Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!purdue!gatech!prism!vsserv!loligo!pepke
From: pepke@loligo (Eric Pepke)
Newsgroups: comp.graphics
Subject: Re: Frame rate (was: Re: HDTV and ATV Glossary (TN32))
Message-ID: <145@vsserv.scri.fsu.edu>
Date: 31 Aug 89 21:11:57 GMT
References: <13130@well.UUCP> <17400008@hpfcdj.HP.COM> <470@telesoft.telesoft.com> <8034@ardent.UUCP>
Sender: news@vsserv.scri.fsu.edu
Reply-To: pepke@loligo.UUCP (Eric Pepke)
Organization: Supercomputer Computations Research Institute
Lines: 43

When you're talking about the difference between interlace and non-interlace
per se, it doesn't make any sense to let every other variable differ as well.
If you do, it's never clear just what you are comparing, and the result is
that you can never say anything meaningful about interlace.

Interlace was a kludge way back in the early days when emulsions were coarse
and amplifiers were slow to minimize perceived flicker GIVEN A CERTAIN
BANDWIDTH.  You don't need to worry about comb filters and aliasing when
you make your amplifiers out of tubes with bakellite bases and all your
images come from an image orthicon that is about as grainy as cream of
wheat.  Now, of course, you can build tiny solid-state current feedback 
amplifiers with transitions sharp enough to pierce the firmament.  However, 
bandwidth is still one of the most important numbers that you have to 
engineer around.  In CG terms it translates roughly to horizontal scan 
frequency times number of dots in a line and directly affects

1) Response time of the phosphor
2) The rate at which you have to grab memory
3) The quality of the fastest amplifiers in the entire loop
4) How long you can make your cable and what you have to make it out of

These are the MAJOR engineering decisions.  The question is, given the 
bandwidth you can handle, the potentially high vertical visual frequencies,
the resolution of the bits of phosphor, etc. is interlace itself going to
make the picture better or worse, or will it matter.  Numbers for existing
systems put on spec sheets by marketing departments have to do more with 
history than with anything else, and freeing one from these bad 40-year-old
decisions that made sense at the time is the whole point of HDTV in the first 
place.

So, when considering interlace vs. non-interlace, keep the bandwidth
constant, as well as the number of lines displayed and pixels per line.  
The kind of niggling about whether 30 Hz is frame rate or field rate is part 
of the reason that cps was abandoned for Hz, so Hz is obviously not a useful 
concept in this context and causes more trouble than it's worth.

Eric Pepke                                     INTERNET: pepke@gw.scri.fsu.edu
Supercomputer Computations Research Institute  MFENET:   pepke@fsu
Florida State University                       SPAN:     scri::pepke
Tallahassee, FL 32306-4052                     BITNET:   pepke@fsu

Disclaimer: My employers seldom even LISTEN to my opinions.
Meta-disclaimer: Any society that needs disclaimers has too many lawyers.