In article ,
 "Michael J. Mahon"  wrote:

> John B. Matthews wrote:
> 
> > http://home.woh.rr.com/jbmatthews/a2/lores.html>
> > 
> > I have a clear memory of the NTSC video red being a dark shade, as 
> > Linards suggests. I have no idea why the Applesoft manual calls it 
> > "magenta," which I identify with a secondary color having sRGB 
> > components (0xFF00FF). Of course, the html rendering is itself a 
> > function of the monitor's calibration.
> 
> Color 1 has only one bit of four set, so its luminance value is
> 25% of white.  The hi-res colors (not white and black) all have
> two bits set, and so have a luminance value of 50%, so lo-res color
> 1 is definitely a "dark" shade.  (Note that 0xFF00FF has a luminance
> of 67%--very much a "light" color.

Ah, thank you, Michael. This sheds light (no pun:-) on Linards' 
derivation, which matches my recollection. I remember the "yellow" being 
more of a gold and the "brown" being a muddy green, but I'd swear there 
were two different grays. I'm curious why his table shows the two grays, 
[5] and [A], having the same sRGB 0x808080. Is that that the difference 
in luminance, also?

-- 
John B. Matthews
trashgod at gmail dot com
home dot woh dot rr dot com slash jbmatthews
--- SBBSecho 2.12-Win32