🚀 go-pugleaf

In article <1dcbcuINNj5s@smaug.West.Sun.COM> Richard.Mathews@West.Sun.COM (Richard M. Mathews) writes:
>music@erich.triumf.ca (FRED W. BACH) writes:
>>     Yes, and speaking of color tv, since my favorite color family is 
>> the violets, it always bugs me when true violets (not blue+red phony
>> attempts at it) are picked up by a color-tv camera.  Now I don't
>> know what *every* camera out there does, but some will simply *not*
>> pick up this violet and display it.  All you get is a pure bright
>> blue.  Is this because the true wavelength of the violet is outside
>> (i.e., not between) the spectral responses of the phosphors used?
>
>I would think that you get a pure blue because the camera's red detector
>does not have the same small peak in the violet that our eye's red cones
>do.  Only the camera's blue detector responds to the violet light, so the
>camera produces a blue image.

Is your reasoning here that the eye distinguishes violet from blue by
the red-sensitive cones responding more strongly to violet than blue?
I don't think this reasoning is correct.  The eye's blue-sensitive
cones have their main peak at 430 nm, well into the violet.  The eye
recognizes the transition from violet at 430 nm to blue at 480 nm by
the red- and green-sensitive cones responding to blue *more* strongly
than to violet, and by the blue-sensitive cones responding less
strongly (it's downhill from their main peak at 430).

The main peak of green is at 530 nm, and red at 560.  At 430 nm the red
output is a mere 10% of what it is at its 560 peak.

There is indeed a violet peak, called the cis peak (I think Cis is
German for C sharp, suggestive of a tweak), present in all three cone
types.  This gives all three a small violet boost, amounting to more of
a tweak than a real peak that just attenuates the roll-off.  I believe
its role is not yet understood.

The retina filters out UV.  UV ordinarily looks like faint violet, but
that's real violet you're seeing.  People who have had their retina
removed perceive UV not as violet but as blue.  Not much is known yet
about UV sensitivity of individual cones so we can only guess why, but
plausibly it is due to the blue response decreasing more than the red
and green responses for UV rather than some unexpected peak (F sharp?)
in the latter responses.

These numbers are based on measurements done at Stanford, in Denis
Baylor's lab, of spectral sensitivity of single cones, which are
micropipetted up and their electrical sensitivity measured.  (No I
didn't volunteer :-)  The above numbers are the most recent from
Baylor's lab.  Wyszecki and Stiles "Color Science" is the "bible" in
more than one sense: the 2nd edition (1982) reports (p.93) that visual
pigments have only been extracted from rods, although of that date
there was no shortage of indirect evidence for their presence in cones
from work done in the 1970's.  Extraction of cone pigments is a
relatively new, tricky, and exciting business.

>Without a TV which includes a violet phosphor, there is no way that any
>camera can make your TV show violet.

Yes, indeed.  See my post of last week describing the gamut of a color
CRT as a triangle in the interior of the CIE chromaticity diagram.
Anyone care to hold forth on the main obstacles to a good violet
phosphor for color CRT's?  Seems like a great opportunity to make a
buck or two!
-- 
Vaughan Pratt				There's no truth in logic, son.