[argyllcms] Re: camera profiling with newest version of Argyll

  • From: Ben Goren <ben@xxxxxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Thu, 27 Aug 2015 09:00:44 -0700

On Aug 27, 2015, at 8:27 AM, Chad Johnson <chad.chadjohnson@xxxxxxxxx> wrote:

I certainly white balance my camera but these files have nothing to do with
the camera which comes along at a later step.

The RGB values in the file come from the camera, no? In your file, they're
close to equal, but they shouldn't be because the patch they correspond to
isn't white but blue.

Imagine if you had done the click-to-balance thing on a (noticeably) light blue
patch. You'd expect a yellow cast to the resulting file, right? And the light
blue patch would come out white instead of light blue, no? That's exactly what
you've done.

The chart you have doesn't have any patches suitable for a click-to-balance
tool. I don't know what tool you're using for RAW development...but damned few
of them even have the ability to set white balance correctly, let alone make it
easy and / or practical....

The odd spectrum is due to the light source which is fluorescent and is
partly what I'm trying to calibrate (ie. at the end of the day the
calibration is for the lightsource + camera).

The spectrum I'm referring to is the reflectivity spectrum of the patch on the
chart that you measured with the ColorMunki. That patch, intrinsically, doesn't
reflect much UV light directly, and instead re-emits that energy as blue light.

I've attached the spectrum for your reference.


It's a mercury-based fluorescent bulb, of course, with a strong UV filter. That
will again complicate things. I would not recommend using that light for
color-critical work; incandescent will give you far superior results. Indeed,
it only has a CRI of 87.8...you can do better at your local hardware store, and
even those aren't so great.

My equipment is not counterfeit. Ridiculous!

At this point, I'm hoping that it is, indeed, a counterfeit chart...I averaged
all the white patches in your earlier file, and there aren't any outliers...nor
any question but that there's FWA in the paper. For X-Rite to have made the
product like that is...well, it'd be even more disturbing for that to be the
case than to think that somebody is selling counterfeit X-Rite products.

b&

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

Other related posts: