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

colorspace: the vile frontier

I think that a lot can be done to more properly handle color-
space conversion.  I think the biggest and most obvious offender
is the Sirius board, hence all the Flame problems.  There are
other ways to help the problem, but apparently none can be done
in the Sirius, so those who use it are stuck for now.  Perhaps
wider word widths, and numerical processing that can accomodate
"negative colors" might be an area of improvement?  Software
conversions instead of fixed hardware? Better understanding of the
YCrCb space by computer folk who traditionally have only had to
deal with RGB?

Also, it seems that some processes are better suited for one color
space or the other. For example, it would seem that compositing
and chroma keying are well suited for YCrCb, whereas lighting and
shading models probably have to be in RGB space.  I wonder if future
software might be able to work in both?

But for now we recommend using the Davinci RGB Legals, and we also
follow the chain with the HP QA-100 limiters.

On Wed, 19 Feb 1997 12:13:30 -0500  Bob Festa wrote:

>At 09:23 PM 2/18/97 -0800, sflippi9 at IDT.NET wrote:
>> If a spot is leaving Xfer and moving on
>>to say Flame, is the colorist trying to conform to RGB legal limits.

>Once bitten...twice running with daVinci legal color limiting enabled.
>Do yourself a favor, make all your recordings "flame retardent".
>Bob Festa

Bill Topazio, VP/Engineering     bt at mte.com   v:212-907-1225
Manhattan Transfer / Edit        New York, NY f:212-687-2719

"The universe is a big place, perhaps the biggest."
                                   -Kilgore Trout