[BC] CBS Color...lived on

[Richard Fry]

RRSounds:
>The "330" you quote is for *horizontal* luminance resolution, which
>is where the first NTSC compromise occurred. Horizontal resolution
>IS bandwidth-dependent. While they aren't really 'lines' but rather the
>maximum number of black-to-white (or vice-versa) transitions per
>horizontal line, 330 'lines' is indeed the typical maximum in the
>luminance channel for an NTSC broadcast transmission.

But that luminance resolution is unchanged from the previous monochrome 
standard, so it is not an "NTSC compromise."

Peter Haas:
>The lower sideband is not bandwith limited, wereas the upper sideband
>is bandwidth limited so as to accommodate the FM audio, which is
>4.5 MHz above the AM video carrier.

If you're talking about the total modulation of the visual carrier, its 
lower sideband is ~limited to  Fv  -1.25 MHz, and its upper sideband is 
~limited to Fv + 4.2 MHz..

If you're talking about the upper and lower sidebands of the color 
subcarrier, they have ~ equal bandwidths for the Q component.  The lower 
sideband of the I component extends to about -1.5 MHz , and its upper 
sideband to about + 0.6 MHz, referenced to the color subcarrier frequency.

Burt Weiner:
>The visual carrier is Vestigial and, as I recall, must be attenuated
>by 40 dB or more at -.1.25 MHz, the lower edge of the channel.  The
>upper sideband has to protect the color subcarrier's lower sideband
>which extends approximately -600 kHz plus guard band (down) from the
>color subcarrier itself and as a result leaves only about 3.2 MHz for
>the luminance channel.

As in the monochrome standard, NTSC luminance is transmitted with ~4.2 MHz 
of video bandwidth.  Chrominance and luminance are "frequency interleaved." 
A modern TV set with a comb filter can remove most of the color subcarrier 
from the luminance channel, leaving the full ~4.2 MHz monochrome video 
signal to drive the display device.

Dana Puopolo:
>If you've never seen it before, you would be amazed at how much
>quality is lost going through the average NTSC transmitter.

The major difference being that the resolution of the luminance channel can 
be greater than ~4.2 MHz as the composite color signal exits the video 
switcher.   I have set up cameras that showed nearly 800 lines of 
horizontal resolution on a test chart (nearly 10 MHz of video bandwidth). 
Of course, your comment applies to the monochrome television standard as 
well as to NTSC.

Otherwise, the composite output of a color camera already has band-limited 
the I & Q signals in the chroma subcarrier to essentially what is going to 
pass through the tx, so color performance won't be greatly different 
before/after transmission.

There is an irony in NTSC in that, while many complain about its design 
limits, most TV sets don't use the full bandwidth of I video, which would 
increase color detail in the display.  Most receivers filter the lower 
sideband of the I subcarrier to make it ~equal to its upper sideband, which 
makes for simpler/cheaper circuits than when dealing with those dissimilar 
bandwidths.

RF