View Single Post
  #12  
Old 09-18-2016, 12:25 PM
old_tv_nut's Avatar
old_tv_nut old_tv_nut is offline
See yourself on Color TV!
 
Join Date: Jul 2004
Location: Rancho Sahuarita
Posts: 7,184
Quote:
Originally Posted by Outland View Post
...Why setup was invented, I'm not really sure. It doesn't seem like it does much if at all.
There were a couple of things added to NTSC standards that turned out to be unneccesary and eventually a detriment.

Setup was one. It was intended to prevent clipping of lowlights by inaccurate black level clamps. Professional gear quickly migrated to accurate back-porch clamping, and TV sets used none at all at first. The setup level was inherently a problem in accurate setting because there is no natural reference for it in a video signal, as there is for zero %. It really got bad when cable TV came in, because now there were thousands of video proc amps in cable systems, each with a different error in setup. As receivers began to be designed with better DC restoration, they just passsed through the errors. In the 70's the TV makers and broadcasters mounted a campaign to get all the broadcasters in a given market to monitor their setup and see that a match was maintained. The big cable operators also participated, but of course it was much more difficult for them to keep all those proc amps under control. The industry made a mistake by not petitioning the FCC in the 1950s to change setup to zero. The adoption of the color standard would have been a great time to do it.

Another area of unnecessary NTSC signal features was the sync equalizing pulses. The idea was to send H sync pulses at twice normal rate during vertical blanking, so the vertical sync separators in receivers would not get any horizontal rate energy, which would cause line pairing (poor interlace) by triggering vertical retrace on whole lines only instead of half lines (there are 262.5 lines per field, for 525 lines per frame). In fact, vertical serrations at twice horizontal rate would have been sufficient, and the main reason for poor interlace in receivers was all the horizontal sweep current running around the chassis and getting into the V sync circuit. The equalizing pulses were specified on the idea that H sync would be edge triggered on the leading edge of horizontal pulses or equalizing pulses- a really bad idea for noisy over-the air signals. All receivers used RC low pass filtering of the H sync to reduce noise, resulting in H sync being related more to the center of the H sync and equalizing pulses. The serrations in the wide V sync pulse were also timed based on edge detection for horizontal. The result was that all receivers had a slight pull of horizontal phase toward one side during vertical retrace. Fortunately, this was minor and not a real problem, but it caused Motorola to include an explanatory note in their data sheet for TV sync chips, since engineers, by the 70s, didn't know about this NTSC quirk, and would think the chip was at fault. The NTSC spec included both pre- and post- equalizing pulses (before and after the V sync), which is mathematically correct if you build a circuit that looks at the whole interval to determine the position of V sync. But, of course, the simple V integrators in TVs (until digital countdown ICs were implemented) had no such memory, and the post equalizing pulses were a waste, as they could only extremely weakly affect the following V sync a whole field later.
__________________
www.bretl.com
Old TV literature, New York World's Fair, and other miscellany
Reply With Quote