On Feb 12, 2009, at 10:18 AM, wsolum wrote:
> I realize I'm in a datacolor group but the question I'm asking is
> general, not product specific. Sorry if that causes any problems.
No problem. Its certainly better than asking product specific
questions, about some other company's products. : )
>
>
> I shoot in RAW and use LR/ACR to post process.
That all matches my own process, and if you are the person who was
using a Canon 5D, that matches my own flow as well (though the 5D Mark
II has now replaced the original 5D for me).
> The red/magenta
> problem is most notable on caucasian subjects when shooting strobe
> lighting.
What color of strobe lighting?
> Even when I use a grey balance card the skintones just
> don't look right to me. I was shooting in Adobe RGB mode
You said RAW above, so the AdobeRGB setting is not affecting your
images, its only applied to Jpegs...
> and editing
> in Adobe RGB in photoshop but I'm switching to sRGB.
Careful there... AdobeRGB is fine in Photoshop, and sRGB is fine for
the web or for sending out to labs too clueless to convert to a custom
profile, but if you assign, instead of convert, between the two color
spaces you can end up with (depending on which direction you go in)
either undersatured skintones or... (drum roll please) overly red
skintones in caucasian subjects... including red noses on older men
that make them look like lifelong drunks. Sound familiar?
> Lightroom uses
> ProColor or something like that and handles all the conversions
> automatically.
Correct, their internal space is not of much consequence here, as long
as you are rendering out to AdobeRGB at the end, and honoring that in
Photoshop.
> I have noticed variations in image renderings in LR
> vs Photoshop but have figured that was a factor of the background
> color (LR is dark grays, PS is mid grays).
There are other possible causes... see above.
>
>
> So I was wondering what affect these pasty walls, flourescent
> lighting or near total darkness were playing on this process.
Pasty isn't quite as good as dead gray, buts its pretty acceptable.
And in darkness, the wall colors matter far less. And you don't have
windows or other bright light sources effecting the eye's color
balance (again, assuming this is the same person with the previous 5D
question), so you are not in bad shape. What color the fluorescent
lightbulb is has some effect, particularly if you are viewing prints
under it. And if the room is too dark, then a bright display can cause
glare and eye fatigue, but thats a different issue.
> Should
> I have lighting at all and what? What white-point is preferred, is
> there a luminance that is best or does it depend on intent? Any
> other factors I'm not considering?
These are all valid questions, and if you owned Spyder3Elite, then the
software would walk you through a fair amount of this. As I recall,
you are the person using a different product, so I really can't help
there. But if you have a calibrated display, you should not be getting
bad color on screen from a 5D. Lets take your own photos and process
out of it. Open the Matrix test image from the Test image folder in
your Spyder folder (I assume you own at least one Spyder3 product, or
you wouldn't be on this list...). Those are precorrected images, with
carefully chosen skintones (some of them shot with a 5D). There is a
blond with what we call a yellow skintone, and a girl with what we
call a pink complexion, to bracket the caucasian range. There is a wax
work of Marilyn, to show tungsten burn to white. And there is a dark
skintone in deep light, more for shadow detail but it also shows the
color balance issues in dark skin. If these images look, and print,
okay for you, then your problem has either to do with your flash and
settings (not my area to diagnose, I avoid flash like the plague), or
its your workflow settings, which I make a few comments on above.
Either way, its not really a Datacolor issue...
C. David Tobie
Global Product Technology Manager
Digital Imaging & Home Theater
CDTobie@...