Spears & Munsil 4K UHD HDR Benchmark (Test Discs) May 15, 2023

[INdetectableMAN]

Finally, my two units of the new edition arrived, which are already together with the other two of the previous edition.

The shipment arrived without customs incidents and I did not have to pay customs taxes.

The first thing will be to test the matter of real images and skin tones, and then measure DV in absolute mode with Calman and ChromaPure to see if everything works as expected.

I will test on Panasonic and LG OLED with Sony X700, Panasonic UB9000 and UB820 players

Thank you for the magnificent work done.

[Stacey Spears]

For the 10K montage, HDR10 and DV play the exact same file on disc. Only difference is DV enables the MEL layer when playing.

For skin tones, DV, HDR10+ and HDR10 are all different files. Skin tone uses FEL. HDR10 and 10+ should be the same encoded file, but 10+ is a copy of the file with the metadata embedded.

Will be curious what you think about the Graded vs. Ungraded and HDR Analyzer as well.

[Geoff D]

Originally Posted by Stacey Spears

For the 10K montage, HDR10 and DV play the exact same file on disc. Only difference is DV enables the MEL layer when playing.

For skin tones, DV, HDR10+ and HDR10 are all different files. Skin tone uses FEL. HDR10 and 10+ should be the same encoded file, but 10+ is a copy of the file with the metadata embedded.

Will be curious what you think about the Graded vs. Ungraded and HDR Analyzer as well.

Interesting, I remember you saying something before about how the HDR10 and DV montages on the first UHD had to have separate encodes, even for the DV's HDR10 base layer, howing to how the Dobly encoding worked, that you could only have a base layer created by the DV engine? Or is that just for FEL?

[Stacey Spears]

Originally Posted by Geoff D

Interesting, I remember you saying something before about how the HDR10 and DV montages on the first UHD had to have separate encodes, even for the DV's HDR10 base layer, howing to how the Dobly encoding worked, that you could only have a base layer created by the DV engine? Or is that just for FEL?

That is usually just for FEL. On the last disc, I did not encode the DV stuff as it was done by Pixelogic, so it was a different encode than HDR10 10K.

[INdetectableMAN]

Since I just received it today, I haven't been able to test it as thoroughly as I would like, and it's an album that takes time, but from what I've seen it seems like a fantastic album that will exceed all my expectations. A great job.

The skin tones section seems to me, for lack of a better word, wonderful.

Originally Posted by Stacey Spears

More notes.

DTM
The skin tone content has a real color checker chart next to each model. The lower left chip is the brightest chip. It should measure somewhere around 132-134 nits. We have seen that dynamic tone mapping has raised that chip to over 200 nits, which is incorrect.

I've done a little test on my Panasonic OLED, and I've measured the bottom left frame in DV mode, DV Dark mode on Panasonic.

It has measured 133 nits

[Stacey Spears]

Originally Posted by INdetectableMAN

Since I just received it today, I haven't been able to test it as thoroughly as I would like, and it's an album that takes time, but from what I've seen it seems like a fantastic album that will exceed all my expectations. A great job. :roca:

The skin tones section seems to me, for lack of a better word, wonderful.

I've done a little test on my Panasonic OLED, and I've measured the bottom left frame in DV mode, Dark mode on Panasonic.

It has measured 133 nits

Thank you for the feedback and 133 is right on target!

Does Panasonic have a DTM mode for HDR10? If so, I would be curious to now what it measures, when you have time, of course.

[DisplayCalNoob]

Originally Posted by Stacey Spears

I have never asked them how many ms it takes to process a frame. I will ask them and see if I can get an answer.

Are you experiencing A/V sync issues with DV?

Anything processing that takes longer than the time they have to present in milliseconds (ms) will introduce delay in the video. Any video processing algorithm can do this if not fast enough.

With my new Pioneer VSX-935 audio receiver, the difference is only .5 ms, my older Sony had was even greater. There's an additional. 5 ms with HDMI chroma subsampling on(TV to player, second HDMI port on player used for audio only).

[wxman2003]

Originally Posted by Stacey Spears

Thank you for the feedback and 133 is right on target!

Does Panasonic have a DTM mode for HDR10? If so, I would be curious to now what it measures, when you have time, of course.

Has there ever been discussion amongst the experts about setting luminance based on distance from the tv? For example, the meter reads 130 nits sitting on the tv. And as you move further away, light diffuses and 130 nits may be 100 nits at 4 feet, and 70 nits at 10 feet from the screen for a random example. Not saying those are actual numbers, but it will not appear as bright the further you sit from the tv. So if one wants the same effect at 10 feet that you get at 4 feet, would calibrating the tv to 200 nits for 10 feet away give you that effect as if you sat 4 feet from the tv and calibrated to 130 nits? I always wondered if there was some sort of equation to take into account distance and diffusion.

[Stacey Spears]

Originally Posted by wxman2003

Has there ever been discussion amongst the experts about setting luminance based on distance from the tv? For example, the meter reads 130 nits sitting on the tv. And as you move further away, light diffuses and 130 nits may be 100 nits at 4 feet, and 70 nits at 10 feet from the screen for a random example. Not saying those are actual numbers, but it will not appear as bright the further you sit from the tv. So if one wants the same effect at 10 feet that you get at 4 feet, would calibrating the tv to 200 nits for 10 feet away give you that effect as if you sat 4 feet from the tv and calibrated to 130 nits? I always wondered if there was some sort of equation to take into account distance and diffusion.

Not that I am aware of. I have asked about screen size, such as a theater, and light output. e.g. 10k nits on a 40' screen is much different than 10K nits on a 40" screen.

That sounds like a great question to ask when we do the stream with Dolby! Please remind me. I will post about each one before to gather some questions.

[Stacey Spears]

Still thinking about the custom case that holds all four discs, which would include two new sleeves for discs 1 and 2 to include the 1 and 2 on the spine. Would a new black case for discs 1 and 2 be interesting so they all match or just have blue and black mixed? OCD anyone?

If we included a USB stick with 8K patterns, would that be interesting? Not a complete set, but a few dozen maybe. Since there is no 8K disc format, it would be more interesting.

Embossed or sticker for artwork if we did a steelbook style? I like this visual style with the rounded corners and clear spots for each disc.



We are thinking we would do something Kickstarter style. We would open it up for a month. Then order an amount slightly larger than the pre-order itself. It would be a one time event.

We will hold off for a while as we ponder what to do. Any interest in something like this? You can PM if you like or post on here. Jason brought up the My So Called Life lunchbox, but I would prefer something that goes on the shelf where I can see the spines. We could even offer steelbook cases for each that are empty and you move your discs into.

Any and all feedback is welcome.

[INdetectableMAN]

Today I was able to do one more small test, this time with the LG C2.

First I measured Grabiella's white box in HDR10 Cinema mode calibrated with Calman, and it measures 140 nits, which is fine. However, with DTM enabled it goes to 180 nits, which is excessive.

Then I measured in Dolby Vision Cine calibrated 1D LUT with the settings file loaded (670 nits), and the Gabrilla frame measures just 119 nits.

after tried DV analysis patterns in absolute mode using Calman and ChromaPure and it seems to work very well.






I've taken the DV measurements based on a calibration I had done some time ago, and now all I've done is reload the configuration file with the correct values.

In short, after calibrating the LG in DV, checking in absolute mode, everything seems to be fine, but then the Grabiella box is below target. It's not much, but it's down.

The color temperature pattern seems like a great idea to me, I have used medialight's own D65 light reflected on a neutral wall as an optical comparator and it works well. An optical comparator will be better, but a D65 light placed strategically behind the screen can be very useful to use the pattern.

Originally Posted by Stacey Spears

Thank you for the feedback and 133 is right on target!

Does Panasonic have a DTM mode for HDR10? If so, I would be curious to now what it measures, when you have time, of course.

Yes, my Panasonic has a DTM, and tomorrow I am going to try what you say and I will even measure the Grabiella frame, which is the one I have taken as a reference, in HDR10+

[Stacey Spears]

Originally Posted by INdetectableMAN

Today I was able to do one more small test, this time with the LG C2.

First I measured Grabiella's white box in HDR10 Cinema mode calibrated with Calman, and it measures 140 nits, which is fine. However, with DTM enabled it goes to 180 nits, which is excessive.

Then I measured in Dolby Vision Cine calibrated 1D LUT with the settings file loaded (670 nits), and the Gabrilla frame measures just 119 nits.

after tried DV analysis patterns in absolute mode using Calman and ChromaPure and it seems to work very well.





I've taken the DV measurements based on a calibration I had done some time ago, and now all I've done is reload the configuration file with the correct values.

In short, after calibrating the LG in DV, checking in absolute mode, everything seems to be fine, but then the Grabiella box is below target. It's not much, but it's down.

The color temperature pattern seems like a great idea to me, I have used medialight's own D65 light reflected on a neutral wall as an optical comparator and it works well. An optical comparator will be better, but a D65 light placed strategically behind the screen can be very useful to use the pattern.

Yes, my Panasonic has a DTM, and tomorrow I am going to try what you say and I will even measure the Grabiella frame, which is the one I have taken as a reference, in HDR10+

Let me check the source for each model and confirm the levels of the patch. I may have just spot checked as the lighting really should not have changed, but I will confirm. It was a fair amount of time between each model.

We have thought about writing up a short paper describing exactly what you did using a gray card and light source. I have been asking Jason to make me a keychain D65 light. Of course I also wanted him to make D65 bulbs for my RAV4. I replaced all the interior with LEDs, but they are not the best.

Thank you for the feedback on the color temp pattern. I wanted to do something different as everyone does the same thing, including us, by just having gray steps.

Per Phil Holland, who was DP/Director of the skin tones, we did not use the D65 patch to white balance. We used the one to the left of it as it is better for skin tones.

Phil has agreed to do a video/interview and we will discuss the skin tones, the LA skyline shot in the montage and anything else people might be interested in. For those that don't know, Phil shot the Apple TV screensavers. He did not dive or go to space, but all of the aeriel was him traveling the globe. He has some great stories, maybe we can discuss that. His 12k content on YouTube is another fun topic.

