From: Philip Nasadowski
Newsgroups: alt.folklore.computers,sci.electronics.design
Subject: Re: Digital divide by ten, 1949 style
Date: Sun, 22 Sep 2002 21:59:22 -0400
Organization: Biker/metalhead from hell!!!
Message-ID:
References:
NNTP-Posting-Date: 23 Sep 2002 02:02:50 GMT
X-Harley: '99 FXDWG
User-Agent: MT-NewsWatcher/3.1 (PPC)
In article ,
Brian Inglis wrote:
> It's been many years, but I found few similarities between PAL
> and NTSC colour encoding when I first came over here and looked
> at the dufferences -- colour appeared to have been designed into
> the PAL system, whereas it appeared to be a kludge added onto the
> side of the NTSC system.
Well, NTSC *was* a kludge designed into an existing system. Albeit,
it's a very clever one, and pretty cool.
http://www.ee.washington.edu/conselec/CE/kuhn/ntsc/95x4.htm
Gives a broad overview of NTSC, but goes on to say PAL is much like NTSC
save for minor changes.
Naturally, if the compatability issues didn't exist, NTSC could have
been a lot more boring. But being backwards compatible with B&W TV was
needed, given the rate of TV sales, the cost of color TVs, and the lack
of programming at the time. A few different ideas were tried before I/Q
modulation came along anyway.
Actually, NTSC really isn't that hard save for two minor things - the
fact that the color subcarrier is supressed, and the sidebands are
somewhat lopsided (limited on one side to prevent hitting the audio),
plus, Q is a narrower bandwidth than I. The result of the latter is
NTSC effective transmitts full color for large areas, Orange-cyan for
medium areas, and detail in B&W. Few TVs even make use of the Q axis
infor, given that narrowband X-Z demod is stanard these days (or was
until recently, though RCA played around with all sorts of demod systems
in the 50's).
Anyway, a color TV is a B&W TV with a bandpass amp, a setup for
detecting the colorburst and syncing a local oscillator to the burst,
and 2 mixers (well, they're called demodulators, but they're more like
mixers IMHO) that mix the bandpass filtered signal, one in phase with
the osc, the other 90 degrees out. The result is B-Y and R-Y info, from
which G-Y is extracted.
If anyone's curious, ask and I'll post a schematic of the chroma section
from my '65 RCA CTC-16...
--
To email me, change 'usermale' to 'usermail'