HDR Discussion Thread

[Shalashaska]

Do you guys think HDR will ever go mainstream?

Even with all its benefits, will people care if they can't see anything on the screen in a bright living room setup?

[HeatEquation]

Originally Posted by Shalashaska

Do you guys think HDR will ever go mainstream?

Even with all its benefits, will people care if they can't see anything on the screen in a bright living room setup?

What do you mean by mainstream? On cable/satellite TV? If so, then yes, I do. The format that would be used for broadcast HDR would be HLG, which is suppose to look good in various different environments, including in bright rooms.

Also, HDR is already being incorporated in a lot of streaming content, and on virtually every 4K disc.

[Jason 831]

Originally Posted by Shalashaska

Do you guys think HDR will ever go mainstream?

Even with all its benefits, will people care if they can't see anything on the screen in a bright living room setup?

HDR benefits from brighter displays. TVs will get brighter help and brighter displays will help the viewing in a bright living room setups

[JohnAV]

Originally Posted by Staying Salty

http://www.patentlyapple.com/patentl...-industry.html

None of the more active Apple Rumor sites that have been following the ATV5 development came up with that. Why? Does anyone here think that Apple will cozy up to Samsung so easy?

[Shalashaska]

Originally Posted by HeatEquation

What do you mean by mainstream? On cable/satellite TV? If so, then yes, I do. The format that would be used for broadcast HDR would be HLG, which is suppose to look good in various different environments, including in bright rooms.

Also, HDR is already being incorporated in a lot of streaming content, and on virtually every 4K disc.

I'll admit, I've never fully understood how HLG works, so maybe they've already found a way to bypass that issue. If that's the case, hooray for HDR!

As for its growing presence in streaming and physical media, that doesn't really mean it's gone mainstream. 99% of TV owners will still be watching SDR content on an SDR display over cable/satellite TV or streaming. It's still an enthusiast format for enthusiasts with new expensive TVs. 3D was everywhere for a while there, but it never seemed to catch on in the home. It didn't suit most people's viewing conditions, not unlike HDR if they don't find a way to make HDR watchable in a brightly-lit room.

Originally Posted by Jason 831

HDR benefits from brighter displays. TVs will get brighter help and brighter displays will help the viewing in a bright living room setups

That's what you'd think!

http://www.hdtvtest.co.uk/news/4k-vs-201604104279.htm

[Geoff D]

Originally Posted by Shalashaska

I'll admit, I've never fully understood how HLG works, so maybe they've already found a way to bypass that issue. If that's the case, hooray for HDR!

As for its growing presence in streaming and physical media, that doesn't really mean it's gone mainstream. 99% of TV owners will still be watching SDR content on an SDR display over cable/satellite TV or streaming. It's still an enthusiast format for enthusiasts with new expensive TVs. 3D was everywhere for a while there, but it never seemed to catch on in the home. It didn't suit most people's viewing conditions, not unlike HDR if they don't find a way to make HDR watchable in a brightly-lit room.



That's what you'd think!

http://www.hdtvtest.co.uk/news/4k-vs-201604104279.htm

Think of HDR10 as having a light level which is completely fixed for any given shot: if it's 10 nits then it MUST be 10 nits, if it's 1000 nits then it MUST be 1000 nits. If the display can't meet these figures then this is where the tone mapping kicks in, re-adjusting the content to fit the properties of the target display but, as we know, this is easier said than done in lieu of content-derived metadata and can introduce a whole slew of problems. This Perceptual Quantiser EOTF is referred to as having absolute luminance.

HLG works more like the conventional gamma system in that while it may have been mastered to a set amount of nits it does not assume that x value of brightness will be waiting for it at the display end. Instead of rendering the content as absolute values first and then letting the TV decide best on how to map it as per HDR10, HLG builds the rendering of the content into the actual decoding of the signal by the display so that the brightness, range etc is instantly best adapted to whatever the target display is outputting. And because it leaves the decision on how to show the content up to the display in the first place it needs no metadata. This Hybrid Log Gamma EOTF is referred to as having relative luminance.

"Right, but wouldn't a dynamic metadata system like Dobly be just as useful in mapping the image to a target display then?" Fine question, but it doesn't take into account the kinds of logistical challenges to create metadata on the fly for a live TV production, not to mention the handling of mixed sources like now/next graphic overlays that broadcasters do nowadays as Penton pointed out (because we can't let one programme finish without letting you know what's coming up, people having the assumed attention spans of goldfish). And there's always the potential for dynamic tone mapping to actively reduce the APL of any given scene in order to best balance highlights and brightness, whereas HLG's simpler rendering of the visual intent means that the brightness/APL will always be properly matched to however you've got your specific display set up which makes it a boon for watching in the daytime, unlike that HDTVtest, er, test which was done with 'absolute' HDR10 content.

HLG doesn't need any super duper updates to the actual transmission systems either, and as the first part of the curve resembles normal SDR anyway it can thusly be viewed on a SDR display without any need for mapping, metadata etc which is itself another little bonus for broadcasters (created by the BBC and NHK as it was, this is no surprise). It could be argued that HLG does neither HDR or SDR as good as other forms of those two systems, but in terms of the ephemeral nature of most TV broadcasts (especially sports) and getting HDR into that specific viewing environment then its advantages outweigh its disadvantages IMO.

