Re: video
- To: Lonnie Reed <Lonnie.Reed@xxxxxxxxxxx>
- Subject: Re: video
- From: Soeren Tiedemann <tiedema@xxxxxxxxxxxxxxxxxxxx>
- Date: Wed, 17 Mar 1999 11:22:32 +0100 (MEZ)
- cc: misc
- In-Reply-To: <199903170034.QAA15822@tellurian.Eng.Sun.COM>
Hi Lonnie,
> Since f21 only puts out 16 colors right now, and since you have multiple
> f21's could you have one say "dedicated" to blue, anther to red and
> a third to green. Then could you mix the video output to produce mixed
> colors? Then it would seem you could produce many more colors. Would the
> F21's have to be syncronized (maybe run off the same video clock)? Can
> you just mix video at all?
Jeff himself mentioned something like this. One F21 for each color signal.
It should be possible without too much problems to get 800*600 with 24
color-bits. Using all the I/O-power should even lead to higher
resolutions. Let's calculate a bit ...
> Basically is it nontrivial (doesn't take a lot of additional cktry) to
> mix video signals? And is RBG easier to do this with than NTSC?
I would prefer RGB because there are some different TV standards. At the
moment I'm trying to solve the puzzle how to change MuP21-NTSC video
signal into the German PAL standard. It's possible but not easy.
Perhaps I have to use a different video-clock-cycle ( Chuck uses
4*3.58=14.32MHz to provide 455 cycles 140ns each, the PAL-norm is
4*4.43=17.72MHz ). In easy words, converting NTSC to PAL requires not only
to increase the rows ( 525 -> 625 ) but also to decrease the 30Hz image to
25Hz. It's a nice puzzle, because even in the best case I have to rewrite
most of Chuck's video driver code. To avoid such converting problems I
would suggest to use RGB, it should be more suitabale for graphics.
Soeren <tiedema@mail.uni-freiburg.de>