[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: DL Color Space

In a message dated 98-01-09 07:25:12 EST, Dave Satin wrote:

<< In my post, I said that SGI had fixed the color space problem internal to
the DIVO and GIVO hardware. >>

In the most recent Silicon Graphics "Pipeline" (their bi-monthly tech
publication), SGI did a piece on DIVO and GIVO.  Interesting mainly because it
appears to be way behind the times.  Among the several points I noted, they're
still calling what we all know as ITU-R BT601-4 "CCIR 601 digital video," and
it's been more than a year since the old CCIR went defunct and the title of
the standard changed!

I asked one prominent non-linear online system manufacturer whose products run
on SGI hardware why they don't get around the color space issues by using
digital color space conversion.  Sierra Design Labs makes a standalone box
that they claim does just that.  The online system company told me they don't
see the need for any such foolishness.  Yet, color space conversion is one of
the big differences between the DIVO board and Sirius.  Evidently there is
still a lot of confusion in the computer industry about the handling of
television-style video, and the science still has some maturing to do.

One of the things the computer industry is going to have to come to grips with
sooner or later is that Y/Cr/Cb video came into being for a very good reason.
Since the human eye doesn't see a great deal of detail in color, it is
sufficient--as far as the typical 25" TV set in the typical American living
room is concerned--to let the luminance channel (Y) carry the detail and use
lower bandwidths for Cr and Cb.  If we couldn't pull this trick, we'd have to
use 30% more tape to record RGB, and 30% more transmission bandwidth to get it

Now the FCC's vision of the future of broadcasting is that every radio and
television station will eventually become a "bit blower" (my term).  They will
throw data packets into the air that can be for absolutely anything--audio and
video programming, paging systems, data relays, whatever.  Different types of
receivers will only decode the packets they are meant to receive.  Even after
compression, RGB still takes more packets to transmit than Y/Cr/Cb, which
means that there will be a strong economic incentive to stick with Y/Cr/Cb and
not switch to RGB.  Regardless of what Bill Gates says!  

So it's not just a color space problem that SGI, DL, and the other computer
companies need to resolve, it's also going to end up being a "color-cash" one!

A happy weekend to all~
Christopher Bacon

Thanks to Jan Janowski for support of the TIG in 1998
TIG subscriber count is 910 on Fri Jan  9 07:35:54 PST 1998
complete information on the TIG website http://www.alegria.com/tig3/