[argyllcms] Re: Absolute light meter calibration?

  • From: Marwan Daar <marwan.daar@xxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Fri, 7 Aug 2015 16:52:27 -0400

I wanted to do something similar (I'm more interested in repeatability
measurements). After consulting with a friend, the idea was to use a
tungsten filament. When it's fed a current, it acts like a pretty good
black body radiator, from what I understand. If you combine a power source
with a voltage regulator (to maintain the voltage at a constant, non
fluctuating, level), you might be able to calculate the expected spectral
flux, and use geometry to figure out irradiance or radiance (and therefore
illuminance and luminance). You'll probably need an oscilloscope to monitor
power levels.


On Fri, Aug 7, 2015 at 3:50 PM, Ben Goren <ben@xxxxxxxxxxxxxxxx> wrote:

Nearly all photographic uses of light meters are ultimately geared towards
relative measurements of one kind or another. Nobody typically cares about
the _actual_ value for W/m^2 or Lux or what-not; just that, when the meter
says to set exposure for such-and-such a combination of aperture, shutter,
and ISO that the resulting photograph is what you expected.

I happen to, for the first time I can think of, have a use for getting a
truly accurate (within typical photographic margins of error, ideally less
than 1/3 stop) absolute measurement of local solar radiance. It's for an
esoteric project that might or might not pan out for making
colorimetrically-accurate photographs of the Moon and, perhaps, other Solar
System objects. The measurement would get combined with a spectrographic
measurement of the Sun and the ASTM E-490-00 extra-atmospheric solar
spectrum and all sorts of other math and what-not to, amongst other things,
eliminate the color cast caused by the atmosphere.

Can anybody suggest some sort of reliable way to get an accurate
calibration in absolute units of either a typical Sekonic meter or an i1
Pro in ambient mode or something like that? I'm thinking, ideally, of a way
of creating a light source of dependably-known illuminance that I can
measure. An unfrosted incandescent bulb, maybe? And some way of verifying
that the wattage printed on the bulb matches what it's doing? And the math
to calculate the rest?

It's both a photographic exercise and an excuse for me to do some hands-on
basic physics, so I'd welcome something, for example, suited for an high
school or college physics lab...but I _do_ want the final result to be
reasonably reliably accurate to no worse than 1/3 stop.



Other related posts: