[opendtv] The "real" problem with OFDM in the U.S.

  • From: "John Willkie" <JohnWillkie@xxxxxxxxxx>
  • To: <opendtv@xxxxxxxxxxxxx>
  • Date: Sun, 6 Mar 2005 12:35:04 -0800

Several weeks back, mono-note Dermot, while continuing his consistent praise of 
OFDM modulation over VSB, spilled the beans on why OFDM was inappropriate as a 
replacement for NTSC in the U.S.
I don't know if my friend Dermot fully realized the upshot of his comments, and 
I suspect that only a few list members caught the import of his comments.  It 
took me a week or two of pondering to realize the implications.

What the heck am I talking about?  Well, we were discussing probablity eg, 
f(50,50) of reception with various schemes.  Rember, in this context, that 
8-VSB is supposed to be the equivalent of NTSC, not the better of it.

Dermot bragged how goold fashioned COFDM routinely offered F(50,90) reception 
probability over entire coverage areas, and with -H, it was F(50,99) (or was it 
F(99,99)?

These figures sound great.  For the uninitiated, that means that at fifty 
percent of the reception locations, at least 90% of the time, a sufficient 
signal level was received.  (Analog also imposes a grade of service figure 
before these reception levels are relevant, but I'll table that here.)

Unfortunately, as any any good and many bad engineers know, NTSC reception 
specifies a contour signal level and probability level.  Unfortunate because 
the best (at the city-grade contour) that NTSC has to deliver is F(50,80).  At 
the grade A and grade B points, the probability is F(50,50).  And, the 
inferference contour is F(50,10), wich a desired to undesired dB figure 
specified.

Upshot:  for all it's superiority, permitting COFDM would have provided a 
superior signal level to that of NTSC, and -- aside from narrow technical 
modulation issues -- would have unwound the carelfully (politically) negotiated 
replacement of NTSC with an equivalent service level from 8-VSB.  To permit 
COFDM, broadcasters would have had to pay a heavy price (or prices).

At least, back then.  I don't recall anybody offering to pay the government 
money to permit COFDM operations.  Might make sense now, if it's so superior.

John ("i'm glad to see Dermot supporting my 'political' position on COFDM in 
the U.S. with empirical data) Willkie

P.S.:  I saw a posting requesting information about the origins of 30-foot 
reception criteria.  I'm suprised M. Schubin didn't respond; the 30-foot rule, 
as I've pointed out here several times) comes from the FCC's R-6602 
(Longley-Rice) study, which, if memory serves me correctly, was first published 
in 1946 or so.

P.P.S.  I've also seen postings about Ibiquity/IBOC.  Sounds great, an engineer 
of long tenure in the industry told me, except for it to work, Shannon's law 
has to be repealed, and he pointed out, the folks at Ibiquity are "real pearls" 
to work with.










 
 
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:

- Using the UNSUBSCRIBE command in your user configuration settings at 
FreeLists.org 

- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word 
unsubscribe in the subject line.

Other related posts: