HDR Discussion Thread

[Penton-Man]

Richard's back.

Originally Posted by Richard Paul

I am curious to see if HDR10+ can make it into ATSC 3.0 since Dolby has a lot of connections in the US broadcast industry. So far the only amendments that have been voted on are to add full range video, ICtCp, and Dolby dynamic metadata. Also any broadcasters that want to avoid metadata can use HLG.

Yes, specific benefits of metadata free operation…..

[Penton-Man]

^ middle thumbnail pic is to illustrate squeeze and tease compositing at the end of a program, i.e. difficult with a metadata based broadcast approach

[Geoff D]

So even compositing in them little now/next boxes as broadcasters are wont do (very familiar with the above from the BBC, being a Britisher) messes with the metadata? Makes sense.

[Penton-Man]

Originally Posted by Geoff D

So even compositing in them little now/next boxes as broadcasters are wont do (very familiar with the above from the BBC, being a Britisher) messes with the metadata? Makes sense.

Correct and apart from live graphics offered during operation, also something which I didn’t add to in the bullet points of the above pic is that with regards to consumer equipment, specifically, picture in picture operations and overlapping things like graphics for subtitles, closed captioning, etc. is more difficult with a metadata based approach.

[Penton-Man]

Originally Posted by Geoff D

from the BBC

Also, notably, it has been proposed by the BBC, and many in the industry hold this to be of merit, that HLG inherently holds an advantage over PQ-based approaches in brighter viewing environments, e.g., superior viewing with HLG than PQ when watching sports (soccer, football, etc.) during the daytime in one’s family room with friends…..or with any content in a non-light controlled room.

I know that BBC have done additional unpublished subjective tests, with the aid of a studio, on that matter and I hope that by year’s end, perhaps at IBC 2017 or the SMPTE 2017 Annual Technical Conference & Exhibition in Hollywood, both of which I plan on attending, that the OOTF for any given surround light environment be further addressed and although this timeline may be pushing it for their math wizards, that the results of this new testing yield a precise formula presented in a paper at either of these two upcoming events for public analysis.

[Geoff D]

Makes sense also, as the relative luminance of HLG and the rendering of the OOTF in the display device rather than being assumed and baked into the content should mean that brightness/APL doesn't take anything like the kind of hit as it does with tone mapped PQ content.

[Shalashaska]

I appreciate this thread and the posters in it very much, despite the fact that I really have no clue what's going on here most of the time.

[Penton-Man]

Originally Posted by Shalashaska

I appreciate this thread and the posters in it very much, despite the fact that I really have no clue what's going on here most of the time.

Thank you , as to me, then I promise not to discuss the usage of Barten as opposed to Schreiber ramps for determining the threshold for visible banding in images. I can’t speak for others, but I can tell you that you’re good in my book as long as you don’t ever cause Verity to cry…..

The Harklets in "The Dirty Cinema Screen" - YouTube

especially when we get
direct LED Cinema installations.

[Penton-Man]

Originally Posted by Penton-Man

Also, notably, it has been proposed by the BBC, and many in the industry hold this to be of merit, that HLG inherently holds an advantage over PQ-based approaches in brighter viewing environments, e.g., superior viewing with HLG than PQ when watching sports (soccer, football, etc.) during the daytime in one’s family room with friends…..or with any content in a non-light controlled room.

I know that BBC have done additional unpublished subjective tests, with the aid of a studio, on that matter and I hope that by year’s end, perhaps at IBC 2017 or the SMPTE 2017 Annual Technical Conference & Exhibition in Hollywood, both of which I plan on attending, that the OOTF for any given surround light environment be further addressed and although this timeline may be pushing it for their math wizards, that the results of this new testing yield a precise formula presented in a paper at either of these two upcoming events for public analysis.

Note to Robert Zohn,
Kevin Miller should especially be interested in the yet to be published BBC’s findings ^ in regards to non-reference viewing conditions, given his presentation next Tuesday.

[Penton-Man]

Originally Posted by Penton-Man

....given I’ve been talking the value of ACES with words and motivational cartoons

for years now, for example

rather than again doing a lot of typing, this time with regards to VFX, for further explanation, I’ll just link with a timestamp to an animated ‘image crafter’…..https://www.youtube.com/watch?v=kUOo...sktop#t=45m11s
followed by Bobby who henceforth I think will be in Josh Pines’s doghouse.

Also, since the panelists couldn’t talk about it, a past example of a motion picture having an ACES workflow, namely, Life…..
http://www.digitalcinemareport.com/a...r#.WUb9LOvyu1s

Speaking of ‘image crafters’ who are more budget-minded DSLR inclined, are there any Panasonic GH5 afectionados - http://www.pocket-lint.com/review/13...il-the-4k-king reading?

I’m wondering whether the GH5 got the HLG capture firmware yet?

