Re: [APML] Synthetic Luminance Channel for Film Images

[Roland Roberts <roland@astrofoto.org>]

>>>>> "Jerry" == Jerry Lodriguss <jml@astropix.com> writes:

>From your first post

    [...snip...]

    Jerry> Notice they are making the luminance layer out of a
    Jerry> combination of the red, green and blue filtered images.

    Jerry> I wonder if they know they are not gaining anything by
    Jerry> doing this because they did not shoot a separate high
    Jerry> resolution luminance channel with a higher signal to noise
    Jerry> ratio to start with.

    Jerry> The signal to noise ratio from combining the RGB channels
    Jerry> is the same if it's color, or if it's black and white.

A while back we had an extended discussion about tri-color photography
and filters; specifically whether or not filters should overlap at the
100% point of the 50% point.

I don't think they are doing the "wrong" thing but agree that they are
not "gaining" anything.  They *are* making the most out of the
information they have.  They happen to know 4 pieces of information;
the total photon count in a particular spot, integrated over all
wavelengths, and the fraction that falls into each of three different
wavelength bands.

With CCDs, adding the RGB components does give you the same things as
a luminance shot due to the linearity of the CCD.  It doesn't give you
higher resolution, but it gives you a higher S/N than if you didn't
combine them.  Insofar as you are just counting photons, whether you
count the red, green, and blue ones all at during an exposure or only
count one color during each of three exposures, it makes no
difference.

Mind you, your knowledge about the *color* is not enhanced by this
process.  Further, there is a correlation between color error and
luminance error which is not present when you have a separate
luminance shot.

roland
-- 
		       PGP Key ID: 66 BC 3B CD
Roland B. Roberts, PhD                             RL Enterprises
roland@rlenter.com                     76-15 113th Street, Apt 3B
roland@astrofoto.org                       Forest Hills, NY 11375

--  APML Archives at <http://astro.umsystem.edu/apml/>  ---
             Unsubscribe at <majordomo@seds.org>