[opendtv] Re: Let the games begin

  • From: Tom Barry <trbarry@xxxxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Wed, 08 Nov 2006 21:58:36 -0500



John Willkie wrote:
IIRC, they're offering 16-20 linear video (each with two audio) program services,
plus 'clip channels' and a handful of audio and data services in a multiplex
that occupies a 6-mhz channel using 22 or so MBPS per mux.


Are you sure about that? 22 mbps (reliably) in a 6 MHz mobile channel would seem exceptional, and far far better than A-VSB could approach.

- Tom

In strong signal areas, the frame rate is 24 or 30 per second; otherwise, it's 
12 to 15 per second.  But, remember that those video channels at best are half 
the resolution of NTSC, but with image sharpening technology.

Remember: people don't watch technology, and they tend to seek quality when it 
is available and has a wide variety of content.  (This includes the betamax 
argument, since it wasn't widely available and had limited variety of content.)

John Willkie

-----Original Message-----

From: Tom Barry <trbarry@xxxxxxxxxxx>
Sent: Nov 8, 2006 5:22 PM
To: opendtv@xxxxxxxxxxxxx
Subject: [opendtv] Re: Let the games begin



John Willkie wrote:


What is confusing to you here, Tom? Qualcomm is offering (or will be

within a few months or a year or whenever) 1.5 mbps video mobile video channels 320x240. For $15 per month. It's highly likely that A-VSB services will (start out at least) as free to air. Guess which one will win out on that ground?

To me it's just bits, like Craig says. Let's call it bits/second/Hz, the spectral efficiency. That is, at a reasonable, typical, and equivalent (very bursty, non-random) error rate, how many bits per second can be delivered in how much channel bandwidth by the various methods? For instance, what size channel (in MHz) is Qualcomm using for that 1.5 mbps premium service?

Or how much would they take if they wanted 4.5 mbps like A-VSB? (close to enough to do scaled 540p HD just fine, competing with today's typical OTA quality)

- Tom



What is confusing to you here, Tom?  Qualcomm is offering (or will be within a 
few months or a year or whenever) 1.5 mbps video mobile video channels 320x240. 
 For $15 per month.  It's highly likely that A-VSB services will (start out at 
least) as free to air.  Guess which one will win out on that ground?

Have you ever considered the size of screens that will be most commonly used 
for mobile video?  Let's say for the time being that few to none will be 
1080x1920, or even 1/5th that size.  Probably for a very long time ...

I would think that A-VSB mobile video will be mutually exclusive (on the same 
transport) with HDTV, and probably will be mutually exclusive with data 
broadcasting.  (I can't say why at this point.)  It's just another arrow in the 
increasingly flexible ATSC quiver.

One also needs to consider the ability to offer an "associated small screen service" as 
hierarchical video. In other words, where some of the mobile video is usable in fixed locations to 
augment the main video service. These are application-layer issues: first the transport-layer 
issues need to be resolved, which are "probably" being addressed as I write.

The latest revision of the A/53d (dtv) and A/65c (PSIP) specifications already 
contemplate associated or free-standing video services.

John Willkie

-----Original Message-----


From: Tom Barry <trbarry@xxxxxxxxxxx>
Sent: Nov 8, 2006 8:17 AM
To: opendtv@xxxxxxxxxxxxx
Subject: [opendtv] Re: Let the games begin



Craig Birkmaier wrote:



Let the games begin.

Is A-VSB still in the game?

It appears that after FEC overhead in order to get mobile or good multi path reception it would only have 4-5 mbps of payload. Did I understand that correctly? If so then it obviously doesn't do anything for HDTV reception since with MPEG-2 that is certainly not enough for HD.

But I guess A-VSB is really only for mobile. Can anybody say what throughput can be expected for mobile reception on a 6 mhz channels using the various DVB or OFDM options? Is A-VSB competitive here?

I guess most telling (for 8-VSB HDTV) is the quote from (Samsung's?) Mike Simon about 17 minutes into the audio stream about possible receiver side improvements to 8VSB:

---- quote -----
... To give you just a flavor for what we have and what we can have with this proposal and why it does make a difference, graphically if you can look at this it's kind of intuitive. Here's the normal 8-VSB physical layer as I show on my left here, and it shows it's one data frame made up of 2 data fields and it has a PN511 and other PN sequences that occur in this data field sync area, but that only occurs every 24 milliseconds, so, if I want to be able to track anthing that's changing, it better be changing awfully slow. Or I'm not going to be able to track it and, uh, as good as blind equalization techniques have been, and I guess will even be, uh, you know, we don't, SAMSUNG DOES NOT BELIEVE THAT RALEIGH CHANNEL PERFORMANCE WILL EVER HAPPEN BY JUST RECEIVE-SIDE ONLY IMPROVEMENTS. [emphasis added]
---- /quote ---

Is Raleigh channel performance something we need in gen-x receivers to handle dynamic multi path properly?

