AVS Forum

AVS Forum (http://www.avsforum.com/forum/)
-   Latest Industry News (http://www.avsforum.com/forum/286-latest-industry-news/)
-   -   NAB 2014 TWiT Special (http://www.avsforum.com/forum/286-latest-industry-news/1526902-nab-2014-twit-special.html)

Scott Wilkinson 04-11-2014 11:52 AM

I roamed the show floor of the 2014 NAB (National Association of Broadcasters) convention to interview representatives of companies at the forefront of the content-creation industry. I was looking for the latest technologies that may affect what consumers will see and hear in the future, especially in the realm of Ultra HD/4K. The companies you will hear from in this special coverage include Canon, LG, JVC, Panasonic, Sony, Christie, Dolby, DTS, Fraunhofer, Datacolor, SpectraCal, NHK, and more.

 

 


Orbitron 04-11-2014 09:20 PM

Scott, thanks for the excellent coverage.

tigerfan33 04-12-2014 06:14 AM

How was the pq on the OTA 8k booth compared to some of the 4k native content?

Masterbrew2 04-12-2014 07:49 AM

Fraunhofer's sound frame seems like an idea that be will dead in the water.

 

Who would want that contraption on their wall? If you would, you could get better sound from classic speakers that utilise their volume better.


TVOD 04-12-2014 11:28 AM

The
Quote:
Originally Posted by tigerfan33 View Post

How was the pq on the OTA 8k booth compared to some of the 4k native content?
From my eye I didn't see any noticeable artifacts in the 8K transmission. It was hard to make a direct comparison to 4K as the 8K booth only had 8K. The transmitted material was a pre-recorded segment. The OTA transmission used two streams, one on horizontal polarization and the other vertical, using h.265 and 4096QAM. They had spectrum analyzers and constellation displays on each polarity. I asked why not circular polarization instead of discrete H/V and they said the H & V used different power levels. So far in their test transmissions they have gone 17 miles.

I wasn't as impressed with the pseudo 22 audio channel display. They had foot prints marked on the floor where one was to stand to listen, but even then, perhaps due to the environment, there wasn't (at least to me) a strong sound envelope experience.

They also had a Sony F65 with a 8K interface live to a monitor and of course that looked great. It was a big improvement over the rather large Ikegami 8K camera with the 16 HDSDI connections shown in the past.

8K in 6Mhz is quite a feat!

TVOD 04-12-2014 12:01 PM

The Dolby Vision demonstration unfortunately used some rather small monitors. They had the same material on two side-by-side LCD monitors, the left using Dolby Vision and the right without. They didn't appear to the the same model monitors. The Dolby Vision one did have local dimming. Aside from any dark area performance differences between the monitors, the effect of bright highlights was stunning, even a little bit painful, at times. In an interesting way the slight discomfort added to the believability of the image as such contrast differences could likely occur in real life. Even with just specular highlights, there was quite a noticeable improvement. It is the complement to dense and detailed dark areas to improve realism.

I asked if they had published any specifics on their gamma curve, but there didn't seem to be anything yet. They did say the curve goes deeper into the blacks compared to the 709 linear point.

One interesting thing they told me was they were abandoning the idea of using metadata added to a traditionally encoded disc to re-create Dolby Vision video. I don't know if this is a temporary issue, but it sounded like they stopped pursuing that direction. I'm curious to see if there is any contradictory information coming.

EDIT: Apparently this is NOT true about the metadata. See Scott Wilkinson's post below.

My biggest disappointment of the show was not running across Scott Wilkinson to give a quick howdy from the peanut gallery.

Orbitron 04-12-2014 12:18 PM

Scott - OLED vs LED with HFR, in your opinion, which has a better image?

AVS153 04-12-2014 12:43 PM

Impressive video at NAB 2014

Is 2000 NIS screen with Dolby's Vision 
It can not hurt the eyes?


TVOD 04-12-2014 01:50 PM

Quote:
Originally Posted by AVS153 View Post

Impressive video at NAB 2014
Is 2000 NIS screen with Dolby's Vision 

It can not hurt the eyes?
Yes it can hurt your eyes, the same way a high power audio system can hurt one's ears. The idea is to allow a reserve for peaks on smaller areas. On the rest of the image the brightness is nearly the same. It's customary now when grading to use a curve (the knee) to gradually limit all the bright highlights. A sharp limit will result in splotches of white or a solid color with hard edges. When comparing a knee curve to the extended highlights in Dolby Vision, it just looks dull. It still may be desirable to limit the brightness of larger areas, but they can still extend past what is used with "traditional" gamma.

An interesting idea might be that dynamic range could be controlled by the viewer. If the extended brightness causes excessive eye fatigue to some viewers, they could have a way of controlling it to a comfortable range.

tenthplanet 04-12-2014 07:01 PM

I see the video says part 1, will there be a part two.

Daniel Chaves 04-13-2014 02:07 AM

Wow I just listened to the DTS Headphone X demo and I honestly felt like I was listening to actual distinct speakers placed around the room, it was mind blowing, cant wait for that tech to take off. smile.gif

AVS153 04-16-2014 08:32 AM

Quote:
Yes it can hurt your eyes, the same way a high power audio system can hurt one's ears.

If it hits the eye, How  there is no standard  to stops selling  2000 NIS screen with Dolby's Vision   ?

No  Standards Institute certification?


HockeyoAJB 04-16-2014 10:03 AM

Quote:
Originally Posted by TVOD View Post

...

One interesting thing they told me was they were abandoning the idea of using metadata added to a traditionally encoded disc to re-create Dolby Vision video. I don't know if this is a temporary issue, but it sounded like they stopped pursuing that direction. I'm curious to see if there is any contradictory information coming.

