HDR Discussion Thread

[Kris Deering]

Originally Posted by Penton-Man

Where is Kris?
Reading the Navy Times? - https://www.navytimes.com/news/your-...rrier-captain/

You’re not a faceless loony, we’ve got eyes on you - https://forum.blu-ray.com/showthread...5#post16639962

I'm around, I've just been busy. I find Vincent's video interesting for sure. It does show how the single pixel level drives the MaxCLL, but also shows how overall luminance is still mainly restricted to 4000 nit (which makes sense since you wouldn't have wanted clipping on the Pulsar used for grading). I was bummed he didn't go just a bit past the scene when the car gets sucked up into the tornado and explodes as there is a lightning strike just after that scene that actually has illegal values for color that has thrown off a couple of different tone mapping solutions.

But the single pixel thing does illustrate why MaxCLL is not always the best indicator for what your display should be set for in terms of its peak value or clipping point. If you were to setup the peak white of your display to 10,000 nits for this particular movie, you would be wasting a ton of dynamic range, making everything look dark when it doesn't need to be.

The other thing this has me thinking about is the fact that blue in the YCbCr space is supposed to be limited to 7% of the video signal (if memory serves), so peak level for blue should only be 700 nits max. I may have to consult Stacey on that and get his thoughts.

[Kris Deering]

So I chatted with Stacey about my thoughts. With 709 blue was limited to 7% (so peak blue is limited to 7 nits) whereas with 2020 it is 5% (so limited to 500 nits). So when Vincent goes to the Parade screen and shows the blue at nearly 10,000 nits, that is wrong. Parade isn't nits, it is code value. When they derive MaxCLL they arbitrarily assign a luminance nit value to the peak single pixel. So in this case 99xx because of blue. They treat the peak pixel as a white luminance generically, when really parade is meant to show max CODE value, not luminance. So in this case the max blue luminance of that frame would be somewhere in the 400 nit range. Once again, a format that was designed to confuse and frustrate all that dive in!

You can see when he is actually looking at the waveform monitor that shows true luminance that this title is nearly locked at 4000 nits (it looks clipped there in a lot of scenes he shows, which very little overshoot). This makes sense since it is a Pulsar used and the grader would not want to see clipping on the display (the Pulsar doesn't tone map, but it can go nearly to 7000 nits in VERY small window sizes to allow some latitude).

[Penton-Man]

Originally Posted by Kris Deering

I'm around, I've just been busy....

stay safe
I’d hate to lose any of you guys.

[Penton-Man]

Originally Posted by Kris Deering

So I chatted with Stacey about my thoughts. With 709 blue was limited to 7% (so peak blue is limited to 7 nits) whereas with....

yeah, agree with all of thee above, in fact I think blue even in P3 (SDR) comes out to only about 7 nits - https://forum.blu-ray.com/showthread...s#post16445811

[Batman90]

Ok I couldn’t find a thread on this. But can anyone tell me how to calibrate my Vizio V series 4K Dolby vision tv so I’m getting the best picture possible?

[JohnAV]

Originally Posted by Batman90

Ok I couldn’t find a thread on this. But can anyone tell me how to calibrate my Vizio V series 4K Dolby vision tv so I’m getting the best picture possible?

Look at rtings calibration settings for your model.

[Lee A Stewart]

Bitmovin Announces Support for Dolby Vision

Online video technology company Bitmovin April 23 announced it will license and implement Dolby Vision, Dolby’s high dynamic range 4K video format, into its transcoding solution.

“Bitmovin believes that Dolby Vision is quickly becoming the de facto HDR standard for 4K TVs and that streaming services will continue to be at the forefront of Dolby Vision 4K content,” according to a Bitmovin press release.
Dolby Vision uses dynamic metadata to adjust the picture on a scene-by-scene basis for more detail and better color.

“The integration of Dolby Vision in Bitmovin’s Video encoding means that Bitmovin customers will be able to deliver more dynamic and brighter images while using minimum bandwidth,” according to the release.

Bitmovin Video Encoding offers per-title and three-pass features that automatically select the best encoding options according to the complexity of each piece of content, adjusting bitrate requirements to optimize the delivery of each individual asset, according to the company. This reduces bandwidth consumption by applying the minimum bitrate required to maintain perfect visual quality.

“Image quality is essential for survival in an increasingly competitive streaming market, and HDR is just as important as 4K,” said Stefan Lederer, CEO of Bitmovin, in a statement. “By integrating Dolby Vision into Bitmovin Video Encoding, our customers will be able to create and stream content that delivers transformative HDR experiences on minimum bandwidth and provide consumers with the viewing experience they deserve and expect.”

https://www.mediaplaynews.com/bitmov...-dolby-vision/

[Penton-Man]

Follow up from 10/09/2017 - https://forum.blu-ray.com/showthread...n#post14181916

In these challenging times, now offered are some free SMPTE standards and Motion Imaging Journal Articles -
https://www.smpte.org/free-standards-and-publications

one of which might be of interest to some who may have missed that post/link from way back in ’17 located near the bottom of the list is -
How Independent are HDR, WCG, and HFR in Human Visual Perception and the Creative Process?
(click on the redirected pdf)