HTH, and if I've goofed anywhere then there are several actual experts who'll correct me in due course.

[Shalashaska]

Originally Posted by Geoff D

Think of HDR10 as having a light level which is completely fixed for any given shot: if it's 10 nits then it MUST be 10 nits, if it's 1000 nits then it MUST be 1000 nits. If the display can't meet these figures then this is where the tone mapping kicks in, re-adjusting the content to fit the properties of the target display but, as we know, this is easier said than done in lieu of content-derived metadata and can introduce a whole slew of problems. This Perceptual Quantiser EOTF is referred to as having absolute luminance.

HLG works more like the conventional gamma system in that while it may have been mastered to a set amount of nits it does not assume that x value of brightness will be waiting for it at the display end. Instead of rendering the content as absolute values first and then letting the TV decide best on how to map it as per HDR10, HLG builds the rendering of the content into the actual decoding of the signal by the display so that the brightness, range etc is instantly best adapted to whatever the target display is outputting. And because it leaves the decision on how to show the content up to the display in the first place it needs no metadata. This Hybrid Log Gamma EOTF is referred to as having relative luminance.

"Right, but wouldn't a dynamic metadata system like Dobly be just as useful in mapping the image to a target display then?" Fine question, but it doesn't take into account the kinds of logistical challenges to create metadata on the fly for a live TV production, not to mention the handling of mixed sources like now/next graphic overlays that broadcasters do nowadays as Penton pointed out (because we can't let one programme finish without letting you know what's coming up, people having the assumed attention spans of goldfish). And there's always the potential for dynamic tone mapping to actively reduce the APL of any given scene in order to best balance highlights and brightness, whereas HLG's simpler rendering of the visual intent means that the brightness/APL will always be properly matched to however you've got your specific display set up which makes it a boon for watching in the daytime, unlike that HDTVtest, er, test which was done with 'absolute' HDR10 content.

HLG doesn't need any super duper updates to the actual transmission systems either, and as the first part of the curve resembles normal SDR anyway it can thusly be viewed on a SDR display without any need for mapping, metadata etc which is itself another little bonus for broadcasters (created by the BBC and NHK as it was, this is no surprise). It could be argued that HLG does neither HDR or SDR as good as other forms of those two systems, but in terms of the ephemeral nature of most TV broadcasts (especially sports) and getting HDR into that specific viewing environment then its advantages outweigh its disadvantages IMO.

HTH, and if I've goofed anywhere then there are several actual experts who'll correct me in due course.

It's all starting to make some sense! Thanks!

[Shalashaska]

Your first paragraph also reminded me of another question.

Is HDR10 tone-mapped based on its 1000/4000-nit container, or its more content-derived MaxCLL/MaxFALL?

[Geoff D]

Originally Posted by Shalashaska

Your first paragraph also reminded me of another question.

Is HDR10 tone-mapped based on its 1000/4000-nit container, or its more content-derived MaxCLL/MaxFALL?

That's a good question, one that I don't have a definitive answer for because every manufacturer has a different approach to mapping, and according to HDTVtest some discs (dunno which) actually have the metadata strings nulled out anyway because the content providers don't think that static mapping is up to scratch.

I'll hazard a guess though and say that the basic minimum/maximum display mastering values of the container itself are what's being interpreted first and foremost, especially given how dark certain 4000-nit-but-low-MaxCLL titles like Goodfellas (247 nits max) are interpreted as being on some displays. It can end up much too dark because the mapping isn't looking at the maximum content light level but the overall mastering maximum and it's squishing the whole 4000-nit container down to fit x nits peak of the display without realising that the content's brightest point (never mind the MaxFALL or APL in general) only just exceeds 5% of that amount of peak brightness which then drastically affects the overall brightness of the image. (This is apparently how Dolby's own static mapping algorithm works.)

This is pretty much why Sony's 'mapping' in their premium HDR sets is more of a hard clip that's best set to the limits of what the display can handle, with the A1 OLED nearer 1000 nits, the ZD9 nearer 2000 nits and everything above that thrown away (in order to track the PQ EOTF correctly, anyway). While this will lose you some of the brightest specular highlight detail it will also ensure that the APL is kept nice and high for the average piece of HDR content.

[Richard Paul]

Originally Posted by DanBa

"Panasonic and 20th Century Fox have joined Samsung in its plan to offer HDR10+ as a royalty-free alternative to Dolby Vision. The HDR10+ format will be made available to more manufacturers in early 2018.

The three companies will make HDR10+ available to additional manufacturers of TVs, Blu-ray players, set-top boxes, and SoC vendors starting in January 2018. It will be royalty-free but requires that manufacturers pay a “nominal administrative fee”, the alliance said."
http://www.flatpanelshd.com/news.php...&id=1503914052

That is huge news and Dolby must be angry about that. With a major studio supporting HDR10+ it will be here for the long term and even if it doesn't replace Dolby Vision it will always be an option that companies can choose. Options are great for consumers especially when they are free.