[zmarty]

Originally Posted by Penton-Man

I’m wondering whether the GH5 got the HLG capture firmware yet?

I use a GH5 to shoot videos for my channel. The HLG firmware is not out yet, but for HDR post-production it's better to shoot in V-Log L anyway.

The upcoming firmware (this summer) will have HLG and it will also feature a 400 Mbps video codec.

[Penton-Man]

Originally Posted by Shalashaska

I appreciate this thread and the posters in it very much, despite the fact that I really have no clue what's going on here most of the time.

from….http://www.streamingmedia.com/Articl...10-118853.aspx

“FYI, Dolby Vision is natively 12-bit in the ITP color space”…..also known as ICtCp - https://forum.blu-ray.com/showthread...p#post13467638

Folks can learn a bit more about PQ ICtCp from Jack Holm (http://www.displaysummit.com/some-co...-color-volume/)
next Tuesday at the Samsung sponsored event I previously alerted readers to, registration remains open.

[Penton-Man]

Reminds me, on a personal note to ray0414 as I recall you expressed a past fondness for Li …. https://forum.blu-ray.com/showthread...i#post12582945

Well, she won’t be giving the HDR10+ presentation at the event, but the co-inventor for hdr tone mapping with scene adaptive parameterized tone map function from Samsung will…. http://www.displaysummit.com/hdr10-beyond-hdr10/

....in case you were planning on attending

[Penton-Man]

Originally Posted by zmarty

I use a GH5 to shoot videos for my channel. The HLG firmware is not out yet, but for HDR post-production it's better to shoot in V-Log L anyway.

The upcoming firmware (this summer) will have HLG and it will also feature a 400 Mbps video codec.

Thanks z.

[Staying Salty]

Originally Posted by Penton-Man

Well, she won’t be giving the HDR10+ presentation at the event, but the co-inventor for hdr tone mapping with scene adaptive parameterized tone map function from Samsung will…. http://www.displaysummit.com/hdr10-beyond-hdr10/

"HDR10 is being adopted as an HDR video format nowadays. However, it is known that the HDR10 system suffers with highlight details loss or dimming phenomenon due to the intrinsic nature of the associated static tone mapping in a consumer device. In this talk, we will present a novel HDR technology called HDR10+ based on the SMPTE 2094-40, where SMPTE 2094-40 fundamentally provides scene by scene dynamic tone mapping metadata for proper tone mapping at a device. It can be also discussed and shown how HDR10+ improves the HDR quality at a consumer display."

I am still not convinced that the dynamic race horse returns to his stall to eat. Maybe he just slows way down at times during the race?

[Geoff D]

They've got a competing dynamic system to shill so they would say all that. It's gonna be an eye-opener comparing dynamic to static on a TV which already has a form of 'active' HDR processing like the Sony ZD9.

[zmarty]

Here are the technical requirements Netflix imposes on their content producers, including HDR specs: https://backlothelp.netflix.com/hc/e...uirements-v2-1

Some highlights:

Camera Requirements

4K Resolution:
• Camera must have a true 4K sensor (equal to or greater than 4096 photosites wide).

Recording Format:
• Minimum of 16-bit Linear or 10-bit Log processing
• Bitrate of at least 240 Mbps (at 23.98/24 fps) recording
• Recording format must be set to either: RAW (uncompressed or lightly compressed sensor data) OR Log color space (i.e. S-Log3, V-Log, CanonLog3, REDLogFilm, BMDLog, LogC)
• No looks or color corrections should be baked into the original camera files.
• Files must maintain all metadata (i.e. Tape Name, Timecode, Frame Rate, ISO, WB, etc.)



Approved cameras

[Penton-Man]

ray0414, I’m tracking the posts in which I refer to you and you don’t thank me. I view you as an underperformer in that regard and the lack of validation on your part disturbs my self esteem…..not sure how long I can last without member validation.

[Penton-Man]

Originally Posted by Staying Salty

...I am still not convinced that the dynamic race horse returns to his stall to eat. Maybe he just slows way down at times during the race?

Salty, your point is unclear. If you were more coherent, without the metaphorical reference, maybe your concern could be introduced by either the moderator or one of the non-yes men (do any still exist?) reporter-types at the event next week. Although they would never acknowledge it, a lot of lurkers read this thread.

[Penton-Man]

Originally Posted by zmarty

the technical requirements Netflix

[Show spoiler]

And continuing from the far right in your flow chart above, going from the Dolby Vision IMF master to HDR10 is not as simple as some may think…..

Many Dolby Vision masters are in P3 colorspace. For HDR10, Netflix has to convert to rec.2020 color space with some processing and a color adjustment detailed here - http://www.streamingmedia.com/Downlo...flixP32020.pdf which loosely describes how they optimize the quality of the HDR10 encodes derived from the Dolby Vision masters.

If any Netflix critics think they can do better, contact Jimmy….
https://www.linkedin.com/in/jimmyfusil/