[INdetectableMAN]

Originally Posted by Stacey Spears

Let me check the source for each model and confirm the levels of the patch. I may have just spot checked as the lighting really should not have changed, but I will confirm. It was a fair amount of time between each model.

We have thought about writing up a short paper describing exactly what you did using a gray card and light source. I have been asking Jason to make me a keychain D65 light. Of course I also wanted him to make D65 bulbs for my RAV4. I replaced all the interior with LEDs, but they are not the best.

Thank you for the feedback on the color temp pattern. I wanted to do something different as everyone does the same thing, including us, by just having gray steps.

Per Phil Holland, who was DP/Director of the skin tones, we did not use the D65 patch to white balance. We used the one to the left of it as it is better for skin tones.

Phil has agreed to do a video/interview and we will discuss the skin tones, the LA skyline shot in the montage and anything else people might be interested in. For those that don't know, Phil shot the Apple TV screensavers. He did not dive or go to space, but all of the aeriel was him traveling the globe. He has some great stories, maybe we can discuss that. His 12k content on YouTube is another fun topic.

Hi Stacey

I have measured on my Panasonic as you asked, and both in calibrated HDR10, with DTM on/off and HDR10+, my Panasonic measures the same in the bottom left white box of the table. It measures 128 nits in HDR10, DTM ON/OFF and HDR10+, the same.

I've also been measuring with all the models, and the bottom left box measures essentially the same on all of them. And I have entered and exited to avoid that the static table could activate some type of protection that, by slightly attenuating the image, could condition the measurements.

HDR10+


DTM ON


have also measured my Panasonic in absolute DV mode and everything is perfect, it tracks PQ EOTF correctly.

PANASONIC DV absolute mode


However, the LG in the lower left white frame for skin tones is 119 nits and the Panaonic 133, so I think DV tonemapping will do different things between Panasonic and LG, as on LG the picture is a bit darker.

[mrtickleuk]

Originally Posted by Stacey Spears

We have thought about writing up a short paper describing exactly what you did using a gray card and light source. I have been asking Jason to make me a keychain D65 light. Of course I also wanted him to make D65 bulbs for my RAV4.

Ha, that's brilliant. That single sentence has made my day

Originally Posted by Stacey Spears

Phil has agreed to do a video/interview and we will discuss the skin tones, the LA skyline shot in the montage and anything else people might be interested in. For those that don't know, Phil shot the Apple TV screensavers. He did not dive or go to space, but all of the aeriel was him traveling the globe. He has some great stories, maybe we can discuss that. His 12k content on YouTube is another fun topic.

Marvellous, I look forward to it!

I assume he's
https://www.youtube.com/@phfx

and not
https://www.youtube.com/channel/UCkF...0YWNzVctjvLSIg

?

[Stacey Spears]

Originally Posted by mrtickleuk

I assume he's
https://www.youtube.com/@phfx

and not
https://www.youtube.com/channel/UCkF...0YWNzVctjvLSIg

?

You are correct sir! https://www.phfx.com/

Here is the article on the 12K shot using three cameras.
https://www.phfx.com/articles/Filmin...lsOverNewYork/

[Wendell R. Breland]

Originally Posted by Stacey Spears

How do I post a link to a YouTube video on this forum without it converting to a browser choice page?

Not mrtickleuk but the best way I have found: for the desired YT video accessed, copy the address from the browser address bar then use the “insert link” on your post here.

[TbeRw01]

Originally Posted by Stacey Spears

Thank you for sharing. I have forwarded the video to the vision scientists at Dolby to get their opinion. I am not sure if they will share their thoughts with me. I don't use any of those devices myself.

I am also planning to send my ATV measurements of YCbCr and RGB to Vincent in hopes he will follow-up and possibly report on the issue. Apple actually listens to Vincent. Every CE company has someone who influences them. e.g. Vizio will do what every David K. says at CNET. At least they used to, have not spoken to Vizio in a while.

I can confirm that the Sony X700 does true TV-LED and the Google Chromecast does indeed ''fake'' TV-LED (the rpu metadata changes are done by the player).
https://slow.pics/c/3sARstln

So it seems to be an issue with the Android box.

[mrtickleuk]

DTLA ORBIT in 8K - YouTube

This is using youtube.com and not the youtu.be which Youtube's "share link" feature gives.

[Stacey Spears]

Originally Posted by mrtickleuk

This is using youtube.com and not the youtu.be which Youtube's "share link" feature gives.

Thank you, youtu.be vs. youtube.com was/is the issue.

It looks like I can't get the exact timecode to work and also show a preview in the post. The shot starts around 1:57 into the clip.

https://www.youtube.com/embed/Mn24cQLqAE4?start=116

[TbeRw01]

Originally Posted by TbeRw01

I can confirm that the Sony X700 does true TV-LED

Sony X800m2 is also correct.

[Show spoiler]

Dolby Vision TRUE vs FAKE TV-LED RGB Sony X800m2 vs GTV Chromecast - YouTube