[gkolb]

Originally Posted by Richard Paul

That is huge news and Dolby must be angry about that. With a major studio supporting HDR10+ it will be here for the long term and even if it doesn't replace Dolby Vision it will always be an option that companies can choose. Options are great for consumers especially when they are free.

If they get off their duff and get the disk authoring issue solved (if that's the root problem currently), they can make more converts out the casual DV enthusiasts (not the persons that are usually in these forums). The bigger the popular base that they can build early, the more converts they will add to their side each month.

Having a shaky start with "grey bars" talk, and no forthcoming explanations on their part to put the issues in context is leaving a vacuum in space that other parties are trying to fill. Not good!

[Penton-Man]

Originally Posted by Shalashaska

Do you guys think HDR will ever go mainstream?

It will take some time but yes, eventually. As most recently mentioned and shown here, e.g. - https://forum.blu-ray.com/showthread...w#post13941774 , for live sports, SDR broadcast can’t accommodate the dynamic range from bright sunlight to stadium shadow, whereas HDR broadcast can.

In the future, be it at their local retailer or wherever, once the general public witnesses the improvement in picture quality with HDR and such things as soccer and American football especially involving events on a bright sunny day with stadium shadows, it should serve as a catalyst for widespread adoption.

Although, keep in mind that the benefit of HDR is not exclusive to stadiums challenged with shadows, as the visual improvement (colors, contrast) with HDR broadcast involving indoor arenas showcasing basketball and ice hockey matches has also been clearly demonstrated in tests. In fact, to give you an idea as to its superiority, with regards to the later, even the skate marks on the ice become so much more apparent and detailed in HDR that they can arguably prove a distraction to the actual play and the shaders might have to mitigate this.

[Penton-Man]

Originally Posted by Geoff D

....not to mention the handling of mixed sources like now/next graphic overlays that broadcasters do nowadays as Penton pointed out

For those new readers to this thread, see - https://forum.blu-ray.com/showthread...e#post13745238 and catch a few posts in the conversation with Geoff on that page for further elaboration of the finer points.

[Penton-Man]

Originally Posted by HeatEquation

What do you mean by mainstream?

I mean mainstream in terms of widespread adoption by the public, not just availability to consumers – http://ieeexplore.ieee.org:80/docume...ns?reload=true
2017 Update on High Dynamic Range Television
SMPTE Motion Imaging Journal
Issue 7 • Sept. 2017
Page(s):94 - 96

And in due course. eventually supplemented with HFR (HDR plus HFR), because of the benefits of that increased temporal resolution (https://forum.blu-ray.com/showthread...r#post13558917) should prove irresistible to many, not just study participants exposed to HFR content alone, which like HDR, by itself, has shown a significant increase in the quality of the viewing experience.

[puddy77]

Panasonic announced their big firmware update for the GH5 will be released at the end of September. It includes the new HLG direct recording mode: https://www.dpreview.com/news/979912...xpected-extras

[HeatEquation]

Panasonic's IFA presentation on HDR10+ (starts at 19:00):

https://www.youtube.com/watch?v=tTyR...ature=youtu.be

[mrtickleuk]

Originally Posted by Penton-Man

On the topic of HDR10+…..

Note to Oliver K. My friend , if you’re heading down south to Berlin to take in IFA, how about a report on this session - http://www.displaysummit.com/hdr10-hdr10-soc-for-tvs/.

If anyone goes, please can you ask them - beg them! - for a publicly accessibly test clip that we can play? Even just ONE! I don't mind if it's just a 10 second testcard!

It's now 18 months (https://vimeo.com/164619204) since "Dynamic metadata for HDR10" was first demonstrated. Yet we still have nothing to play with!

TIA.

[HeatEquation]

Originally Posted by mrtickleuk

If anyone goes, please can you ask them - beg them! - for a publicly accessibly test clip that we can play? Even just ONE! I don't mind if it's just a 10 second testcard!

It's now 18 months (https://vimeo.com/164619204) since "Dynamic metadata for HDR10" was first demonstrated. Yet we still have nothing to play with!

TIA.

Probably going to have to wait until Amazon releases HDR10+ content.

[HeatEquation]

Steve May‏Verified account @SteveMay_UK 6h6 hours ago
Just had an interesting demo of HDR10+ running on a Panasonic OLED. Looking much better than when I first saw it at CES! #IFA2017

https://twitter.com/SteveMay_UK/stat...95744889360385

Steve May‏Verified account
@SteveMay_UK
Replying to @JohnnyFocal @BigJohnnyArcher
Fully expect the BDA the rubber stamp HDR10+ at CES...

https://twitter.com/SteveMay_UK/stat...15573016866817

Steve May‏Verified account @SteveMay_UK 6h6 hours ago
Fox takes the stage with Panasonic to promote HDR10+, promises a slate of new releases. Firmware update coming to 2017 4k Pro TV range...

https://twitter.com/SteveMay_UK/stat...89181084622849

[Penton-Man]

I am not at liberty to comment on that ^ ("rubber stamp") at this time. Fellow Insider 2themax is and is better informed than any journalist on the topic.