[opendtv] Re: Some results - 1080p @ 60 is Next?

  • From: Craig Birkmaier <craig@xxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Thu, 17 May 2007 11:47:00 -0400

At 6:17 PM -0400 5/15/07, Manfredi, Albert E wrote:
Craig Birkmaier wrote:

 Second, for high action, 720@60P is superior to higher resolution
 formats due to the improved encoding for emission. This DOES NOT
 mean that the acquisition system should be limited to 1280 x 720
 - it means that it is beneficial to oversample in the camera to
 produce the highest quality 720P possible to feed to the emission
 encoder.

But that is a matter of channel bandwidth, codec algorithm, and CPU
processing power.

Channel bandwidth is fixed and there are many alternative applications looking for bits. You like multicasts, but this limits how much quality can be delivered for each program in the multiplex.

Codecs are improving, and this can help, If we do not reduce the bits allocated to a single program. But peak bit rate requirements will still peak beyond what the channel can deliver.

Increased CPU power is largely irrelevant, as we need all of it to implement the new codecs.

So an improvement in the codec and in processing power
will change the equation completely, even without changing the RF
standards. ATSC and DVB-T can both be updated, since both are layered
protocol standards. Changing the RF channels is much more difficult.

Agreed. So the logical conclusion is that these improvements will allow for improved delivered image quality. Unfortunately, this is only one of the potential ways in which the bits can be used. We are now facing the potential of allocating bits for HD and SD broadcasts, multicasts, and now mobile applications that chew up lots of bits just to improve the channel coding.

Do you really think anyone is going to use these bits to improve delivered image quality?


 I am amazed that so many people think we need to move to 1080@60P,
 just because a handful of cameras now exist to acquire it.

I'd test it out and see. This can be done immediately, within the
existing standards, using either 1080 at 30p or 24p, and then everyone
can make an informed decision. I think we'll see that at screen sizes
that will become more common, say above 42", the difference will be
perceptible. So you move to whatever is available at a reasonable price,
and fits in the bandwidth the broadcaster wants to dedicate to the
program.

I think you are basically wrong, at least with respect to average screen sizes. 720P is more than adequate for 98% of all of the consumer displays sold, and can actually deliver BETTER image quality to the really big screens in bandwidth constrained applications...like DTV.


 How are we going to increase the bandwidth of the entire chain
 Bert?

Don't need to. The bandwidth is already there in many modes. You can
transmit your American Idol or Survivor in full 1080 at 24p today, for
example. Or in theory, but probably not well in practice yet, you can
transmit 1080 at 60p using H.264 compression even inside a 20 Mb/s
channel. In time, as H.264 is tweaked more or even replaced by something
better, that would be a credible transmission mode.

Eventually you may be right, although you are not taking into consideration the peak bit rate requirements, of the market driven desire to deliver more stuff, not higher quality stuff.


 Our DTV standard was designed to deliver TV to the masses, NOT to
 videophiles. One can make a very strong argument that we do not
 even need HDTV to deliver high quality images to the AVERAGE sized
 consumer display.

Consumer sized displays are big these days. Like FM radio, the
transmission channel is used by a wide spectrum of receivers, from the
cheapest to the most elaborate in-home and in-car hifi systems.

True. You keep talking about a bell curve for displays. The center of that Bell is STILL less than 30 inches; in five years it might be 36." Even if it eventually gets up to 50", you have already proven that 720P is adequate.

Delivering 1080P to 100 inch displays IS NOT an application for OTA DTV.


It's a bell curve. You want to grab listeners and viewers to include the
portion of the curve that tails off to the right. And that right side
keeps moving further to the right. What seemed unnecessary in the 1990s,
when DVB-T opposed HDTV altogether, is no longer unnecessary. This will
probably happen within the HDTV modes in the near future, as everything
goes upscale.

You, the DVB folks understood how important it was to ride the practical curves for consumer electronics, balancing quality and quantity against the average size of available displays. They were also pragmatic about improving ALL delivered image quality, not just a few programs. As a result, we still have to put up with more than 90% of all programming in the U.S. being delivered at NTSC quality or below, and only a handful of programs being delivered at HDTV quality (but mostly below due to the compression artifacts).

The interest in HDTV is growing world wide. It already IS an important application in the U.S., and other countries are starting to get on board. But the interest in mobile and portable video is growing even faster. And the desire to deliver everything on demand is growing fast too.

This is not a monolithic problem - it has many dimensions and many potential business models. I would point out that broadcasters in the U.S. have not made an extra dime on HD in terms of advertising to date. Some may be using their HD programming as a bargaining chip in retransmission consent agreements, which do drive money to the bottom line.

We are just now deploying a significant infrastructure to produce HD content as defined in the last century. Do you really expect the networks and content producers to upgrade this infrastructure to deliver 1080@60P?


I think lots of people staunchly opposed HDTV in the early days of ATSC,
using arguments on screen sizes and costs. No one "needs" HDTV, was the
mantra. But that was proven false. Screens of 50" or greater won't be so
rare anymore. I viewed a DVD on a 108" screen at a video store, and was
amazed to see it truly doesn't look so hot. It was a rather dramatic
demonstration of why we need HD DVDs (generically speaking). So, stuff
changes.

What is false is your statement.

There was never much question about delivering HDTV - in fact, it was the Trojan Horse behind the entire U.S. DTV standards process. The ability to deliver SDTV was what was added at the 11th hour.

The consumer electronics industry literally paid for the U.S. DTV standards process because they needed a market in which they sell more higher profit big screens. Despite this they crippled the standard with artifacts of CRT-based display technologies including interlace. I can assure you that we were not opposed to HDTV in the early days of ATSC - what we wanted was a system that could deliver high quality content at many resolutions, especially those that were most likely to be exploited by affordable consumer display technologies.

We were right about everything we predicted and what would be important. Hindsight is 20/20.

Resolution requirements scale with screen size and viewing distances. We defined this completely in 1992. Nice that you finally got a chance to see what we were talking about at the video store.

Regards
Craig


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