Re: Stereo Macro

[P3D John Bercovitz <bercov@xxxxxxxxxx>]

Sam replies:
 
> Let's put it another way:  Is there any advantage of medium format
> macro over realist format macro? I can see reduced grain as an
> advantage, but will the image be any sharper at a small aperture like 
> f:45 or f:60?
 
Prologue
OK, I'm clear on the concept now; I got off on kind of a tangent with 
that strict 1:1 stuff.  To take reduced grain as much out of the picture 
as reasonable, I'll say the film + lens resolution is 60 lines per mm 
at the film.  (Grain still has to figure in because it's one of the 
components of fuzziness.)  I'm going to stay away from formulas but I'm 
going to give you some numbers to tell you what I'm comparing.  I don't 
want anyone thinking I'm comparing apples and oranges when really I'm 
comparing apples and pears: If I don't say exactly what I'm comparing 
when I say one thing is better than another, it doesn't mean much.
 
Medium format conditions
First, let's arbitrarily chose 1:1 magnification for medium format.  
Medium format has a diagonal of 80 mm and hence a normal focal length 
of 80 mm.  To get a focal length of 80 mm in a 1:1 situation, we need a 
40 mm lens.  (If we used an 80 mm lens, we'd be operating at 160 mm and 
we'd get _horrendous_ squash when we viewed the result.)  Let's say the 
object (the subject of the composition) is 25 mm high.  Since we are
photographing at 1:1, it will also be 25 mm high on the film.  When we 
view the image, it will subtend an angle at the eye of 17.76 degrees.
 
Miniature format conditions
Now we move on to miniature format.  Let's say a normal focal length is 
40 mm.  To subtend 17.76 degrees, our 25 mm high object now has to make 
an image on the film which is only 12.5 mm high.  So now the magnifi-
cation is 0.5, not 1, as it was in the medium format case.  To get a 
magnification of 0.5 and an operating focal length of 40 mm, we need 
40 mm on the film side of the lens and 80 mm on the subject side of the 
lens.  This means we need a lens of focal length 26.7 mm.  (You do have 
one of those on hand, don't you? 8-)
 
Qualitative results
So now I go to my spreadsheet which considers geometric optics, physical 
optics, and film/lens resolution...  Oh my.  This is interesting.  I set 
the f/number to get the same geometric resolution (at the eye) in both 
cases.  The result of that was that not only was the resolution due to 
geometry the same in both formats, the resolution due to diffraction 
(again at the eye) was also exactly alike in both formats for a given 
depth of field.  So the sole difference is the film resolution.  And 
that only matters because in the medium format case, we view the film 
from a greater distance.  Wow, huh?
 
Quantitative results
     set geometric res = 1.5 MOA
     give film/lens credit for ability to do 60 lines per mm
 
Miniature format (26.7 mm lens, subject dist = 80)
 
near     far      maximum    lens      eff.     summation
dist     dist     st base    f/no      f/no   res, all causes
 
70       90         10        97       147       7.5
72       88         13        77       117       6.1
74       86         17        58        87       4.9
76       84         26        38        58       3.8
78       82         53        19        29       3.0
79       81        106        10        14       2.7
 
 
Medium format (40 mm lens, subject dist = 80)
 
near     far      maximum    lens      eff.     summation
dist     dist     st base    f/no      f/no   res, all causes
 
70       90         10       146       296       7.3
72       88         13       116       234       5.9
74       86         17        86       174       4.6
76       84         26        57       115       3.5
78       82         53        29        57       2.5
79       81        106        14        29       2.2
 
The reason the effective f/number is larger than the f/number written on 
the barrel of the lens is that you've got the lens focussed out so far.  
You have to factor that in when you figure the amount of light needed.
 
What would be a good resolution at the eye?  Well 1 minute of arc (MOA) 
would be nice but obviously we aren't going to get it by conventional 
photography.  Two or three or even four MOA would probably be OK as long 
as you don't have any sharp, high contrast edges to betray the lie.
 
I threw in the maximum stereo base.  A bit gratuitous for these purposes 
but that was what I originally wrote the spreadsheet for so it was easy 
to get.
 
John B
 
PS: I bet you "intuited" this result, huh Sam?  Sneaky guy.  8-)
 
PPS: I could have played with the spreadsheet to balance res loss from 
each cause in each case but that sounded like work.  As it is, these are 
balanced best at the 78/82 range for both formats.


------------------------------

End of PHOTO-3D Digest 1449
***************************
***************************
 Trouble? Send e-mail to 
 wier@xxxxxxxxxxxxxxx 
 To unsubscribe select one of the following,
 place it in the BODY of a message and send it to:
 listserv@xxxxxxxxxxxxxxx 
   unsubscribe photo-3d
   unsubscribe mc68hc11
   unsubscribe overland-trails
   unsubscribe icom
 ***************************