Re:camera profiling
Re:camera profiling
- Subject: Re:camera profiling
- From: Jack Bingham <email@hidden>
- Date: Mon, 11 Mar 2002 08:39:29 -0500
"You are skirting the issue. The issue isn't whether a good camera profile
saves time, makes clients happy or produces good results."
Its funny, I thought we were all in the business of making our clients happy. We can ramble on and on
and on about theory. All I care about is whether the camera profile eliminates most of the hassles and
mine do. I am not seeing the catastrophic failures Andrews talks about and I shoot in every imaginable
kind of light day in day out. I always got pretty good results from Colorblind and ICC Dcam has been
exceptional. I personally have profiled all Phase One models and Nikon models and many of my associates
have done Imacons Canons Kodaks and even Coolpixs using the same software and techniques I use. We are
all getting great results from one profile with the exception of the PhaseOne H20 which required
tungsten and a strobe profiles. Many of us, Derrick Brown the longest, have been trying to make the
point that profiling a scene does not work but profiling a camera does. Here is why. Lets look at a set
in the studio that has been lit. The shot contains 5 red fire extinguishers on a blue background. It is
lit with one softbox overhead. a spotlight from the left rear and an umbrella from the front. The spot
light hits the sides of the extinguishers but not the front of the flat target. The silver umbrella is
old and the flashtube is new so they are both very blue. The softbox is old and the nylon has an
obvious yellow cast. Here is what happens on the target. The light across it is not even, at the bottom
the blue seamless is reflecting up into the lower patches and on the left one of the extinguishers is
casting some red into the left side of the target. The upper right corner is 25% brighter than the
lower left. All the gray patches on the Macbeth have a blue cast. The two front light sources mix
unevenly across the target, warmer at the top from the softbox and cooler on the right from the
umbrella. So given this circumstance does anyone really believe any software can resolve how to render
that target accurately? You might suggest this is extreme but I got just such a target last week from
a photographer I have been helping. Since the software needs to resolve not just the color but the
brightness of each patch it will in the end change the lighting and the color of everything. We contend
that a profile is built from a target that is evenly lit, carefully exposed and color balanced so the
comparison between image and data are free from any outside influences. Then with proper gray balance
(linearization) a camera profile becomes a guide for how a particular camera responds.
[demime 0.98b removed an attachment of type text/x-vcard which had a name of jack.vcf]
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.