Referring to earlier discussion if Canon f3.5 and f4.5 are half stops
or not (even though as *values* they are not):
It seems they *are* half stops, when EOS-3 is set to display half-stops.
Actually based on my measurements I'm sure of it.
If EOS-3 is set to show 1/3-stops then 3.5 and 4.5 naturally are
one-third steps.
I also tested my 28-135@28mm (f3.5) and 100-400@100mm (f4.5) to see
if the max aperture is half stops or something else. See down below.
I made the following test system:
- set EOS-3 in "bulb" mode with remote release attached.
- opened camera back and tweaked the filmback lock to get "bulb" work
(normally EOS-3 uses some fast shutter speed if back is open).
- attached the EF 35/2.0 lens without front cap.
- put the camera on my small light table, lens pointing down.
- while back open I put piece of white copier paper on the shutter
opening (where film usually is located), just behind (actually above)
the shutter blades.
Next I took my flashmeter (Gossen Sixtomat Flash) which can also be
used to measure available light (reflective or incident). It displays
the values in 1/10th of steps.
I removed the spehere in front of it to use it as a reflective light
meter.
Finally I:
- set EOS-3 in 1/3 stop selection mode (was it CF5=0 ?).
- set the lens aperture at certain value (e.g. 2.8).
- pressed the remote trigger shutter release and locked it (so that
shutter remains open).
- while holding the light meter close to the white piece of paper
so that the sensor points in the middle of the shutter area AND
protecting it from environmental light (which was dim but still
significant) I read the light meter value and wrote it down.
So I was measuring the lightness of the backside of the paper
while the light was coming from other side.
- I set the lens aperture at 2.8, 3.5, 3.8, 4, 4.5, 5 and 5.6 and
took 10 measurements at each f-stop.
This showed quite accurately that the aperture was jumping in 1/3
stops.
Then I repeated the same for half stops (CF5=2 I think) 2.8, 3.5,
4, 4.5 and 5.6 and again it was quite apparent that the aperture
jumped with half stops.
In fact, I first tried a similar test without the piece of paper and
with the sphere on in my light meter (as incident light meter) but
the measurements were not linear. E.g. whan I changed the aperture
1 or 2 stops the measured values were changing actually more or less
than 1 or 2 stops, and error was increasing the more I changed the
aperture.
I quess the reason why the light meter could not measure reliably
in this case was that the light was coming from one direction only.
I'm not sure though if it is a problem in the meter design or if
this should be expected? After all, when the sphere is on, the
sensor is probably still measuring only the front portion of the
sphere.
Finally,
I wanted to see how do some of my lenses behave at full aperture.
In most cases I used apertures 'full open', -0.5, -1 and -1.5 stops.
to see what happens. It looks that while 100/2 and 100-400@100mm(f4.5)
full open are really that fast my EF35/2 and 28-135@28mm(f3.5) aren't
quite as fast as promised. It was less than -0.5 stop error, maybe
around -0.25 stops of promised maximum aperture (I don't remember
the 35/2 results exactly).
Because of vignetting (or is it called "light fall-of") I tried to
measure from the middle of the film area. Actually in this test I
moved the meter slightly - while measuring continuously - trying to
find the position for highest values.
Any comments in method or results?
In my opinion everything was as expected.
regards Vesa
_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.
*
****
*******
***********************************************************
* For list instructions, including unsubscribe, see:
* http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************
