At 11:19 PM +0000 3/29/02, olympus-digest wrote:
>Date: Fri, 29 Mar 2002 18:52:03 +0001
>From: "Enrique Cabrera Rochera" <qcabrera@xxxxxxxxxx>
>Subject: [OM] Shooting paintings
>
>Dear friends:
>
>On monday I have to satisfy the demands of one of my most challenging
>customers:
>my mom. She is a semi-professional painter and usually does an exposition per
>year.
>So far, my dad did the photographies with the OM10 that started the saga at
>home.
>However, he never took it seriously enough, and the brochures were never
>great. There
>are a few factors that you should be aware of: [snip]
I too photographed paintings for my mother. I the a method detailed on pages
270-273 of "The Hasselblad Manual", Ernst Wildi, Focal Press, 1980, adapted to
an OM-1 with a 50/1.8 lens. It worked very well, and Mother was happy.
1. The camera has a linear (not circular) polarizing filter mounted on the
lens, and is on a tripod. This method cannot be done with handheld anything.
2. There are at least two lights (hot lights or strobes), each with a big
sheet of linear polarizer in front, tight enough to prevent leakage of
unpolarized light. The optical quality of these polarizers is not important,
but they must all be adjusted to be perpendicular to that of the camera
polarizer. This is accomplished by turning the lights on one at a time, and
rotating the lamp polarizer until the least light comes out of the camera
viewfinder. Do not touch the camera or its polarizer while making these
adjustments. When all lamp polarizers are adjusted, turn all lamps on. You
should see through the camera very deep colors and no surface reflections. The
effect is amazing. Note that metallic objects will look black in this
crossed-polarizer setup, as reflection from metals does not depolarize the
light. (If you do want some surface reflections, rotate the camera polarizer
slightly, to taste. )
3. Use the camera to meter the exposure, but bracket heavily. Incident-light
metering (my usual favorite) won't work here because the crossed polarizers
absorb most of the light; only light that penetrated into the painting's
varnish and paint (becoming depolarized in the process) and came back out again
will make it to the film. Even with perfect polarizers, the loss will be 2
stops, but 5 to 6 stops is likely in practice. The short distance from light
to target helps greatly here. Reflected-light metering can work, once one
knows how many stops the camera polarizer will cost, but I always used the
camera's built-in meter instead.
4. Stop down to ensure that the depth of field is sufficient to handle
deviations from flatness in both the painting and in the optics of the lens.
5. The smallest feature on my mother's painting (oil on canvas) was one or two
millimeters in diameter (more or less the size of the smallest brush), and the
painting was something like 2x3 feet (610x915mm) in size, which becomes a spot
(1mm)/(610/24)= 0.039 mm in diameter on the film, so the required resolution is
something like 0.5/0.039= ~13 line pairs per millimeter. It the camera
polarizer has been adjusted to allow some specular highlights to be seen,
greater resolution may be needed for those highlights to look right, but still
13 line pairs per millimeter is very relaxed, so any reasonable film and lens
should work OK. (I used Kodacolor 100 print film, then one of the best
consumer-grade films around.)
6. To prevent keystone distortion, the film plane must be parallel to the
plane of the target painting. This is essentially impossible to arrange by
eye, without some help. Hasselblad sells an accessory, the "linear mirror
unit", to do this, but it didn't fit 35mm cameras and in any case its cost far
exceeded my budget, so I made my own equivalent:
I purchased two small plate-glass mirrors from a mirror shop, for less than
US$10. These mirrors were square, 3 inches (75mm) on a side. In one, I
removed the backing paint (using paint stripper) and metallization (using
sulphuric acid (intended for an automobile battery) on a Q-tip) from a circle
0.5 inch (13mm) in diameter in the center, leaving a clear window in the
mirror. The exact sizes are not important. Now, if one holds the mirrors face
to face and looks through the little window, one sees what looks like a tunnel.
If the tunnel looks straight, the mirrors are parallel. This test, even with
homebrew equipment, is very sensitive. (This illusion has been called "the
tunnel of love"; don't know why or where it came from. Perhaps from a
misogynist saying it's all an illusion.)
In use, the windowed mirror is held to the lens and camera with two rubber
bands, and the other mirror rests on the painting, and one looks through the
viewfinder to see the tunnel illusion. Adjust camera on tripod until the
tunnel appears straight.
7. I don't recall barrel or pincusion distortion being a problem, which isn't
a surprise. The painting itself was impressionistic, and it would have taken
gross amounts of distortion to be visible unless one had painting and life-size
photograph side by side. Certainy, a few percent distortion of any kind would
have been invisible.
Joe Gwinn
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|