Re: The color of gray
Re: The color of gray
- Subject: Re: The color of gray
- From: Mark Stegman <email@hidden>
- Date: Wed, 03 Sep 2014 00:19:30 +1000
Ben,
Thank you for generous and very informative response. I am aware of the
qualities of perception you refer to in the latter part of your response as
I have used them in my own teaching of 'elementary' colour and its
reproduction. As for the initial explanation on the 'mechanics' of
spectrophotmeters I am unsure if it is consistent with your original
statement.
Mark
On Tue, Sep 2, 2014 at 1:15 PM, Ben Goren <email@hidden> wrote:
> Excellent questions!
>
> On Sep 1, 2014, at 7:50 PM, Mark Stegman <email@hidden> wrote:
>
> > If the illuminant has a restricted, non uniform, spectrum how can the
> spectral response of the object be flat?
>
> The reflected spectrum for an ideal gray surface is exactly proportional
> to the illuminant, less a linear factor.
>
> "Flat" is a convenience of language. A reference spectrophotometer is
> going to have its own source of illumination with a known spectral
> response, typically something broadband and without emission or absorption
> spikes -- an incandescent bulb. The light is bounced off the sample and
> reflected through a diffraction grating to project a spectrum (rainbow) on
> a detector array. The instrument measures the intensities of the various
> wavelengths, compares them with the already-known (from the factory)
> standard of the instrument's light source, and computes and reports the
> difference between the two.
>
> So, the illuminant might have a spectral output of the typical blackbody
> radiation curve, possibly even with a low temperature. And, if you were to
> measure an ideally white surface, the _absolute_ values recorded by the
> instrument's sensor would be a perfect match for the illuminant used -- but
> that's *not* what the instrument reports. Instead, it reports how much less
> (or more, in the case of fluorescence) the sensors record than the
> illuminant emits.
>
> > How can an object reflect parts of the spectrum that are not present?
>
> It can't, which is why spectrophotometers are designed with broadband
> non-spiky light sources. But, in theory, so long as there's _some_ light
> from the whole measured portion of the spectrum, _any_ light source could
> be made to work equally well -- with the caveat that engineering
> practicalities (especially sensor noise) affect the quality of the results.
>
> > What colour is a 'grey' object under a red light?
>
> Ah -- now we're getting into psychology with a good dose of philosophy.
>
> "It depends."
>
> If you're in your living room at night and you tape the same red gel over
> all your lights, you're probably going to see the object as still being
> gray unless you examine it critically. If you return the room lighting to
> normal, set the gray object on a light trap and shine a tightly-focussed
> and hidden red flashlight on the object and ask somebody to step into the
> room and tell you what color it is, that person will tell you that it's red.
>
> There's an easy and mind-blowing experiment I now command you to do as
> soon as possible.
>
> You'll need a styrofoam cup and a digital camera that you can shoot RAW
> and develop on your computer.
>
> Put the cup in direct sunlight, ideally right at the crack of dawn or just
> before sunset (but any time will work, and it's a good idea to repeat the
> experiment at various times). Shoot a picture such that the cup's own
> shadow line is pointed right at you -- that is, so that you're at 90
> degrees to the Sun and, from your perspective, half the cup is in the
> sunlight and half the cup is in the shade. For best results, use the cup's
> own shadow for this, not a shadow cast by some other object.
>
> Now, back at the computer, use the eyedropper to click-set the white
> balance on different parts of the cup. Observe.
>
> Develop the picture with a few representative such white balances. Create
> a neutral gray square (R=G=B=118, or anything in that general range so long
> as all three are equal) on a separate layer and drag it around the scene in
> all developments.
>
> Finally, return to the same outdoors scene the next day, with the cup but
> without the camera. Observe!
>
> Cheers,
>
> b&
>
_______________________________________________
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