Accurate Color
Accurate Color
- Subject: Accurate Color
- From: Asher Kelman <email@hidden>
- Date: Mon, 16 Jul 2007 14:16:35 -0700
On 7/14/07 6:16 AM, "Jack Bingham" wrote:
I tend to think we are ignoring a rather large group of photographers
here who make their living shooting products for reproduction. In
this case we actually want "accurate" output color. My simple
definition of accurate is in the eye of the check holder. The red
sweater has to match or there are so many returns and the customer
comes back to complain about the match and yada yada yada.
If the red sweater has to match, why not take a measurement of it
under an agreed lighting. (Yes, the light might drift but we can
specify that too!)
This is important in the clothing industry where a company orders
cloth for 10,000 dozen jackets. When the cloth is chosen, the color
is known.
It can be measured, same on delivery. The process can be repeated at
intervals along the cloth. All one has to do is define the light and
delta for the measurements and then the color is accurate within
these agreed parameters. One cannot simply return a $100,000 of cloth
you ordered in spring because it appears different when it arrives in
June for fall season!
The same with a sweater being photographed if accurate color matters
to you! How hard is it to measure the color in the light that's
important to you and agree on what Delta would be acceptable in a
range of tonalities. If one wishes to do it, the task is such a
simple one.
Andrew Rodney Wrote:
That's STILL not scene refereed which IS accurate (measured, defined)
color.
Define accurate Jack (without the quotes). OK, you have to use some
metric
like a measurement otherwise we're just talking about what a client
or end
user feels matches. And even when color match, the number often
don't. We
can thank our friend metamerism for this. And every color match?
Matches to
what delta? If so, how do we decide this short of measuring the two?
Pleasing color can and often is the appearance of the output referred
color
producing what appears to be a match to the scene but we simply can't
measure this success we can only agree that they match.
IF you want to say that using sound color management and color
management
products produces matching color that client and creator agree match,
OK I'm
fine with that. But to call that accurate is a stretch since its up to
interpretation and there is no way to prove it using scientific testing
methods.
Well, if we have such limitations on measurement we would have to
have endless calculations in measuring 1 meter since the length would
change as temperature changes and possibly as the earth rotated and
as the gravity compressed the mass of the ruler differently. All
measurements are approximate, except counting integers, where the
counting requirements do not exceed the rate of appearance of
measurable events.
So accurate, at the most, in the context of color means "closely"
matching in parameters considered significant for the purpose at
hand. "Close?" What does that mean? It means near enough!
So an accurate color for delivering a GM color chart would be, of
course different than for delivering a print for a wedding, a dog
show or a Lancome commercial. Only the GM color chart must be printed
to some measured standard and measured tolerance. The rest can be
done with the eye or again by measurement. All would be "accurate"
within the context of use.
Asher Kelman
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden