4K Movies Releases

[Penton-Man]

Originally Posted by raygendreau

http://www.flatpanelshd.com/focus.ph...&id=1435052975

Perceptual Quantizer (PQ) as proposed by Dolby is 0 to 10,000 cd/m2. The test they ran used equipment capable of reaching ,004 cd/m2. I think this caused some confusion and caused some to believe that PQ is .005 to 10,000 cd/m2. Look at the bar at the bottom of the chart in the linked article for the range of PQ.

I think no confusion….but now a concern that I may need new glasses ($$) when it comes to reading 0’s and decimal points on a small screen.

Ray, since I know you like your blacks, to delve into further detail than summarized here….

https://www.youtube.com/watch?v=QLBP...tu.be&t=10m21s , from which the blogger (AV writer/journo) lifted the graph for the preference study….

Specifically (glasses on now), 90% of participants were satisfied with a black level at 0.005 (actually 0.0049 nits to be precise). But, keep in mind that experiment was with a small (23”) panel.

Following that and not revealed in the SMPTE Webcast, another experiment was performed using a 156” screen, something on the order for a small grading theater or a home theater front projector enthusiast . So, given your affinity for black levels, do you know what black level 90% of the participants in that follow-up experiment with the larger screen size were satisfied with? …As there was quite a significant difference….which surprised a little, some of the original investigators.

[Penton-Man]

Originally Posted by Geoff D

Yes, but then he responds with 'why do I want/require that?' so I'm out.

http://www.ign.com/articles/2015/07/...s-at-comic-con

[Geoff D]

Originally Posted by Penton-Man

I think no confusion….but now a concern that I may need new glasses ($$) when it comes to reading 0’s and decimal points on a small screen.

Ray, since I know you like your blacks, to delve into further detail than summarized here….

https://www.youtube.com/watch?v=QLBP...tu.be&t=10m21s , from which the blogger (AV writer/journo) lifted the graph for the preference study….

Specifically (glasses on now), 90% of participants were satisfied with a black level at 0.005 (actually 0.0049 nits to be precise). But, keep in mind that experiment was with a small (23”) panel.

Following that and not revealed in the SMPTE Webcast, another experiment was performed using a 156” screen, something on the order for a small grading theater or a home theater front projector enthusiast . So, given your affinity for black levels, do you know what black level 90% of the participants in that follow-up experiment with the larger screen size were satisfied with? …As there was quite a significant difference….which surprised a little, some of the original investigators.

It wouldn't surprise me if the blacks were measured quite a bit higher on the projection setup than the small monitor, as even in the best theater (SDR) at the moment the blacks seem quite a bit thinner than on my edge-lit LCD TV, never mind on the holy grail OLED.

[Spike M.]

Originally Posted by Penton-Man

Eh…. ^ probably before your time Spike, anyway, YouTube HDR explanation from the Wood-er .

Thanks for the info.

I'm not sure how great of a goal it is to make a TV so bright sensitive that you'd need to wear sunglasses, but it's interesting. Wish I could see it for myself.

[Penton-Man]

Originally Posted by Geoff D

It wouldn't surprise me if the blacks were measured quite a bit higher on the projection setup than the small monitor, as even in the best theater (SDR) at the moment the blacks seem quite a bit thinner than on my edge-lit LCD TV, never mind on the holy grail OLED.

Good logical reasoning, and more significantly, ‘chapeau’ for taking the time and effort to contribute a response in lieu of ray’s absence……..however, keep in mind Dolby engineers can be a clever lot -

Plus, it was an experiment and thusly not restricted by typical entertainment community projector practice for watching motion pictures. So, in this case, for the experimental design, actually a looow (0.002 nits black level) was achieved with a Christie 4K DLP projector (4230) by utilizing a 2.4 neutral density filter which was only used to test the black image set and subsequently removed for the diffuse white and highlights image sets.

So, the projector was actually able to achieve blacker blacks than could the prototype dynamic range display system with its 23” panel pictured at the 10 min. mark of that SMPTE Webcast, which comparatively, was only able to test down to 0.004 nits.

[Penton-Man]

Originally Posted by Spike M.

Thanks for the info.

The info was easy, whereas tracking down that Arlo Guthrie motorcycle song was a bit more difficult because I was blocking his name in my mind.

[Penton-Man]

Originally Posted by Spike M.

I'm not sure how great of a goal it is to make a TV so bright sensitive that you'd need to wear sunglasses, but it's interesting. Wish I could see it for myself.

Well, David was being a bit cautious at the very end of the YouTube clip because prior to the making of that video, he was acutely aware of the landmark HDR viewing tests (quoted and referenced somewhere here on this thread over a year ago, notably before they were ultimately published in the SMPTE Journal this past May), which were performed by both the EBU and EPFL (an academic think tank in Switzerland - https://www.epfl.ch/index.en.html ), which used a 4,000 nit Dolby Pulsar monitor for viewing the HDR test clips and for which….importantly,

the best ambient illumination level for optimal viewing by the participants was as yet unknown????? (Investigators tested at 10…20…and 24nits room lighting levels) and, truth be told, the feedback from some observers was that the combination of the Pulsar and the ambient lighting utilized, caused some of the study participants, eye strain.

BUT, on the other hand, I wouldn’t want to unduly scare folks because one also has to consider *how* (extreme-fair to midland -conservative-subtle) or, to what degree, the material was manually graded/presented in the first place. There are several variables at work (simultaneously) when viewing HDR content, which can impact the total experience. This remains a learning experience for everyone, especially the content creators.

[Richard Paul]

Originally Posted by Penton-Man

Patience ….for perspective, 5 years of discussion was needed to come up with Rec. BT. 2020. Favorable progress has been made with the extended image dynamic range solution for broadcast television, you’ll see.

In fact, both proposals have already been shown (together, in the same room) in demo clips to ‘techfluencers’ outside the hidden recesses of the respective working party/study group.

I think that HDR development is moving at a reasonable speed but I would be disappointed if broadcast HDR isn't supported on the TVs that are released next year.

[Geoff D]

Originally Posted by Penton-Man

Good logical reasoning, and more significantly, ‘chapeau’ for taking the time and effort to contribute a response in lieu of ray’s absence……..however, keep in mind Dolby engineers can be a clever lot -

Plus, it was an experiment and thusly not restricted by typical entertainment community projector practice for watching motion pictures. So, in this case, for the experimental design, actually a looow (0.002 nits black level) was achieved with a Christie 4K DLP projector (4230) by utilizing a 2.4 neutral density filter which was only used to test the black image set and subsequently removed for the diffuse white and highlights image sets.

So, the projector was actually able to achieve blacker blacks than could the prototype dynamic range display system with its 23” panel pictured at the 10 min. mark of that SMPTE Webcast, which comparatively, was only able to test down to 0.004 nits.

Ah, I didn't cotton on that it was an experimental projector setup. You mentioned the level of black that 90% of the participants preferred on the projector, not just what it measured outright, is that the 0.002 figure you mentioned above?

[raygendreau]

The "23 inch panel" used for the Dolby perceptual test was a 6KW digital cinema projector projecting an image on the back of a lcd screen.

[Kris Deering]

Originally Posted by Penton-Man

I think no confusion….but now a concern that I may need new glasses ($$) when it comes to reading 0’s and decimal points on a small screen.

Ray, since I know you like your blacks, to delve into further detail than summarized here….

https://www.youtube.com/watch?v=QLBP...tu.be&t=10m21s , from which the blogger (AV writer/journo) lifted the graph for the preference study….

Specifically (glasses on now), 90% of participants were satisfied with a black level at 0.005 (actually 0.0049 nits to be precise). But, keep in mind that experiment was with a small (23”) panel.

Following that and not revealed in the SMPTE Webcast, another experiment was performed using a 156” screen, something on the order for a small grading theater or a home theater front projector enthusiast . So, given your affinity for black levels, do you know what black level 90% of the participants in that follow-up experiment with the larger screen size were satisfied with? …As there was quite a significant difference….which surprised a little, some of the original investigators.

Black level testing is very hard to do, especially if you are talking about preference testing. Was the testing done in a dark room or a light room? How much time was given in the dark environment before the testing was done? Was there mixed content or was it a true black screen?

Our eyes biasing goes a long way with blacks, especially if you have insane peak whites that were viewed relatively soon before the black. The brighter the image, the more your eyes will hold black at black when you go to a fade out. But give the eyes time to adapt and you will see the problems.

But pure blackouts are extremely rare in real program material for more than a second or two. But it still takes a serious amount of sequential contrast to get the displays black level good enough for very low APL material. This is why displays that cheat by turning off the light source for a black out (laser, LED) are not always the best displays for contrast. They measure well for pure black since they turn the light source off but go one level above black and that jump is huge. I'm more interested in display companies filling in that gap.

[raygendreau]

Here is a better image of the st-2084 curve recently discussed:

http://dci-forum.com/index.php?actio...ach=1594;image

If you are a videophile you aren't going to settle for .005 cd/m2. Videophiles want to see all that they can see. If you've got the rods, you will want .0005 cd/m2.

[raygendreau]

Originally Posted by Kris Deering

Black level testing is very hard to do, especially if you are talking about preference testing. Was the testing done in a dark room or a light room? How much time was given in the dark environment before the testing was done? Was there mixed content or was it a true black screen?

If you listen to the SMPTE webinar at around 10 minutes, Miller provides some parameters for the perceptual test. They had ambient light simulating exit signs in a theater environment.

Also see Figure 6 and narrative for more info:

http://www.mediaandbroadcast.bt.com/...HDTV-final.pdf

[Penton-Man]

Originally Posted by Geoff D

Ah, I didn't cotton on that it was an experimental projector setup. You mentioned the level of black that 90% of the participants preferred on the projector, not just what it measured outright, is that the 0.002 figure you mentioned above?

Correcto-mundo , the projector was able to achieve 0.002nits and turned out 90% of the viewers preferred that black level; whereas nobody was satisfied with a black level of 0.1nits, which satisfied 50% of the viewers in the 23” sized screen experiment.

[Penton-Man]

Originally Posted by raygendreau

The "23 inch panel" used for the Dolby perceptual test was a 6KW digital cinema projector projecting an image on the back of a lcd screen.

I’d thought that self-evident from the explanation and respective slide in the SMPTE Webcast - https://www.youtube.com/watch?v=QLBP...utu.be&t=9m18s

Although, what is neither evident from the slide, nor revealed in discussion is that the display system was calibrated to produce that very wide range….accepting an XYZ input. The color gamut of the system covered the full P3 gamut. However, that particular capability was not critical for testing preferred black and diffuse white levels.

P.S.
Hey, I just saw the water use reported by local O.C. water districts comparing last June to June 2013 and Irvine Ranch Water District slotted in as a water hog . I hope at least you and your neighbors are being water conscious.

[Penton-Man]

Originally Posted by Kris Deering

Was the testing done in a dark room or a light room?

Dark, as I mentioned on the last page… to be precise, in total, the ambient lighting of the room was < 3 lux, (giving < 0.0001 nits.

Originally Posted by Kris Deering

How much time was given in the dark environment before the testing was done? Was there mixed content or was it a true black screen? Our eyes biasing goes a long way with blacks...

The goal was to test viewer preferences based on shorter term adaptation simulating typical motion picture viewing as realistically as possible. The experiment started with a one minute adaptation phase allowing the HVS of the participant to adjust to the environment. Then, to describe it as simple as possible for everyone reading, all possible stimulus levels of the given image were shown (sent 10bit RGB 4:4:4) in a consecutive run at 4 sec. intervals starting from mid-level Log Average Luminance down to the display minimum and then back up to the least *black* image.

Each trial consisted of multiple steps, based on user response which is just too complicated and time consuming for me to go into, at least for right now.

[Penton-Man]

Originally Posted by raygendreau

If you listen to the SMPTE webinar at around 10 minutes, Miller provides some parameters for the perceptual test. They had ambient light simulating exit signs in a theater environment.

Also see Figure 6 and narrative for more info:
http://www.mediaandbroadcast.bt.com/...HDTV-final.pdf

^ Excellent references but doesn’t contain what I described above for Kris.

[Penton-Man]

Originally Posted by Kris Deering

Was there mixed content or was it a true black screen?

The stimulus image filled the entire 23” screen, like so –

[Penton-Man]

Originally Posted by Penton-Man

stimulus levels of the given image were shown (sent 10bit RGB 4:4:4) in a consecutive run...

Something like so -

[Penton-Man]

https://forum.blu-ray.com/showthread...t#post11095874
^
A feather in the cap for the BBC boys….http://advanced-television.com/2015/...-prize-at-ibc/