[Penton-Man]

from 2-15-2020 -

Originally Posted by Penton-Man

The title of this particular lecture is a bit misleading because the presentation (including a valuable slide deck) covers so much more beginning with fundamentals regarding HDR and WCG (which may be beneficial as a refresher to some) and then becomes more advanced with the discussion of test patterns -
https://theiabm.org/iabm-webinar-how...is-color-bars/

Skipping to the later, given the importance of color bars to those in the know, one thing I will add to Norm’s short review at the end (~ the 47:15 timestamp) which I don’t think was covered at all is that ITU BT. 2111 was published a while ago and subsequently when content providers put those recommended HDR color bars thru tone mapping to BT. 709 they got weird distorted color bars with the brighter colors being attenuated which caused all sort of problems.

So, the EBU will be coming out in the coming weeks with new HDR color bars (as well as some other test signals, etc. – luma ramp, saturation check, Quad SDI check ) which should help drive greater consistency in HDR -> SDR down mapping enabling content providers to be more confident and precise in what their HDR signal will look like when it is viewed by SDR audiences.

Follow-up.
the document –
https://tech.ebu.ch/publications/tech3373?

[Penton-Man]

Originally Posted by Penton-Man

even the Pomeranian was impressed –

(technical, Camera -> AJA’s FS-HDR

From Claudio as DoP, Editor, Production Design, Animal Wrangler, Self-Isolationist, a newly minted professional video -
https://www.instagram.com/p/B91-2DWJI5F/

https://www.instagram.com/p/B91-2DWJI5F/

[LordoftheRings]

Originally Posted by I DO BLU

Because he was exposed as a fraud with his FAKE HDR video

Really!?

* We need Vincent back here so that we can discuss this HDR business with our questions.

[Penton-Man]

Yeah, sorry Blu, but I found that comment a bit harsh too. Vincent is not a ‘fraud’. I find him to be somewhat of an unnecessarily over-the-top showman in terms of trying to sell his business and promote his career unlike others who are just as qualified but mostly fly under the radar. And he’s simply made some proven mistakes with his past videos and assertions (completely aside from that of the ‘fake HDR’ thing which I don’t care to get into because it’s already been beaten into the ground).

[sapiendut]

Anybody can eventually make mistakes.

[Penton-Man]

Originally Posted by sapiendut

Anybody can eventually make mistakes.

absolutely true , but for someone who is as visible and followed as he is on the YouTube and the tweeter he should be held to a higher standard than most

problem is, his followers accept everything he claims as long as it sounds good and they are overwhelmed with technical talk

[LordoftheRings]

No better man than Vincent himself to defend his interests and professionalism as a video calibrator expert.

[Penton-Man]

Originally Posted by LordoftheRings

No better man than Vincent himself to defend his interests and professionalism as a video calibrator expert.

Certainly, he doesn't need wild-eyed advocates. And I think in general overall, Vincent provides much more of a service than disservice to the AV community.
But don’t dismiss the greater strength of peer review by those who don’t flaunt their credentials. For instance, I put Geoff D. at the same level as Vincent without the fanfare of posting on a multitude of platforms.

I’m done quibbling about this situation. I've already said that Blu's post was too harsh and I don't endorse it.

P.S.
And as evidenced by discussion with Kris on this page (linking back to concepts first introduced in 2015) I am completely unqualified to discuss video. So I recuse myself from that.

[Waboman]

Now lets get back to what's important. More pole pics!

[Penton-Man]

Originally Posted by Waboman

Now lets get back to what's important. More pole pics!

What, you no like Poms in a video by a DoP who has arguably the most experience shooting feature films with HDR iterations and as compared to other DPs flies way under the personal interview publicity radar?
Here - four poles holding up a water tower lighted up for a good purpose - https://www.hollywoodreporter.com/ne...orkers-1290380

[Penton-Man]

https://www.hollywoodreporter.com/he...huffle-1291805

[Geoff D]

Originally Posted by Penton-Man

absolutely true , but for someone who is as visible and followed as he is on the YouTube and the tweeter he should be held to a higher standard than most

problem is, his followers accept everything he claims as long as it sounds good and they are overwhelmed with technical talk

To be absolutely fair to Vincent that's not exactly a phenom unique to him, especially on a public forum like youtube where the tribalism and confirmation bias is off the charts and actual facts can often be very hard to come by. People would literally say the exact thing about me and my 65K worth of posts on this forum of course, that all the technojabber blinds people to actually holding me accountable for what I say. I even get people that say that they didn't understand a thing that I said but that it sure sounded impressive anyway!

Him and Adam are doing great work in that vast YT cesspool, dealing with facts more than fiction, and, although their views will occasionally get skewed a little (or a lot) according to their own personal tastes, they talk so much sense most of the time that even the spat w/ref to Mando hasn't soured me on them. And with them actually doing 4K reviews now with all that heatmap wizardry I'm gonna be out of a job! Well, I would be if I actually got paid or otherwise employed to write all the claptrap that I do about 4K UHDs.

Originally Posted by Waboman

Now lets get back to what's important. More pole pics!

I wonder: did the Pomeranian pee on the pole?