[opendtv] Re: All colors of rainbow in new display concept

Olivier Houot wrote:

> It seems to me that the colors are obtained by diffracting a
> white light source. So if you want a very precise color
> inside the rainbow, it will represent a very tiny fraction
> of the incident light. There would be a compromise between
> color accuracy and light intensity output. Even with a well
> thought out compromise, i wouldn't expect those displays to
> be very bright.

Good point. There's a lot of light energy being wasted in the projecion
of unwanted colors to the back of the display mask, i.e. away from the
hole assigned to that pixel. Interesting. What do the greens think about
that? Think of all that heat.

> Also there are many colors out there that are simply not in
> the rainbow (say brown, for example). How would you render
> those?

They do mention using two gratings per pixel, for just such cases. I
suppose a decent brown can be made out of two rainbow colors? Like
yellow-orange and red-orange?

> As for the MPEG encoding, that might not be such a big trouble:
> if you consider that points close to one another will likely be
> of a similar colour, except on a transition from one object to
> the next, and if the colour is varied by applying different
> analog potential to a flexible element, then the driving signal
> will be an analog signal with a reasonable degree of continuity,
> except in some places.

True, although this is still a different method of transmitting color,
compared with the current MPEG/JPEG schemes. But perhaps the problem
goes away if you look at Don Munsil's post. Use an expanded version of
the RGB (Y Cb Cr) scheme we already have to transmit the info. Then
convert to whatever is required by the display local to the display.
Ditto for the camera.

I think it's more fun to dream up a totally new scheme, but infinitely
more practical to "make do" with what we already have. C'est toujours la
meme chose, even though some on this site find this hard to swallow.

> Why couldn't that be encoded just like present day luminance
> or chrominance? Plus, there would be  only one signal instead of
> three components. I expect that would compress just as well.

I think philosophically, no matter how you slice it, if you provide more
information (in the form of more colors), you will require more bits.
All else being equal. Don did mention that even to use the currently
illegal Y, Cb, Cr to expand your color gamut, you'd need more bits
assigned to the MPEG block coefficients. Otherwise, the additional
colors would come at the expense of more coarsely defined colors. Which
sort of defeats the purpose.

Bert
 
 
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:

- Using the UNSUBSCRIBE command in your user configuration settings at 
FreeLists.org 

- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word 
unsubscribe in the subject line.

Other related posts: