RE: 2 degree, 10 degree and observer cmf differences
RE: 2 degree, 10 degree and observer cmf differences
- Subject: RE: 2 degree, 10 degree and observer cmf differences
- From: Roger Breton <email@hidden>
- Date: Tue, 23 Aug 2011 09:19:33 -0400
Tom, Bob, Graeme, Edmund,
Thank's for changing the title of the thread,
One question I always had but never seem to find an answer for. Here goes. I
see lumps of industrial groups flocking to the 1964 10 degrees Observer for
various reasons. The ink industry, for one. Textile, paint and plastics have
always used the 10 degree Observer. Now, possibly, monitor manufacturers?
Leave me space to breathe. To me, it is one thing if research show that 10
degrees is "more applicable" or "better explain" visual judgements on an
emissive device, comparing to a hardcopy proof. But I've missed a few CIC
meetings, I must say, and journal entries so I may be behind, sadly. But on
the other hand, I always read and am a firm believer that rods, by "design",
cannot operate at Luminance levels for which emissive devices are used,
which are "photopic" by nature. My understanding of rods is that they
operate in very low light, below a couple of Lux. Please correct me if I'm
wrong but, naïvely, if I believe the scientific account I read (it's
everywhere), rods get "saturated" at photopic levels and therefore cannot
contribute to light or color sensitivity. Now that I get rods out of the way
conceptually, I agree, intuitively, that using a "larger" area of the fovea,
to take into account cones that are not situated right into the central
retina (fovea centralis) may make for different spectral sensitivities that
better explain some color judgements? **
Is that what the "shift" to the 10 degrees Observer is about?
I have my doubt. Even if research show that comparing two monitor
full-screen white, side by side, use of the 10 degrees Observer would appear
to better reconcile our visual impressions of the perceived differences, the
fact of the matter remain that, when viewing typical imagery for
reproduction, a face, a tree, an animated street, an animal, a bottle of
shampoo, a model wearing a yellow dress, we are still dealing with small
screen real estate areas. That's why, to me, all this "large FOV" discussion
is moot.
And how would you propose to mix D50/2 from reflective originals with D50/10
from emissive device in an ICC transform: just use the 10 degrees numbers as
is? Let the ICC profiler built a monitor profile based on 10 degree
colorimetry? Keep that as a secret to the CMM? Mixing and matching the two
sets of cmf in a profile conversion to the screen? Allow me some to be
sceptical , I doubt that it will be any better than 2 degrees colorimetry.
To me, there are other phenomenon at work that affect our color vision still
beyond a given choice of Observer, over and above the factors that Graeme
suggested. But that's my humble opinion.
Best / Roger
** I remember attending the CIE Expert Symposium in Ottawa, a few years ago,
and the question came up to the floor that, why bother with the 2 degree
Observer? Why not adopt the 10 degree Observer? There was arguments from
both side of the room to the question but, in the end, no one wanted to
change the state of the Observer because, as was said, the 2 degrees still
worked better "on average". Other color matching functions were evoked, in
the heat of discussion, like the 1959 Stiles and Burkes. If memory serves,
the consensus was to expend more efforts on LMS which have been established
physiologically, whereas the 1931 Observer only model our spectral
sensitivity indirectly. I've lost track of where color science has advanced
today :(
-----Original Message-----
From: colorsync-users-bounces+graxx=email@hidden
[mailto:colorsync-users-bounces+graxx=email@hidden] On
Behalf Of Thomas Lianza
Sent: August-23-11 6:32 AM
To: Bob Frost; email@hidden; email@hidden
Subject: 2 degree, 10 degree and observer cmf differences
Hi
I've changed the name of the thread to get it more on target. In my
discussions with some engineering folk in Japan they mentioned that they
felt that displays should be assumed to be a "greater than 10 degree event"
The point is that when two displays are put side by side our impression of
the white point and natural adaptation is based on the wide field presented
by the scene. You will find some displays leave the factory using 10 degree
assumptions of D65 or D50. There is some real merit to this thinking, in my
opinion. Viewing emissive scenes is quite different than viewing a
reflective scene because all colors are combined from the three primaries,
plus a relatively constant level of flare which is dependent upon the
ambient illumination. This is why small amounts of white point deviation
have such large effects upon the absolute color. This is why you can have
really sloppy light booths that "appear" to be neutral and yet a small shift
in white point on a display is very obvious to a trained observer. With
respect to the concept of "chromatic adaptation" emissive displays are
completely different than a reflective environment.
In wide gamut situations, the primaries are necessarily narrow (if there are
only three). This will tend to increase differences between observers. I
am working on some papers in this area ; mostly review documents and
retrospectives to demonstrate how we got to where we are today and some
discussions of the relevance of sRGB in the brave new world of display
technology. The benefit of wide gamut displays can be masked by the
artifacts that narrow band viewing carry. For instance, it is more
difficult to the display white point with narrow band radiators. In high
ambient situations, it is nearly impossible to not have a visible cast in a
narrow band display using standard instrumentation to set white point.
Regards to all,
Tom
_______________________________________________
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