[opendtv] Re: Demand for free DTV rising in Australia
- From: "Manfredi, Albert E" <albert.e.manfredi@xxxxxxxxxx>
- To: <opendtv@xxxxxxxxxxxxx>
- Date: Sun, 24 Jun 2007 18:31:27 -0400
Barry Wilkins wrote:
> I agree that the course is obviously set for the US and
> you should make the most of it. However, the suggested
> reason for the 8-VSB path choice/COFDM rejection being
> a promise to the ill informed of better things just
> around the corner has a certain lack of depth to it.
It did happen, however. The improvement was indeed just around the
corner, in the sense that it was demoed in March of 2002 (in a
lab-configured rack). It appeared in the very first integrated sets in
late 2003, and it FINALLY appeared in STBs in early 2007.
> The European assessment of the time rejected 8-VSB for
> the very reasons that caused major problems with early
> US receivers and is still an inherent weakness of 8-VSB
> versus COFDM, namely dynamic multipath. This must have
> been a flag of warning to the US. Suggestions that the
> USA had a very specific population density and Tx site
> characteristic unrelated to anywhere else where COFDM
> was chosen also appears a weak argument.
To me, it was always a defensive-sounding and ridiculous argument. The
real argument was, or should have been, that there are tradeoffs, and
sticking with the single-carrier scheme, *IF* the promises of better
performance were true, would have its own advantages. Which are, the
lower C/N margin for any given spectral efficiency, and the lower
peak-to-average ratio, which makes for a really good-looking spectrum.
:) As well as reducing co-channel interference.
COFDM has the advantages of, still today (potentially), greater echo
tolerance, and simpler signal processing for good robustness. And thanks
to the great echo tolerance potential, one can more easily implement
SFNs. *BUT*, confound it, show me someone who waxes poetic about SFNs,
and I'll show you an uninformed person.
> No alternative standards in development such as ISDB-T
> used n-VSB modulation.
Actually, the Chinese standard does. It also takes great liberties with
its COFDM modes, by depending *not* on a GI, but rather on the same
exact training sequence it uses in its single-carrier mode, for
multipath distortion correction.
And the same issue was being debated in ADSL circles, multitone vs QAM,
so it's not purely something in the TV community. Signal processing vs
naturally robust modulation scheme. The faster and cheaper chips become,
the more you can opt for fancy processing. (Does anyone use COFDM for
space communications?)
> We have already decided to go with DAB (Eureka 147) as
> mentioned elsewhere. But if there is a superior standard
> out there I want to know all about it and expect that
> because there is still time, the decision makers will
> reconsider any and all developments.
Everything is a tradeoff. Let me ask you a really simple question: why
not stay in the FM for radio, and use one single digital radio standard
for the AM, FM, and the SW bands? DRM would allow that, and IBOC "as is"
would allow that for AM and FM bands. No technical reason why IBOC
couldn't be applied to SW too, btw.
I'm not pushing IBOC at all, though. It has the disadvantage of making
digital very low power, until analog is shut off, therefore not very
robust. But it sure makes the transition easy. But why not DRM? Do you
folks prefer radio to go to the high VHF frequencies or up in the L
band? I don't see any advantage to that. Over here, those upper VHF
frequencies aren't available. And I fail to see any advantage to going
way up in the L band for radio, which is supposed to be very easy to
receive.
> The 8-VSB decision is a bit like NZ going at this stage
> with DAB using the old MP2 codec format. There is only
> one argument in favour of that and it is simply that
> there are large quantities of receivers already out there.
> Everything else is a negative.
Only because no one will do a decent set of updated comparison tests.
But I'm not saying that picking DVB-T is wrong at all. In many ways, I
wish they had made the switch in 2001. But I also don't mind seeing cool
new work to make 8-VSB robust.
Bert
----------------------------------------------------------------------
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:
