> > Sorry, I checked 100-400@100mm (f4.5) only when EOS-3 was set for
> > half-stop scale (it was around 2 a.m. anyway when I made the tests).
> > I may do (just for "theoretical interest") it once more with
> > 1/3-scale selected and see how the jumps are between 5.6 - 5 - 4.5.
> >
> > IF f3.5 is f4 + half-stop it is also interesting to see if f3.5
> > at 1/2 and 1/3-scales is equal or not.
> >
>
>I think you already did a plenty good job, but yes, this would be
>particularly interesting.
>
>I don't think they are equal. If they were, what would f4 + 2/3
>stop be with 1/3 scales? ;-) What I can't figure out yet is the
>following: A nominal f/2 lens can be set to f/4.5 on both,
>1/2 and 1/3 scales. And I suppose it means 1/2 and 1/3 stop
>closed down from f/4 in these cases. But I can't imagine an f/4.5
>lens could not be stopped down to f/4.5 (=1/3 stop from f/4) with
>a 1/2 scale. It has to be possible somehow ;-) Otherwise it would
>be a waste of lens speed and the lens would be faster when used
>with an 1/3 scale than it would be with an 1/2 scale.
>Confusing ... :-(
Thanks Thomas (& possibly others if any?)
Last night (after visiting the Finnish Nature Photograps of the Year
happening which I unfortunately didn't win) I made some further tests.
As you see I don't have enough opportunities to take real photograps...
Here are the results and some comments as well.
Using the same method I checked several lenses. The syntax below is
"fx.y: a.b/c" where
- fx.y means the adjusted aperture
- a.b/c means the measured value where a.b is full stop and c is the
tenths of full stop.
So, as an example f4.5 2.0/5 means when the lens was adjusted at
aperture 4.5 the measured value (which has nothing to do with the
adjusted absolute value) was 2.0 plus 5/10 steps light which in
practice means aperture 2.5 (assuming 2.5 is 2 plus half-steps).
In theory the measured value for each lens should be the same for
the same aperture but because of the method there may be some error.
If someone has a good spotmeter and uses some more homogenous material,
e.g. plastic (instead of the piece of paper I was using) more accurate
results could be achieved.
It seems that my EF35/2 and 50/1.8mkI as well as the Sigma zoom let
more light go through the lens than the others. I double checked the
EF35/2 with the same results (after measuring the other lenses).
Between extremes the difference at e.g. f4 is 0.5 stops which
surely may have some practical meaning as well.
The values in the tables are the "most common values" I saw when I
was measuring repeatedly in each case. The variance was often +-1/10
step between adjacent measurements.
For 1/2-stop measurements the c value should increase by 5
(e.g. 2, 7, 2, 7, ...) and for 1/3-stop measurements by 3.33
(thus e.g. 0, 3, 7, 0, 3, ...) I assume.
For EF35/2 I listed the expected values as well for the last three
lines. You see some unexpected results which may be explained partly
by vignetting which was visible and which may affect the results
(spot meter would have been a better measuring device).
About the case which is especially interesting to you Thomas:
Unfortunately (?) I have only a couple 3.5 or 4.5 lenses but based
on these two cases your theory, Thomas, seems to be the better one
although I'm not totally sure yet because we can't forget the
expected/assumed quality difference between 28-135 and 100-400
and also the 100-400 results wide open at 100mm do not match exactly
either theory. They are totally in different focal length ranges
as well.
Who knows what the Canon tolerances are, perhaps +-0.25 stops?
Maybe some day I do real world test, thus take out-of-focus photos
of my lightbox with some lenses and see if they really look
differently exposed. Interesting isn't it...?
regards Vesa
EF35/2 (1/2 stops)
------
f4.5: 2.0/7
f4.0: 2.8/2
f3.5: 2.8/7
f2.8: 4.0/1 (should be 4.0/2)
f2.5: 4.0/5 (4.0/7)
f2.0: 4.0/7 (5.6/2, thus is -5/10 = -0.5 stops)
EF24/2.8 (1/2 stops)
--------
f4.5: 2.0/4
f4.0: 2.0/8-9 (couldn't decide)
f3.5: 2.8/3
f2.8: 2.8/7
EF50/1.8mkI (1/2)
------
f4.5: 2.0/5
f4.0: 2.8/0
f3.5: 2.8/5
f2.8: 4.0/1
f2.5: 4.0/5
f2.0: 4.0/9
f1.8: 5.6/1
EF100/2 (1/2)
------
f4.5: 2.0/3
f4.0: 2.0/7
f3.5: 2.8/3
f2.8: 2.8/8
f2.5: 4.0/2
f2.0: 4.0/7
EF200/2.8mkI (1/2)
------
f4.5: 2.0/4
f4.0: 2.0/9
f3.5: 2.8/4
f2.8: 2.8/8
EF28-135@28mm (f3.5) (1/2 stops)
------
f5.6: 1.4/8
f4.5: 2.0/3
f4.0: 2.0/7
f3.5: 2.0/9 (only 0.1 stops wider then the expected f4.0 value)
(1/3 stops)
f4.0: 2.0/7
f3.5: 2.0/9 (only 0.1 stops wider then the expected f4.0 value)
Sigma 28-70/2.8 @70mm EX (1/2 stops)
------
f4.5: 2.0/6
f4.0: 2.8/1
f3.5: 2.8/5
f2.8: 2.8/7 (expected 4.0/1 thus -0.4 stops)
@50mm
f2.8: 2.8/9 (thus -0.2 stops)
EF100-400@100mm (f4.5) (1/2 stops)
------
f6.7: 1.4/3
f5.6: 1.4/8
f4.5: 2.0/4 (actually +0.1 stops from expected)
(1/3 stops)
f5.6: 1.4/8
f5.0: 2.0/2 (expected 2.0/1, so +0.1 stops)
f4.5: 2.0/4 (expected 2.0/5, so -0.1 stops)
END-OF-TABLES
_________________________________________________________________________
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
***********************************************************
