[opendtv] Re: FCC Opens ATSC Patent Costs Proceedings

  • From: Kilroy Hughes <Kilroy.Hughes@xxxxxxxxxxxxx>
  • To: "opendtv@xxxxxxxxxxxxx" <opendtv@xxxxxxxxxxxxx>
  • Date: Fri, 18 Sep 2009 02:45:25 +0000

The majority of video streamed and downloaded from the Internet is encoded at 
actual picture height and width with square samples (and progressive).  

It can then be optimally framed to whatever display it lands on according to 
the type of device, user preference, etc. (You might want a "full screen" 
center crop of a 1.85 AR movie on a portable device, but want full width 
letterboxed (different amounts) on your 4:3 or 16:9 TV, and not letterboxed at 
all on your 2.0 AR iPod or 21:9 display)

An Internet video may land on 640x320 AR 2.0 screens like iPods, 720x480 NTSC 
AR 1.333 screens like Playstation Portables, adjustable windows on PC screens, 
4:3 NTSC screens, 16:9 square pixel screens, 320x240 and 416x240 cell phones, 
etc.  A device with one screen aspect ratio will plug into external displays 
with different aspect ratios.

Active Format Descriptors are a kludge to repair the screen-specific formatting 
encoded, but they can't be relied on, so are mostly ignored. The problem is 
bigger than a descriptor and requires an end to end system model where 
producers/encoders know what they are supposed to do, and devices know what 
they are supposed to do.

It is a multiscreen world today, so it makes no sense to do bad things to the 
encoded source material to frame it for old TV screens or new TV screens.  Our 
motto should be "Do no harm"; encode the content as accurately and close to its 
original form as possible, and let the render and display devices frame 
according to their particulars.  

Same motto applies to encoding 24P content as is, not adding 3:2 pulldown that 
will just confuse your 120Hz or 240Hz display.  Same motto applies to not 
flicker filtering the encoded content, and leaving filtering to any renderer 
unfortunate enough to have to output fields to an interlaced raster display.  
Same motto applies to capturing the full gamut and letting displays reduce as 
necessary.  A corollary is that displays should stop scaling up the pictures 
they are given in order to crop the edges a few percent just because analog 
recording used to screw up the edges and CRTs used to have bad power supply 
regulation ("overscan").  Trust encoder/post to crop any garbage before 
encoding and show all 1920 or 1280 or 704 NTSC or 640 wide.

Introducing new standards usually involves throwing away lots of obsolete boxes 
(like NTSC TVs), unless you designed and required extension mechanisms from the 
start ... rarely the case for cost and vision reasons.  In computerland, we 
have to work much harder to make the hardware obsolete because new standards 
can be implemented with software updates.  Case in point:  Many companies are 
shipping the bits of HTML 5 they like even thought the standard isn't finished. 
 Definite flexibility/stability tradeoff.

Kilroy Hughes   

-----Original Message-----
From: opendtv-bounce@xxxxxxxxxxxxx [mailto:opendtv-bounce@xxxxxxxxxxxxx] On 
Behalf Of Manfredi, Albert E
Sent: Thursday, September 17, 2009 12:46 PM
To: opendtv@xxxxxxxxxxxxx
Subject: [opendtv] Re: FCC Opens ATSC Patent Costs Proceedings

Craig Birkmaier wrote:

> The ATSC chose to use MPEG-LA to administer the patent pool since it
> has some overlapping standards and the infrastructure in place to
> deal with collection of these royalties.

Aha. Thanks, Craig, that was the part I was missing.

> Seems a bit late for the FCC to be sticking its nose under this
> tent. The damage is done.

Could be, different administration, different priorities?

> At least one can now understand why so many companies spent so much
> to develop the ATSC standard. To bad the world is now moving on...
>
> 21 x 9 displays

As of now, according to Jeroen's description of 21:9 (which is very close to 
Cinemascope 2.35:1), what this super wide screen buys you is "only" the ability 
to zoom into the super-wide content, which in 16:9 displays would have a bit of 
letterboxing top and bottom. In the 21:9 display, such content under normal 
circustances would display both pillarboxed and letterboxed.

So surely that much you can do with any existing DTV standard, including ATSC. 
It's only a function of the display. Until DVDs and BluRay discs come encoded 
with anamorphic 21:9, you won't be able to actually benefit from the full 
resolution potential. And when anamorphic 21:9 becomes reality, all the DTV 
standards could be updated to incorporate this.

> 1080 @ 50/60P
> Extended color gamuts
> H.264

Just as M/H with H.264 has been added to ATSC, and just as HDTV with H.264 was 
added to DVB-T, I don't see why these new standards can't be introduced. Even 
the wide gamut now available with MPEG-2, as Ron pointed out. These are all 
layered standards, plenty upgradeable. Too bad that the US keeps mixing in so 
much proprietary content in its standards, but then again, that seems all the 
rage. China, Japan, and even Brazil have picked up on that. Just look at the 
digital radio insanity worldwide.

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.


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