[opendtv] Re: Finally anamorphically compressed 480i

  • From: Craig Birkmaier <craig@xxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Sun, 23 Jan 2011 09:54:39 -0500

At 6:53 AM -0500 1/22/11, John Shutt wrote:

The problem with making all SD 16:9 is that absent AFD there is no way for a
converter card to detect the difference between 480i 4:3 and 480i 16:9.  So
when do we pillar box and when do we not?  At best it could only detect
letterboxed SD by detecting the black bars.  But a full screen of pixels is
a full screen of pixels to a card who cannot tell when Oprah is being overly
stretched horizontally.
I think this is overstated. If you are starting with an MPEG-2 transport stream, the metadata is there to instruct a decoder what to do. If, as I expect is the case, John is talking about baseband source, as it may arrive from a programmer, then it may be more difficult to determine whether the source is 4:3 or 16:9. But a human can do this when the check the asset into the stations system. One could argue that a satellite feed may carry either 4:3 or 1`6:9 SD, but most of the stuff being delivered by satellite these days is a compressed MPEG-2 transport stream with metadata. 

So what am I missing here John?



I suppose we could make the same assumption that we make when upconverting
SD material to HD, that absent AFD, SD is assumed to be 4:3.  But then we
would be lowering the effective horizontal resolution of the 4:3 material,
which still makes up the vast majority of programming on an SD channel.
Do you have a human involved in the process of upconverting the asset? Do you have a human master control operator who can look at the feeds and see that a program is not being presented properly (i.e. 4:3 pillarbox versus 4:3 letterbox or widescreen?
I agree that it would be desirable to automate this. Frankly I cannot understand after all of these years why this has not happened in the broadcast industry, UNLESS, it is a matter of budgets and investments in "proper" equipment. 



I'm sure as time moves on, we will see more 480i 16:9 subchannels, most of
which are being fed downconverted HD material.  Almost 100% of the AFD
encoded material we receive from PBS is HD material, with AFD instructions
for downconversion.
No doubt this will continue to happen in broadcast multiplexes, and 480i will continue to be a factor for SD DVD. But interlace is almost gone for Over-The-Top video delivered via the Internet. What you see the most now is 480i being converted to 360P for encoding; and for anything new, progressive acquisition is pretty much the norm, except for the poor bastards who invested in 1080i. 



Hell, I'd just like to see consumer 16:9 televisions that are being sold
today recognize and use AFD.  Incredibly, the vast majority still do not.
That would go a long way in eliminating the postage stamp problem.
I'd like to see consumer televisions with some intelligence (including the source interconnects), some form of networking, and the ability to deal with metadata as intended. AFD is an interesting approach, but it is an unnecessary band-aid, since the information needed to do the job properly is in the transport streams.
This is all the expected outcome of a transition where some proponents stubbornly held onto legacy technologies to protect earlier investments in digitized analog video (i.e. systems using the ITU-R bt.601 standard). And it is a predictable outcome, given the fact that many of the broadcast and CE proponents of the ATSC standard believed interlace would create a barrier to competition from the computer industry.
Stations COULD HAVE made the investment in high quality standards convertors that deinterlaced SD before emission encoding, rather than forcing this functionality into every CHEAP TV. AND the decision to include an interlaced HD format to protect the investment made by the Japanese in HD technology was very shortsighted. 1080i has proven to be a step backwards, which has delayed the entire transition to appropriate acquisition technologies for moving imagery. What is appropriate you ask? 

1. Progressive scanning
2. Oversampling relative to the emission format.

Regards
Craig
P.S. Now that broadcasters are beginning to understand what DIGITAL television really is they are faced with another fork in the road:
One path leads to a digital plant that is built around the handling of compressed assets and a move to an automated system for delivering streams to the broadcast multiplex. Does anyone believe that there is some huge master control system at YouTube where humans push buttons (or monitor an automation system) to deliver every file? In essence, much of the current investment in the upgrade to DTV was inappropriate and needs to be replaced with appropriate technologies.
The other path leads to the digital cliff, where many stations will take the cash for their spectrum and move on, rather than making another investment in the technology they will need to survive... 

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:

  1. Home
  2. opendtv
  3. Archive
  4. 01-2011