Bob,
I would add to your elaborate test schema the control of doing the
same recording onto B&W film, and comparing the light falloff
differences. EV 0.6+/- is a mighty fine distinction. But we can do
better than that with the digital sensor as we can look at the 12bit
quantized data... ;-)
My prediction: You're not going to see any center-edge illumination
differences in results with SLR lenses at 40-50mm focal length that
aren't also apparent on film sized to the sensor format. 40-50mm
focal length SLR lenses do not present the issues that shorter focal
length lenses might: the primary nodal point with such lenses is well
forward of the sensor, so the ray trace to a 16x24mm sensor does not
deviate significantly from a telecentric path. These kinds of effects
(mosaicing, light falloff, chromatic aberration due to lack of
telecentric light path) would be much more apparent with focal
lengths in the 24mm range and shorter, and more particularly with an
Epson/Cosina RD-1 body using older, short, non inverted-telephoto
lens designs.
Oh, you might see sensor back-reflections in some cases: that would
be the result of lens coatings, which is another aspect of lens
tuning performed on the DA and D-FA lenses.
Unfortunately, I only have the A50/1.4 in the range of lenses you
want. A test comparing the DA14 against the A15 might be more
interesting from the point of view of comparing performance on the
digital sensor vs film, as there are no other "digital vs film"
design prime pairs to evaluate. A similar comparison between, say,
the FA20-35, FA-J 18-35, DA18-55, and DA16-45 would also be interesting.
Godfrey
On Sep 6, 2005, at 9:17 AM, Bob Blakely wrote:
Illumination only. This should be intuitively obvious to the most
casual observer. Most other characteristics are independent of
whether the lens is "for digital" or "for film", are independent of
flatness of field, etc., and are simply a function of tradeoffs in
lens design or care in manufacture. This "light falloff" due to
sensor designs has been the bone of contention in many circles.
-------------------------------------------------------
Procedure is:
Snap with the digital lens.
Snap with the film lens.
For each lens:
For each of red, blue, green:
Separate into layers.
Pick a small center spot in each and average all pixels.
Using green as "standard", Calculate the multiplying factors
for red & blue
in the averaged spot.
Multiply the red & blue over entire image by each's
multiplying factor.
We now have the complete field for each lens calibrated to a
center "gray".
Multiply the film lens's entire field (each RGB layer)
by the digital center gray / film center gray.
We now have each field, D lens & F lens, calibrated for equal
center grays.
Pick four small edge spots and four small corner spots in each lens
and average all pixels.
For each of the eight spots for each lens:
calculate the difference, F lens / D Lens in EV.
-------------------------------------------------------
For me, if the difference is less than 1/6 stop, I don't give a
damn, at least at that aperture.
Regards,
Bob...
----------------------------------------------------------------------
------------------------------
By all means, marry. If you get a good wife, you'll become happy;
if you get a bad one, you'll become a philosopher.
- Socrates
From: "Godfrey DiGiorgi" <[EMAIL PROTECTED]>
But what are you trying to compare, Bob? Evenness of illumination,
resolution, contrast ... what? A gray card will only show you
evenness of illumination. At f/2.8, it's a toss up between the
two 50mm lenses to predict which will do better, although at
copystand distances I'd put my bets on the dedicated macro lens.
A D-FA50 macro lens is designed to be best at flat-field imaging,
stopped down to f/8-f/32, in the near-focusing range. An A50/1.4
is designed for general pictorial use at wide apertures, and will
not perform at its best at copystand distances.
Comparing a DA40 and a M40 pancake makes more sense as they are
both designed to do the same job of general pictorial work.
Unfortunately, the only lens I've got of the above four is an
A50/1.4. I can shoot a gray card for you, but I doubt that it's
going to tell you very much.
Godfrey
On Sep 5, 2005, at 1:52 PM, Bob Blakely wrote:
What I would be interested in, then, is D-FA50 Macro wide open
and the A50/1.4 at the lowest common stop, or the DA40 and the
SMC 40 at the lowest common stop.
Regards,
Bob...
--------------------------------------------------------------------
-- ------------------------------
By all means, marry. If you get a good wife, you'll become happy;
if you get a bad one, you'll become a philosopher.
- Socrates
From: "Godfrey DiGiorgi" <[EMAIL PROTECTED]>
There are no "digital" prime lenses between 20 and 35mm.
Pentax lenses designed for digital SLRs, to date, include:
DA14
DA16-45
DA18-55
DA40
DA50-200
D-FA50 Macro
D-FA100 Macro
If I get a chance to set up my copystand and lighting, I'll
make a gray card exposure with the A24/2.8 or FA35/2, but I
doubt very much that you'll see anything different compared to
making the same exposure on a 35mm film negative and then
cropping the negative to the 16x24 sensor format dimensions.
Godfrey
On Sep 5, 2005, at 12:56 PM, Bob Blakely wrote:
Yes. Also, if one of you digital dudes would shoot a gray
card frame edge to frame edge with one of your digital primes
(say between 20 and 35 mm or so) and one of your film primes
of *equal* focal length, I would be interested in comparing
light fall off at the corners. Tripod mount camera on
centerline wit perpendicular from the center of the gray
card. Light gray card evenly with two side lights placed on
each side of the camera and such that there is no equal angle
reflection from the light to the card to the lens. No flash.
Just like you were photocopying art.
