[opendtv] Re: Video compression artifacts and MPEG noise reduction

  • From: John Willkie <johnwillkie@xxxxxxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx, opendtv@xxxxxxxxxxxxx
  • Date: Thu, 19 Jun 2008 18:49:33 -0700 (GMT-07:00)

Most of their spectrum?  I will assume that you mean "most of their bandwidth" 
since spectrum is a much wider term.

I never know what you mean by MPEG-4.  Do you mean ISO/IEC 14496 part 4 or 
ISO/IEC 14496 part 10?  Both are MPEG-4.  The latter is also called h.264 or 
AVC.  The former is usually called MPEG-4 or MPEG-4 part 4.  

I'm on two or three subcommittees actually working on these improvements to the 
ATSC toolkit.  I can't say much about that, outside of what I've read in public 
prints, but without going into detail, I haven't heard about any proposal that 
would permit use of "most" (meaning more than half) of channel capacity for 
mobile services.

You don't have to read very far to hear that the screen size being bandied 
about is something like 320x240.  Assuming AVC, how much bandwidth do you think 
a single program service would use up?  (Hint: it's much less than MPEG-2 
640x480.)

Before February 18, 2009, the ATSC should release candidate standards on this 
stuff, and we can see just how close or far from the mark your opinions are.  
You have heard the one-liner about opinions being like a**holes, because 
everyone has one?

It appears to me that not all broadcasters will want to do this with their 
channel(s), for business, financial or technical reasons, and other 
broadcasters will devote significant bandwidth to mobile services, perhaps so 
that erstwhile competitors can a slice of their channel for mobile services.  
But, unless I've missed something, the upper limit that can be devoted wouldn't 
permit the use of most of a channel for mobile.  HDTV is still important to 
MANY broadcasters, and depending on the content, it might not be possible to 
simulcast HDTV and mobile video.

The market, not you, will determine how successful these services are.  One 
thing to keep in mind: not all the services will be traditional, linear program 
services.  I can say that many people, from different disciplines, are working 
hard on making this stuff real, and "our leader" is Mark Aitken.

"Something is happening here, and you don't know what it is, do you, Mr. Jones."

I think work on all but the fringes of MPEG-2 stopped around 2004, if not 2000. 
 So, you are predicting the past on that point.

John


-----Original Message-----
>From: Bob Miller <robmxa@xxxxxxxxx>
>Sent: Jun 19, 2008 4:05 PM
>To: opendtv@xxxxxxxxxxxxx
>Subject: [opendtv] Re: Video compression artifacts and MPEG noise reduction
>
>Seems broadcasters intend to use part of the spectrum for mobile and
>use MPEG4. If it is successful and/or if others using spectrum above
>51 are successful broadcasters will be tempted to use more and more of
>their 6 MHz for mobile. Especially if few viewers are actually using
>8-VSB for fixed reception which also might fit into broadcasters
>plans. Everything points to the broadcasters using most of their
>spectrum for mobile services IMO.
>
>I don't see anywhere that MPEG2 fits in their plans and therefor think
>that few will spend much time improving MPEG2.
>
>Bob Miller
>
>On Thu, Jun 19, 2008 at 3:31 PM, Manfredi, Albert E
><albert.e.manfredi@xxxxxxxxxx> wrote:
>> Craig Birkmaier wrote:
>>
>> -----------------------------
>> http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel5/76/27384/01218192.
>> pdf?temp=x
>>
>> Summary: This paper presents an overview of the transform and
>> quantization designs in H.264. Unlike the popular 8/spl times/8 discrete
>> cosine transform used in previous standards, the 4/spl times/4
>> transforms in H.264 can be computed exactly in integer arithmetic, thus
>> avoiding inverse transform mismatch problems. The new transforms can
>> also be computed without multiplications, just additions and shifts, in
>> 16-bit arithmetic, thus minimizing computational complexity, especially
>> for low-end processors. By using short tables, the new quantization
>> formulas use multiplications but avoid divisions
>> ------------------------------
>>
>>> So h.264 provides a significant improvement in arithmetic accuracy
>>> and the visibility of artifacts is reduced by acting upon smaller
>>> 4 x 4 regions of the image.
>>
>> Actually, you will not see that "accuracy" is improved mentioned
>> anywhere, but rather that computational complexity is reduced. An
>> integer transorm is not inherently more "accurate" at all, even if the
>> inverse doesn't introduce any additional error. And yes, the variable
>> block sizes will make blocking artifacts less obvious.
>>
>>> So my educated guess is that products like the Algolith device
>>> will not find their way into TV receivers. The shift to improved
>>> compression techniques will make this unnecessary.
>>
>> I don't buy your reasoning, Craig, because a shift to improved
>> compression techniques is not something a CE manufacturer can introduce
>> into his product all by himself. It requires a standards change and a
>> migration of the broadcast system.
>>
>> This box, very simply, can be used to extend the life of MPEG-2, both at
>> the transmission end and at the receive end. Whether it is hugely
>> successful or not is another matter. It took a whole lot of time to get
>> that point across.
>>
>> 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.
>

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