...

 

Is there an alternative to the metadata approach that would allow the same content to be shown on two different displays...one w/ Dolby Vision, in which case you would get the HDR effect, and one without, in which case you cannot?  Or does this mean that content that has Dolby Vision HDR information in it will be incompatible with displays that do not support Dolby Vision, meaning content providers will have to make two versions of said media and consumers will have to choose the appropriate version for them?


AVS153 04-16-2014 12:17 PM

Quote:
Originally Posted by tenthplanet View Post

I see the video says part 1, will there be a part two.

Joining to the question


Scott Wilkinson 04-16-2014 02:11 PM

Quote:
Originally Posted by tenthplanet View Post

I see the video says part 1, will there be a part two.


Part 2 was done by another TWiT host, Fr. Robert Ballecer. I hadn't planned on posting it, but maybe it's a good idea. Look for that shortly...


Scott Wilkinson 04-16-2014 10:13 PM

Quote:
Originally Posted by TVOD View Post

The Dolby Vision demonstration unfortunately used some rather small monitors. They had the same material on two side-by-side LCD monitors, the left using Dolby Vision and the right without. They didn't appear to the the same model monitors. The Dolby Vision one did have local dimming. Aside from any dark area performance differences between the monitors, the effect of bright highlights was stunning, even a little bit painful, at times. In an interesting way the slight discomfort added to the believability of the image as such contrast differences could likely occur in real life. Even with just specular highlights, there was quite a noticeable improvement. It is the complement to dense and detailed dark areas to improve realism.

I asked if they had published any specifics on their gamma curve, but there didn't seem to be anything yet. They did say the curve goes deeper into the blacks compared to the 709 linear point.

One interesting thing they told me was they were abandoning the idea of using metadata added to a traditionally encoded disc to re-create Dolby Vision video. I don't know if this is a temporary issue, but it sounded like they stopped pursuing that direction. I'm curious to see if there is any contradictory information coming.

My biggest disappointment of the show was not running across Scott Wilkinson to give a quick howdy from the peanut gallery.


I just checked with Dolby, and they are NOT abandoning the use of metadata for backward compatibility. The base layer will be Rec.709, and the HDR etc. will be encoded as metadata. A legacy display will simply ignore the metadata, while a Dolby Vision-capable display will use the metadata.

 

Regarding gamma, Dolby Vision does not use conventional gamma; instead, it uses something Dolby calls "perceptual coding." I don't know the specifics of that yet, but I intend to find out.

 

I'm sorry we missed each other at NAB!


TVOD 04-17-2014 08:13 AM

Quote:
Originally Posted by Scott Wilkinson View Post


I just checked with Dolby, and they are NOT abandoning the use of metadata for backward compatibility. The base layer will be Rec.709, and the HDR etc. will be encoded as metadata. A legacy display will simply ignore the metadata, while a Dolby Vision-capable display will use the metadata.

Regarding gamma, Dolby Vision does not use conventional gamma; instead, it uses something Dolby calls "perceptual coding." I don't know the specifics of that yet, but I intend to find out.

I'm sorry we missed each other at NAB!
Thanks for checking, it seemed questionable being a major departure from their previous announcements. The reason given was due to complexity. I'm glad they are still going in that direction.

Perceptual coding uses Dolby's "Perceptual Quantizer" or "PQ" to better match gamma to visual sensitivity which makes the quantizing more efficient. "Traditional" gamma was based on the correction required for CRTs. A good explanation is here:

"Reply from Mike Rockwell:
Dolby Vision delivers a 12-bit per component signal to the display. However, even with 12-bit, using the standard gamma encoding there is not enough precision to carry the entire range from 0 – 10,000 nits. To solve this, Dolby developed a new quantization curve that is based on human visual perception. Basically, what we did was research how much the luminance had to change in order for you to see it. This test was performed with multiple viewers across many demographics. From this research, we were able to create an optimal quantization curve where each numeric change was below the human perceptual threshold in order to avoid image artifacts. We call this curve the perceptual quantizer or PQ. This curve is currently being standardized through the Society of Motion Picture and Television Engineers (SMPTE)."

Since this has been submitted for standardization there should be some specifics.

Scott Wilkinson 04-17-2014 11:39 AM

Quote:
Originally Posted by TVOD View Post


Thanks for checking, it seemed questionable being a major departure from their previous announcements. The reason given was due to complexity. I'm glad they are still going in that direction.

Perceptual coding uses Dolby's "Perceptual Quantizer" or "PQ" to better match gamma to visual sensitivity which makes the quantizing more efficient. "Traditional" gamma was based on the correction required for CRTs. A good explanation is here:

"Reply from Mike Rockwell:
Dolby Vision delivers a 12-bit per component signal to the display. However, even with 12-bit, using the standard gamma encoding there is not enough precision to carry the entire range from 0 – 10,000 nits. To solve this, Dolby developed a new quantization curve that is based on human visual perception. Basically, what we did was research how much the luminance had to change in order for you to see it. This test was performed with multiple viewers across many demographics. From this research, we were able to create an optimal quantization curve where each numeric change was below the human perceptual threshold in order to avoid image artifacts. We call this curve the perceptual quantizer or PQ. This curve is currently being standardized through the Society of Motion Picture and Television Engineers (SMPTE)."

Since this has been submitted for standardization there should be some specifics.


Ah, yes, "perceptual quantization," that's the term they use. Thanks for the clarification and quote!



All times are GMT -7. The time now is 04:24 PM.

Powered by vBulletin® Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.