- Tom




At 5:54 PM -0500 11/7/06, Allen Le Roy Limberg wrote:



In the EVSB, ASB, Philips and ETRI proposals the information content is
coded in such a way that it is not possible for legacy receivers to decode
it.  This means information content has to be transmitted twice if legacy
receivers are to receive it and new receivers are to receive it with full
robustness.  This is a major waste of code capability.

The original goal in the Request for Proposals from ATSC was to ruggedize
HDTV broadcast or multi-cast SDTV. The various proposals seem to focus on replacing HDTV with other services that halve or quarter code rate. Perhaps
this is a tacit concession to business opinion that HDTV broadcast and
multi-cast SDTV will disappear from terrestrial transmitter service.

Al


Thank you Al for beginning a discussion that I hope will help legitimize our community - by this I mean that we may be able to contribute solutions, rather than observing and analyzing the DTV transition. I believe that one objective of the OpenDTV community is to help turn the vast wasteland of Digital TV in the U.S.A., into a critical piece of the infrastructure for communications in the rest of this new century. By so doing, we may also contribute an acceleration of the DTV transition globally.

So here we are, a decade into the DTV non-transition in the U.S., and we are trying to invent ways to layer complexity onto a standard that was never designed to do many of the things that now seem important to the folks making business decisions about the future.

As a starting point for further discussion, I would suggest that everyone who wants to contribute, take a look at a recent presentation about the A-VSB proposals, at the Iowa DTV Symposium, put together by the "original" IPTV people, Iowa Public Television.

http://www.iptv.org/dtv/2006/media_pres.cfm?ses=5&id=7&Track=T

This page contains an Adobe flash file with the presentation slides, and a downloadable MP3 file with the audio from the session. It's almost like being there...

;-)

I have reasons to believe that the Chairman of one of the ATSC working groups working on and helping to test these proposals will be addressing some of the questions that Bert raised about the mobile tests in Buffalo. At least Mr. Aitken said he was working on a response when I talked with him Tuesday.

Both Al and Bert have focused some attention on the issue of the payload overhead that may be required to utilize the new tools that are being proposed. The presentation covers some of the trade-offs and provides an example of the overhead needed to support the training sequences and a robust 1.5 Mbps service. Mr. Aitken noted in our conversation, that they are just starting to test these tools in the real world, and it is still too early to know how to optimize their use and determine to actual overhead.

Needless to say, there is plenty here to sink your teeth into, and to generate serious discussions.

But these discussions cannot take place, without also considering the alternatives.

I have just completed a column on the Digital LPTV transition in the U.S. There were thousands of applications for digital licenses for Class A, LPTV and TV translators, and the FCC will begin to issue licenses next year. This industry has the unique opportunity to develop its own platform for DTV and to build SFNs from the ground up. They are questioning whether they should be saddled with legacy ATSC, help drive A-VSB deployment, or simply join Qualcomm and others, who are deploying OFDM solutions in former TV spectrum.

The FCC has not authorized the use of OFDM by these low power broadcasters, however, the airwaves will soon be carrying OFDM broadcasts that will be received by devices being used by U.S. consumers. One can build a very strong argument to let these licensees have the flexibility to choose the technical parameters of their infrastructure as Sinclair proposed for full power DTV broadcasters.

And there is still another alternative.

START OVER

There is plenty of precedent for this - most notably, NTSC.

As was the case when RCA brokered the NTSC deal, it need not take a long time to come up with a viable solution. So ultimately, this discussion should also consider how the tools that we have today could be used to create a digital terrestrial broadcast network that is:

Interoperable, Scalable and Extensible, but most important, economically viable and able to drive innovation and new services in the future.

Let the games begin.




----------------------------------------------------------------------
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.



--
Tom Barry                       trbarry@xxxxxxxxxxx     
Find my resume and video filters at www.trbarry.com


----------------------------------------------------------------------
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.





----------------------------------------------------------------------
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.



--
Tom Barry                       trbarry@xxxxxxxxxxx     
Find my resume and video filters at www.trbarry.com


----------------------------------------------------------------------
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.



----------------------------------------------------------------------
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.



--
Tom Barry                       trbarry@xxxxxxxxxxx     
Find my resume and video filters at www.trbarry.com


----------------------------------------------------------------------
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: