[Leica] Narrow DoF on the Leica E-3

[pklein at 2alpha.net (Peter Klein)]

Richard:  Thanks, this is what I expected.  If you shoot a gray card or an 
evenly-lit blank wall, out of focus, and convert the image to grayscale, 
it's a little easier to interpret.  I'm sure Photoshop has some sort of 
eyedropper tool that measures the pixel values--most editors do.

Anyway, to clear up the questions others asked. . .

If you shoot at several different f-stops, but adjust your shutter so all 
exposures are the same EV, then all pictures should be the same 
brightness.  If I shoot a gray card at 1/60 @ f/2 and again at 1/125 @ 
f/1.4, the resulting pictures should look the same.  And with my film OM-2 
and 50/1.4 Zuiko, that's what happens.  But put the same lens on the E-1, 
and the picture taken at f/1.4 is much darker.  I found that I had to set 
the shutter 2/3 of a stop slower than the expected value in order to get 
the same grayscale value as the f/2 picture (the E-1 can adjust the shutter 
in 1/3 stop increments).

f/2 and f/1.4 are one full stop apart.  But on the E-1, when you open up 
from f/2 to f/1.4, the sensor doesn't get a full stop more light.  It only 
gets 1/3 stop more.  The sensor is somehow not receiving all the light that 
the lens is transmitting.

I'm only detailing one test I did here.  But the results were confirmed by 
some practical available light photography.

So no, we're not talking about a lens that's only transmitting 1/3 stop 
more light when you open up the full stop from f/2 to f/1.4.  With film, 
the lens performs as it should.  But when that same lens is placed in front 
of a 4/3 sensor, something different happens.  My guess is that some of the 
rays get cut off by the pixel wells.  It must be a matter of angles, where 
the node is for this film lens vs. a digital-specific lens that is more 
telecentric.  The Leica 25/1.4 Summilux does not have this 
problem.  Somebody on dpreview tested it the other day, and it behaves as 
it should.

What it boils down to is that if you want a lens that truly delivers f/1.4 
on a 4/3 camera, you're not going to get it with an OM 50/1.4.  You have to 
buy the Leica lens, or the Sigma 30/1.4.  Or maybe another brand's "normal" 
f/1.4 lens

--Peter

At 07:15 PM 3/4/2008 -0800, Richard wrote:
>Hi Peter, is there a way for Photoshop or something to give the
>average EV value or whatever so I don't have to eyeball the results?
>I ended up using something like
>...
>1/2500 @ F2.8
>1/5000 @ F2.0
>1/8000 @ F1.4
>
>The shutter speed max out at 1/8000. So in theory, the F1.4 is
>overexposure by 25%? Eyeballing the resulting RAW ORF files, the 1.4
>is actually darker, so it may support your theory that the difference
>is less than the one stop.
>
>All the other files look similar, except that *may be* at F11, it is
>brighter by a tad, but may be it's the cloud moving away :-) Not very
>dramatic though. F16 is fine again.
>
>So not very scientific, but may support your thesis...