On 4/6/2010 12:53 PM, Ken Norton wrote:
> OK, so here is my test:
>
> 1. Mount lens to E-1 via genuine OM-43 adapter
> 2. Set focus to infinity
> 3. Set camera to manual exposure
> 4. Set ISO to 100
> 5. Set aperture to F16
> 6. Set shutter speed to 1/4 second.
> 7. Hold lens up flat against LCD computer monitor with white background
> 8. Take picture.
> 9. Open up one stop
> 10. Adjust shutter-speed down one stop
> 11. Take picture
> 12. Repeat 9-11 until done.
>
Is the combination of #1 & #7 valid? I'm not saying it isn't, just
asking? Is it the same using a lens in focus vs. WAY out of focus?
Is it possible that, as the shutter speed gets higher, the sync
frequency of the screen starts to affect the exposure? I was concerned
with both possibilities when testing using a small, daylight fluorescent
light table.
> Comparing histograms and general appearance (this isn't my first dance with a
> camera), I'd say that all images are within 1/8 of a stop of each
> other--which is what I expect to see from the E-1 as there is a little bit
> of exposure slop with it. (10 identical pictures will have a couple exposure
> variations). However, the final one at F1.4 is underexposed by 2/3 of a stop,
> which is exactly what I've been seeing all along with the exposure
> meters.
>
> Carefully examining the lens, the aperture blades at F1.4 are fully retracted
> and behind the round shroud just in front of the blades. So there is no
> evidence whatsoever that the blades are getting hung up at all. None.
>
> So, what am I to make of this?
I'm not entirely sure. As I said and showed with images, I've had
similar results with Canon DSLRs. On the other hand, the diaphragm
definitely occludes part of the apparent aperture when moved from
f1.4-2.0, so there sorta, kinda hasta be less total light coming through
the lens.
Looking through from the front with a transparent ruler, I took rough
visible diameter measurements. It's hard to do accurately and the edge
of the aperture is uneven at f2, sut roughly, I came up with an area at
f2 of 70% of wide open, where it should be 50%. It's a very rough
measurement and in the worng direction. Looking from the back, I can't
find a way to measure. BTW an old (281,xxx) 50/1.4 looks the same.
The mystery, then, as I see it, is:
1. Why has Oly, apparently from the beginning, had a smaller aperture
difference between marked f14. and marked f2 on their 50/1.4 lenses?
2. Why, even when there is still significant mechanical stopping down of
the aperture, is there less apparent change in light on at least some DSLRs?
I know the Canons aren't of much interest to you. I mention them because
the highlight how the problem may be in part something to do with
digital sensors. When I know the aperture is closed down part way, but
get the same actual digital exposure with the same f-stop at either
physical aperture, it raises a question about how that can be.
I have no answer, except to guess that DSLRs may not be good measuring
tools for this issue, for as yet unknown reasons, at least to me.
--------------------
I still differ with your assumption that the smaller aperture opening
change means the lens is not f1.4. That conclusion can't be logically
drawn from the purely relative data.
The 1984 Modern Photo test of 50/1.4 #1,136,552 shows a tested aperture
of f1.44. As the tested focal length is slightly long at 50.94mm, the
calculated aperture if it were exactly a 50mm lens is 1.4134, almost
exactly the theoretical value of 1.414. I tend to go with MPhoto, as
they were in the forefront of lens testing at that time.
If we believe the optical bench test - that Oly is not cheating on
maximum aperture - the question becomes why they "fudged" the step down
to the next aperture. I can imagine all sorts of possible reasons to do
with the practical difference between thin lens theory and real world
performance on FF film. It may be important to remember that f-stop is
simply a ratio - of focal length divided by aperture diameter. What you
are actually trying to measure is T(ransmission)-stops.
Ready to do a test on film? Got a densitometer?
Remembering when the original 50/1.4 was designed, practical even
exposures with aperture ring changes had been much more important than
theory for the whole history of photography. TTL metering was a very new
thing, while the engineers had been designing lenses for some time.
Moose
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|