Re: Explanation of Component Video

[bieling@terra.es]

On May 9, 7:50 pm, "Michael J. Mahon" <mjma...@aol.com> wrote:
> biel...@terra.es wrote:
> > On May 7, 11:39 pm, "William Garber" <willy4...@garberstreet.com>
> > wrote:
>
> >>For those of you, as I, who were having trouble
> >>with this, here is a site that really does it a
> >>good turn. Have a read.
>
> >>http://www.projectorcentral.com/component.htm
>
> > For this to be true :
>
> > "Once we've got our video information packaged up in component video
> > format we've reduced bandwidth requirements by a factor of 3 to 2."
>
> > Even though it's not mentioned in the article, going from RGB to
> > YPbPr, the video image has to be chroma-subsampled, somehow "NTSC-
> > izing" it, although to a lesser degree.
>
> Strictly speaking, the chroma signals don't *have* to be bandwidth-
> reduced, but in the video transmission/recording context for which
> component video was designed, they always are.

You're right, I know. I wanted to note that if they weren't
(subsampled), the bandwidth wouldn't be "reduced by a factor of 3 to
2" simply by encoding as Y+Pb+Pr.

> This is a consequence of the reduced "chroma" resolution of the
> human eye.  Our sense of "sharpness" comes from the luminance
> component, and we are broadly tolerant of reduced chroma sharpness.
> Clearly, computer graphics can lead to very different requirements--
> like a narrow font that has the same luminance as its background.
>

Yep. The NTSC color encoding relied (cleverly!) on this fact.

--Jorge.