As I promised, here are my test images from shooting the Tokina
28-80/2.8, Kiron 28/2 and Zuiko 28/2.8 at apertures from max to f/8 on
the Canyon 5D. Image size: 4368 x 2904
The test conditions were really not very good. In the morning whilst I
was mowing the lawn it was lightly overcast and the wind was dead calm.
By the time I actually got out to do the shooting there was bright sun
playing tag with the clouds and a fairly stiff but variable breeze. I
tried to shoot when the breeze died down but I can't say that I was
always successful so fine detail may be obscured at times by motion.
Large aperture shots were at pretty high shutter speeds, however, with
the Kiron f/2 shot being at 1/8000th second. Slowest was the Zuiko f/8
shot at 1/100th. No special precautions other than a tripod. No mirror
lock-up, no remote cable release, no self-timer. Given the variability
of the lighting any judgements about contrast are probably dangerous
unless you can see the same shadow intensity in the center crops.
I intended to shoot JPEG's with camera defaults but forgot I had the
camera set for raw plus a 4MP JPEG. So I had to process the raw images
which I did with Capture One LE. I used defaults except for forcing the
white balance to the center of the white door and adjusted exposure down
by less than 1/3 stop on a few where the door was slightly blown in
direct sun. Exposure was AV mode which proved to be dead on with the
Tokina and required a -2/3 stop adjustment with the Kiron and -1/3 stop
adjustment with the Zuiko.
Some fallout of the test. I discovered that the reason I hadn't used my
28mm Zuiko in eons is that there appears to be a problem with the
aperture ring detent mechanism. I am unable to open the lens wider than
f/5.6 so the only Zuiko comparison shots are at f/5.6 and f/8. I was
also confusing it with my 35mm Zuiko which is a silvernose. The 28mm is
actually a multi-coated black nose although is is not marked "MC".
The three links below show:
The entire image with yellow rectangles showing the center and edge
sections which were extracted to form comparison matrices
<http://www.chucknorcutt.com/temp/Test%20image%2C%20center-edge%20markers.jpg>
The center performance comparison matrix
<http://www.chucknorcutt.com/temp/Center%20performance.jpg>
and the edge performance comparison matrix
<http://www.chucknorcutt.com/temp/Edge%20performance.jpg>
The Kiron edge performance images (except the last) make it appear that
the Kiron image is larger and thus the focal length might be longer than
28mm. But based on the last image it appears to be only my error in
cropping out the right size piece.
Be sure to click on these last two comparison images after they load to
be sure they're being displayed full size. Lens ID and aperture labels
are at top and left side. Each of the test snippets is a 300x200 pixel
crop. When displayed on your screen at 1:1 you need to keep in mind
that what you're seeing would be like taking a snippet from a 30x45"
print or even larger depending on your screen's pixel pitch.
Conclusion: There are certainly some differences bewtween these lenses
and the Kiron is notably softer at the edge until f/8. There also may
be a bit of vignetting with the Kiron at f/2 but it's hard to say.
Beyond that I don't think the differences amount to a hill of beans
unless you're really into 30x40 prints examined at reading distance.
This is also too much work. If I had to do it again I'd probably shoot
it in the basement using Alien Bees for light. :-)
Chuck Norcutt
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|
