mm, at aperture 2 "focus then reframe" is OK from 2 metres
>>               at aperture 1.4 "FTR" is OK from 3 metres.
>> 
>> At shorter distances, one has to reduce the focussing by, say, 10 - 12%.
>> 
>> I did not consider focussing on an object which will finally be right in a
>> corner (no practical interest).
>> 
>> OF COURSE we luggers will take much, much better pictures now :-)
>> 
>> If anyone is interested in having the equivalent results with a 50mm, let
>> me know (private e-mail), I may well open another page on my web site.
>> 
>> Jean
> 
> While these calculations and formulae are correct, the point is still
> moot due to the almost universal field curvature, especially of fast
> lenses, at shorter distances.
> 
> After these calculations, you won't be closer to the truth or focus,
> _and_ your subject will be gone to sleep or just plain gone.


This might as well be computer code.

IF you were a tree would you know you were in a forest?


Mark Rabiner
Photography
Portland Oregon
http://rabinergroup.com/




In reply to: Message from henningw at archiphoto.com (Henning Wulff) ([Leica] "focus then reframe" (aka fungus in camera))