[Coco] Documentation of "complex" artifacting (i.e. more than black/white/blue/red)?

[John W. Linville]

On Sat, Sep 25, 2010 at 06:01:38PM -0500, john dumas wrote:
>  On 9/25/2010 5:16 PM, Robert Gault wrote:
> >John W. Linville wrote:
> >>On Fri, May 21, 2010 at 05:05:41PM -0400, John W. Linville wrote:
> >>>Are there any articles in Rainbow or similar CoCo-oriented publications
> >>>that documented use of "complex" artifacting for producing additional
> >>>colors (e.g. purple, green, etc) with a CoCo?  Can anyone refer me
> >>>to one?
> >>
> >>I suppose it is bad form to reply to oneself, but at least if I
> >>reply in this thread then I am more likely to be able to find this
> >>information when I need it again. :-)
> >>
> >
> >You need to understand how the NTSC system works to understand
> >color artifacts. A good place to start is here:
> >http://en.wikipedia.org/wiki/Color_Graphics_Adapter

Thanks, Robert.  But while that helps explain the 01 and 10 patterns
(which fit into the time slot for a single NTSC color pixel), I don't
think it addresses the patterns involving more than two bits.

> Excellent explanation.  It was so long ago that memory is dim, but I
> think the none of us were aware of the possibility of artifact color
> when the VDG was designed. Certainly I was not. Terry might have
> been when he worked on the spec since he knew more about NTSC
> than others in the design group. But I suspect not.
> 
> As I recall, we were all surprised to see the artifacts with the first
> silicon. But, some 20/20 hindsight pointed to the single pixel patterns
> being seen by the receiver as chroma data and compared to the color
> burst, producing and artifact color. Two colors actually since there
> was a flip-flop uninitialized at reset. [That's why I think Terry didn't
> suspect the "problem" - he's too swift to let something go by if he
> knows there's a problem.]
> 
> As far as John's question about multi-bit patterns, I can only guess.
> [Been way, way, way too long ago!] My guess would be that those
> patterns have enough energy in the first harmonic for - some -
> receivers to treat the harmonic as chroma data. The resultant
> differences in the pattern, compared to "real" data might have
> a different phase relationship to the color burst and explain
> the different color(s).

Hmmm...well, that seems at least initially plausible to me.
Thanks! :-)

> If any of this is actually true, I would expect different results from
> different receivers - particularly old vs. new ones...
> 
> Perhaps someone who has current in depth knowledge of NTSC
> could say if that's correct or false................

FWIW, my earlier experimentation with another program that claimed
to produce the "complex" artifact colors showed only the "simple"
colors on an LCD screen and only the faintest hint of the purple/green
"complex" colors on a mid-90's vintage Sony TV.

John
-- 
John W. Linville                Someday the world will need a hero, and you
linville at tuxdriver.com                  might be all we have.  Be ready.