Re: Targeting for different illuminants
Re: Targeting for different illuminants
- Subject: Re: Targeting for different illuminants
- From: Dan Reid <email@hidden>
- Date: Tue, 03 Apr 2001 10:54:35 -0600
on 4/2/01 10:42 AM, Steve Upton at email@hidden wrote:
>
At 4:14 PM -0600 4/1/01, Dan Reid wrote:
>
> Hey color geeks,
>
>
Heya Dan
>
>
> Any body given some thought to targeting images for specific illuminants?
>
>
Yup...
>
>
> Most of our profiles are built using a D50 illuminant L*a*b* space. When we
>
> convert from one profile to the next the profiles usually contain a D50
>
> L*a*b* PCS.
>
>
Or a D50 XYZ space, we really don't have a choice with this one as it
>
is hard coded in the spec in the CMM's
>
>
> I am thinking that we would need to use a L*a*b* space with the
>
> targeted illuminant (i.e. Illuminant A [incandescent], L*a*b* space) and
>
> then use this as either an abstract profile or simulation profile in three
>
> fold transform.
>
>
You can hack the flavor of Lab by using an abstract profile but as
>
David Tobie mentioned you really should do it from the Spectral data.
>
>
If you are using the Spectrolino, download their tool ColorLab from
>
www.logosoft.de. It will allow you to read targets with spectral data
>
for each patch rather than D50 Lab. You can then convert the spectral
>
values to many different flavors of Lab including D50, D55, D65, D70,
>
A, C, ideal) you can also convert them to various flavors of density
>
if you want. (whether or not D55, D70 are recognized standards or
>
just convenient notations is a different issue)
>
>
> I know we have a D65 L*a*b* profile but not sure if another
>
> illuminant could be used instead? Any thoughts on this topic?
>
>
Again, I think you can hack the Lab flavor but you should really
>
start with the spectral data if you can.
>
>
We use this technique in our custom profiling solutions for clients.
>
Tweaking the Lab flavor when profiling can complicate the issue but
>
when profiling ink/papers that exhibit metamerism it can sometimes be
>
the only way to go to get neutral looking good under specific
>
lighting conditions. It is important, however to know the lighting
>
conditions you are shooting for so sometimes we make onsite visits to
>
determine the environment first.
Interesting approach. Fooling around in Colorshop the other day I noticed
when I switched illuminants the XYZ data numbers changed but not the Lab
values. I am still not clear as to how we handle *images* and not measured
spot colors. Maybe good ol' Steve could post different illuminant flavors of
Lab at profilecentral.com for us to use as abstract profiles? wink, wink,
<g>
--
Dan B. Reid
RENAISSANCE PHOTOGRAPHIC IMAGING
Color Imaging Solutions Provider
http://www.rpimaging.com | email@hidden
Toll Free: (866) RGB-CMYK [ 866-742-2695 ]
Local: (505) 471-4126