Re: [UFRaw-Devel] WB presets for cameras with many presets

[Udi F. <udi...@gm...>]

> The fine tuning sets balance for an illuminant with some temperature
> (CCT probably).  Using temperature and green is awkward (I'd like to
> see temperature with green 1.0 do the right thing mostly, and then
> only fine-tune green if needed).

As I see it, Green=3D1.0 should mean that your light source is a perfect
black body. I Don't expect fluorescent light to give me Green=3D1.0.

> But, the real point is to get an
> image that's "what I would have gotten from camera had I selected foo".

Shooting raw, I always set the camera to automatic WB. Therefore, the
camera WB presets are not of much interest for me, anyway I don't use
them on the camera.

> Kelvin presets include green level, so there's no double control.
> But, they aren't the same values, and there is stuff to figure out
> here.
>
> As Bill said, surely camera spectral response varies, so the point of
> presets is to match camera behavior.
> Given all that, perhaps the temperature controls for a camera with
> kelvin presets should interpolate among them.

What I'm trying to do is calculate the temperature in RGB space and
convert it to the camera color space using the color matrices. This is
where the camera dependent behaviour should come from. Assuming that
this linear transformation captures the full camera spectral response
is naive, but I think that it is the best that can be done.

I would like to have the D2X multipliers for Kelvin presets, to see if
my code can reproduce them, at least approximately.

> With the D2X there's the additional challenge of the analog white
> balance multipliers.  It's not clear to me how those get recorded in
> the makernote.

What are those? dcraw reads the WB setting from the raw file, or are
you talking about something else?

> I wonder about normalizing the average to 1, or something that tends
> to result in the same values as the camera jpegs for color
> temperatures near Camera WB.  So perhaps using values that result in
> magnitudes near the raw values processed by original camera multipliers.

If you are near the Camera WB, the change in luminosity is very small.
I'm not sure what you are trying to fix here. Can you give me a
concrete example where changing the WB in UFRaw changed the luminosity
in an undesirable way.

Udi