[mythtv-users] coax vs. composite

[Cory Papenfuss]

> Your question is really about the limitations of NTSC video, and not
> the type of cable. For NTSC video the vertical resolution is fixed at
> 480 lines. But horizontal resolution will depend primarily on the
> video source. This page
> (http://lyberty.com/encyc/articles/tech/vid_horizontal_resolution.html)
> claims broadcast signals have about 330 pixels horizontally. This page
> (http://members.aol.com/ajaynejr/vidres.htm) has some more info, and
> states that signals provide about 80 pixels of width for each MHz of
> bandwidth. As Brian said TV stations provide about 4.2MHz bandwidth
> which also gives about 330 pixels.
>
 	I hate adding to the pedantry, but the video resolution issue is 
even more often confused than "coax" connectors.  The first article is 
almost correct, but not quite.  In terms of old-school analog "lines of 
resolution," it *is* the number of vertical black and white pairs that can 
be discerned horizontally.  Where he's not correct is that the standard is 
to define it in terms of "lines of resolution per picture height."  In 
other words, one needs to count the vertical lines that can be discerned 
*in an area of the screen whos width equal to its height*.  Thus, the 330 
lines often touted for broadcast video only means the equivalent of 330 
pixels if the video format was a 1:1 ratio.  SDTV has a 4:3 aspect ratio, 
so a 330-"lines-of-resolution" signal requires 330*4/3 = 440 pixels 
horizontally to adequately capture.  That's why SVCDs are at 480x480 and 
why it's largely considered a good resolution for broadcast-quality.

 	To extend this, consider a DVD.  If played on a 4:3 TV, the 720 
pixels horizontally translate to 540 "lines of video resolution."  If the 
video encoded at 720x480 is intended to be played on a 16:9 widescreen TV, 
it only has 405 "lines of video resolution."  Yes it's confusing to have 
720x480 represent 540 "lines" in one case, or 405 "lines" in another... 
Think of it as stretching the pixels horizontally more across the 
widescreen... since you only count pixels across in a *SQUARE* that's as 
wide as the picture is high, you only count to 405 before you're as wide 
as the picture is tall on the widescreen.

 	Strictly speaking, there are other perception-based "sharpness" 
things like Kell factor that I'm not too familiar with.


 	As others have said, the quality of the cables rarely influences 
the capture quality much... despite what Monster and other snake-oil 
peddlers would have you believe.  When the signal is crap to begin with, 
putting it over a better cable doesn't make it any less crap.

-Cory

-- 

*************************************************************************
* Cory Papenfuss, Ph.D., PPSEL-IA                                       *
* Electrical Engineering                                                *
* Virginia Polytechnic Institute and State University                   *
*************************************************************************