[Tig] another new monitor? (could be a sony-beater?)

[Bob Friesenhahn]

On Thu, 3 Jul 2008, glenn chan wrote:
> And then other technologies like 1-chip DLP and some (but not all)
> plasmas have weird artifacts (e.g. DLP rainbows).  So it raises the
> question... what should our monitoring practices be?  What should the

The DLP rainbow issue happens when the viewer moves his head. 
Otherwise the single-chip solution provides better image registration 
since no special alignment is required.  The "wobbulation" technique 
used with some DLPs is more objectionable for reference use since it 
introduces some jitter at the pixel level.

> standard be?
> IMO, a reference display
> [A] shouldn't have any overly objectionable artifacts... e.g. flesh
> tone color shifting, 1-chip DLP rainbows
> [B] let the user spot motion problems such as judder and flicker (e.g.
> flicker from bad fluorescent ballasts).  Also need to spot content
> that might introduce seizures (it's happened before).
> [C] Otherwise don't worry about motion reproduction differences
> between displays.

It seems to me that a reference display should display the provided 
content exactly.  It should not be over-clocking the video or 
including motion estimation logic to try to make the result seem 
smoother.  If the content is 24 frames per second, then the display 
should update cleanly at 24 frames per second. The reference display 
should not try to emulate a CRT or film projection since neither of 
these display the provided content exactly.

Bob
======================================
Bob Friesenhahn
bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/