[opendtv] Re: Push for Ultra HD 4K forces picture quality tradeoffs - CNET

  • From: Mark Schubin <tvmark@xxxxxxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Tue, 27 May 2014 15:58:52 -0400

They showed real-time decoding of 4K in 5 Mbps at NAB 2013 and real-time encoding at NAB 2014.


On 5/27/2014 12:14 PM, donald.koeleman@xxxxxxxxxx wrote:
Jeroen, I saw a good h.265 demo at Fraunhofer's booth last IBC, 1080P in 2 Mbit/s average, they could use that existing software encoder/chip combination for 4K using four and a combiner, they confirmed when I asked them about this. So, 8 Mbit/s is not a lot of bandwidth, I haven't heard otherwise, so assume that Dutch Cable companies are still fed directly the (public) broadcasters SD service at 8 Mbit/s constant bitrate. We are today seeing the first terrestrial HVEC 4K broadcasts in France from the French Open Tennis Tournament. Even with DVB-T2 this is a bandwidth limited platform, but capable enough to carry a single 50fps 4K channel. As for production, well, if the content and service uptake warrants the effort... Not much different than colour, stereo, hd in that respect, they will make it work.

You are right, Philips has become a licensing house when it comes to CE products, Gibson being the latest licensee.

Hope to see some of your great work soon.


Donald

Quoting Jeroen Stessen <jeroen.stessen@xxxxxxxxxxx>:

Hi all,

Mark and Donald are right that motion portrayal is important.
High brightness makes your eyes respond faster, making good
motion portrayal even more important.

I once looked at a 60 fps CRT monitor and it flickered. Then I
put on my sunglasses and the flickering was gone.

HDR and HFR are orthogonal to each other, you can implement
one with or without the other. 4K UHD is not so orthogonal,
it costs a lot of bandwidth, and once you have a higher static
resolution you're also badly going to wish for a higher dynamic
resolution.

The "visible spectrum" covers the entire horseshoe in the (x,y)
or (u',v') plane, i.e. all the colors that human eyes can ever see.
The BT.2020 gamut covers only a large triangular part of that,
i.e. all the colors that you would reasonably want to see except
maybe some extreme cyan. Choosing an RGB-based color gamut
is a trade-off between gamut size, accuracy, and wasted codes.
The (u',v') representation is so much more efficient than (Cb,Cr)
that we can easily afford to cover the entire visible gamut.

I remind you that Philips no longer makes TVs, that has all been
licensed to various companies like TP Vision, my former employer.
I am now with Philips IP&S, who do research and write standards.
Our work can land in any TV brand in the world.

Best,
-- Jeroen



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