[Coco] For those of you following the RGB2VGA FPGA

rcrislip RCrislip at neo.rr.com
Mon Sep 29 23:00:42 EDT 2014


On Mon, 29 Sep 2014 10:40:46 -0400
Gene Heskett <gheskett at wdtv.com> wrote:

<snip>
> Absolutely Nick. The reason for that is that the color subcarrier out
> of the composite jack if installed, or the rf jack to feed a tv, is
> not properly interlaced.  The subcarrier frequency is generally ok,
> but the H line length is an even number of clock cycles, whereas its
> even + 1/2 in real NTSC. A tv with or w/o a comb filter to separate
> it again will have atrocious dot crawl because that lack of timing
> emphasizes it as opposed to trying to cancel it, AND will destroy the
> luminance sharpness in its attempts. So picking them off and shipping
> them out as RGB must be done.  
> 
> The only semi-workable alternative would be to translate to S-VHS,
> putting the luminance on one pin, and the color subcarrier modulated
> signal on another pin which never mixes the color into the
> luminance.  There are places in the coco1-2 where that can be done,
> and possibly in the coco3 too, under the modulator shielding of
> course. I have no clue if the signal levels are usable but most S-VHS
> circuitry had an automatic gain control based on the amplitude of the
> incoming color synch signal, usually called the burst, sitting on the
> back porch of the H synch signal.
> 
> Trying to cobble a color signal into the NTSC b&w signal was the
> biggest cobble and con job ever perpetrated on the american public.
> But was forced on us by the then limited bandwidth of a 6Mhz wide
> channel. Mathematically it should have worked well, and I have seen
> it work well, but the Conrac monitors that made it look good were
> also $6000 a copy.  In the consumer affordable products, the math
> used would have forced a whole new meaning into the phrase "fuzzy
> math".
> 
> There, I've said it, since NTSC, which paid me well for decades is 
> officially dead except for your old vhs and dvd players.
> 
> But, an internal adapter, picking off the signals and making them fit
> the S-VHS plug, sure seems like a worthwhile project to me.  For the
> coco3, a combiner to mix the rgb into luminance, possibly with a net
> positive gain for the blue might work to make the S-VHS luminance.
> That can be gotten from the RGB output transistors, or possibly from
> inside the modulator can.  The color then would be a connection to
> the color modulator chips output taken before it went any place
> else.  It should not be that hard technically, or expensive but would
> probably need someone handy with a soldering iron to hook it up. Even
> if a video speed opamp or 4 are needed, those have been available
> from the likes of TI for at least 10 years at less than a buck a copy
> prices.  And they'll run on 5 volts!
> 
> Gotta run, need to go skinny dipping since I spent the day yesterday 
> cutting and machining steel, and go take my Toy in to get some
> recalls taken care of.

<snip>


Didn't Chris Hawk have something like that? I pretty sure he did.


More information about the Coco mailing list