Re: Neutral grey under different lighting
Re: Neutral grey under different lighting
- Subject: Re: Neutral grey under different lighting
- From: Roger Breton <email@hidden>
- Date: Thu, 06 Sep 2007 17:30:51 -0400
Peter Morovic? Jan's brother? Wow! Glad to see you on the ColorSync List.
> When two spectrally different surfaces match under one set of viewing
> conditions (observer and illuminant*) they are potentially metameric.
Potentially? I'll read on...
> If
> then there exists another set of viewing conditions (illuminant or
> observer*) for which the match no longer holds, the two surfaces are
> strictly metameric.
A true "matameric pair" then?
> So, when two surfaces match under one light and
> mismatch under another we observe the effect of illuminant metamerism.
Ah! So far you have not list me. Thank's.
> If there is no illuminant or observer for which the two spectrally
> different surfaces mismatch, there is no metamerism and the two surfaces
> can be considered identical in terms of their color rendering properties
> as they induce identical color stimuli regardless of viewing conditions.
> [*not an exhaustive list of "viewing conditions"]
What are the conditions, in your opinion, that would make such a match
exist? Maybe you give an answer below, I'll read on...
> Instead, when a surface induces a certain color stimulus under one light
> source, but induces a distinctly different color stimulus under another
> light source, we observe color inconstancy.
Crystal clear.
> This is not surprising if
> you look at the vastly different spectra of common illumination, e.g. a
> tungsten filament based light source SPD and a fluorescent SPD. So, if
> some surfaces (a single surface) look neutral under one light source but
> exhibit a hue shift under another, this is color inconstancy, not
> metamerism.
OK. That much is clear to me. But in the same way we have "metameric
failure", would it be a terminological or conceptual abuse to coin the
expression "color inconstancy failure" to describe the apparent hue shift?
-- Just a crazy idea, I have not reviewed the litterature...
> While the two terms seem to be liberally interchanged and used as
> synonymous, they refer to fundamentally different phenomena caused by
> altogether different underlying mechanisms of the human visual system:
Again, crystal clear.
> Metamerism is a direct consequence of the fact that the human visual
> system only records a (lossy) compressed representation of the
> illuminant reflected off a surface,
A reduction of information, so to speak.
> i.e. the color signal: we only have
> three types of cones on our retina (RGB cameras are an analogy). So
> there are (infinitely) many surfaces that under a particular light give
> exactly the same XYZ tristimuli.
An infinite variety of surfaces under a particular light that could
potentially induce the same XYZ or visual sensation. I like that ;-)
> As to the usefulness or otherwise of
> metamerism: input devices (cameras, scanners) struggle with device
> metamerism, output devices (printers, monitors, projectors) cannot live
> without observer metamerism - color reproduction is intrinsically
> metameric (for now).
"Device" metamerism, Peter? Do you mean different spectras inducing the same
device RGB code values or response? Why, because of the choice of color
filters used with the CCD or CMOS?
> Color inconstancy instead is linked to the post-retinal (and most likely
> cortical) process of color constancy
Wow, that's a conceptual quantum leap.
> - a mechanism that helps us
> approximately discount the chromaticity of the viewing illuminant
Does color inconstancy has any conceptual links with chromatic adaptation
then? It would seem that way from your description of the phenomenon. But I
could be wrong or confused.
> (a
> mechanism shown to be a necessary condition for efficient object
> recognition for example). However it has also been shown to be
> imperfect.
There is research on this?
> So, a sheet of paper will be perceived as approximately
> "white" regardless of the illuminant (assuming some state of adaptation
> and the absence of a reference under other illuminants),
Interesting. This seems to merge with the idea of the of Markus where he was
arguing that he could "see" when a photographic gray *is* truly neutral.
> the key here
> being "approximately". So (again), if some surfaces look neutral under
> one light source but exhibit a hue shift under another, this is a
> failure of the mechanism of color constancy, or color inconstancy.
You would only call this a "failure of the mechanism of color constancy or
color inconstancy?
> Peter
Roger Breton
_______________________________________________
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