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

[Scenic Labs]

That's roughly what I estimate the add-on would cost. It would only be available from dealers to avoid people buying on Amazon, etc. and finding that the main disc is missing. If any potential filter materials are especially exotic (or if we need multiple filters for different displays), it might cost a bit more.

Originally Posted by Stacey Spears

(Should be less than $10 if all goes well) Jason may kill me for saying that. At the time the add-on comes out, the dual disc set will go up to $47.95. If we can make a filter work, then it would also come with the add-on. While we are targeting early next year, we are waiting on Dolby 4.0 encoder tools that support profile 7.

[Stacey Spears]

Originally Posted by Penton-Man

P.S.
Good to see you back

Meant to ask, do you ever attend the MITC meetings?

[Penton-Man]

Originally Posted by Stacey Spears

Had a kidney stone to deal with.

I wanted to use superscript for the 2, but did not see it as an option in the forum tools. Even my courier, which is a fixed width, did not format correctly. I used that to align the values. On the disc, we do use superscript for the 2.

wrt to the stone, I presume the smiley face indicates it’s been successfully passed, zapped (litho) or scoped

[Penton-Man]

Originally Posted by Stacey Spears

Meant to ask, do you ever attend the MITC meetings?

yes – https://forum.blu-ray.com/showthread...c#post16156498
even saw Phil Holland at one of them….or at least, I think it was him

But anyway, I feel it’s frowned upon among some folks by me speaking/writing about the meetings(s) until something has been formally published, e.g. -

Originally Posted by Penton-Man

MITC Progress Report 2018, now (as of just this week) publicly available to all -
https://cms-assets.theasc.com/SMPTE-...20181005133408

so I’ve been hesitant to talk about them since

[koberulz]

Originally Posted by Stacey Spears

1. His DV patterns are profile 5, not profile 7 (UHD BD).
2. According to his notes, I assume his L1 is created by Resolve.
3. MEL based patterns. (10-bit) Fine for window patterns.
4. IPT encoded. BD uses YCbCr.

He has stayed on top of new stuff with one of the first UHD HDR discs, HLG patterns and now DV. He also has a CalMAN workflow to go with it, which is nice.

What does all that mean in English? I have no idea what profiles 5 and 7 are, or what L1, MEL, and IPT are.

[Geoff D]

The profile he uses is for streaming DV, not for disc.

L1 refers to the Level 1 automated metadata pass carried out when first analysing content in DV. Level 2 (L2) is the manual trim pass whereby the metadata is adjusted to achieve the best possible downconversion. Resolve refers to a suite of grading software which actually has this L1 automated metadata pass included for free with certain packages.

MEL is minimum enhancement layer, a process for DV which only includes the tone mapping metadata and is not the full enhancement layer (FEL) which is difference data derived from the source master and rebuilds the image with full 12-bit fidelity. Streaming DV is essentially MEL by default, on disc then it depends on the studio as to what version they use, if at all.

ITP (also ICtCp) refers to the proprietary colour space created by Dolby to use with HDR, as conventional YCbCr encoding was not designed for use with HDR and can display chroma crosstalk errors as a result. Streaming DV is usually encoded with this process, disc DV uses YCbCr as per the UHD Blu spec.

In other words: Ryan's patterns are 'Dolby Vision' but they don't necessarily replicate how a setup will perform with DV on disc. Alas, neither will Stacey's disc as it doesn't include any DV test patterns, only a video montage.

[Geoff D]

Pertaining to the above:

Stacey, could you let us know what parts of the montage hit whatever nit levels? Not second by second or anything, but anywhere where it exceeds 2000 or 4000 nits perhaps?

[Scenic Labs]

The discs just cleared customs, and since it's still pretty early, I'm hoping that they will be delivered to us today.

[Stacey Spears]

Originally Posted by Geoff D

Pertaining to the above:

Stacey, could you let us know what parts of the montage hit whatever nit levels? Not second by second or anything, but anywhere where it exceeds 2000 or 4000 nits perhaps?

Good idea. I have a shot list in an xlsx that I assembled. The plan is to fill out stuff like location, lens, f-stop, mm, compression ratio, etc... I can look at adding peak nits per shot.

For the add-on, one of the new montages will be the HDR Analyzer version. This will allow me to capture the info you want. I should have a first version of this in a couple of weeks encoded as SDR. Then once I get the Analzyer, and FW update from AJA/CF, I will create an HDR version. Just need to figure out which version of the WFM I want on the upper left. I have asked Bill F. for a color volume graph as well.

Of course, this will be for the version that we call 10,000. Here is a quick graph that Mathias (madVR) ran for me back in December. This uses the same technique as maxCLL uses, which is not a true nit value.

You can use the frame numbers in the xls to align with the frame numbers along the graph below.



Here are some actual analyzer captures from Transkoder. These are just single frame graphs from each shot. Does not mean its the peak. The first shot is a duplicate of the second in this batch. I have posted these before, but for some reason this forum and dropbox don't always work well together.

What are others using for posting images?

[Penton-Man]

Originally Posted by Geoff D

The profile he uses is for streaming DV, not for disc.

L1 refers to the Level 1 automated metadata pass carried out when first analysing content in DV. Level 2 (L2) is the manual trim pass whereby the metadata is adjusted to achieve the best possible downconversion. Resolve refers to a suite of grading software which actually has this L1 automated metadata pass included for free with certain packages.

MEL is minimum enhancement layer, a process for DV which only includes the tone mapping metadata and is not the full enhancement layer (FEL) which is difference data derived from the source master and rebuilds the image with full 12-bit fidelity. Streaming DV is essentially MEL by default, on disc then it depends on the studio as to what version they use, if at all.

ITP (also ICtCp) refers to the proprietary colour space created by Dolby to use with HDR, as conventional YCbCr encoding was not designed for use with HDR and can display chroma crosstalk errors as a result. Streaming DV is usually encoded with this process, disc DV uses YCbCr as per the UHD Blu spec.

In other words: Ryan's patterns are 'Dolby Vision' but they don't necessarily replicate how a setup will perform with DV on disc. Alas, neither will Stacey's disc as it doesn't include any DV test patterns, only a video montage.

‘only’ Geoff?
Heck if the montage includes any flowers as Stacey indicated - https://forum.blu-ray.com/showthread...s#post16225098 that’s the only thing that I could get my wife to sit down and watch at least a portion of the disc with me

Note to koberulz, everything that Geoff explains has been discussed (in detail) at one time or another on the HDR Discussion thread, months….years ago, e.g. ICtCp when it was being developed, Delta ICtCp, L2 before/at the time of launch, etc.

If need more detail, can search with respective keywords.

[Stacey Spears]

Originally Posted by koberulz

What does all that mean in English? I have no idea what profiles 5 and 7 are, or what L1, MEL, and IPT are.

Geoff did a great job answering in detail. I am sorry I did not, I sort of assumed everyone has dove in head first and knows everything about HDR.

The short answer is that his stuff will work for streaming or USB, not for optical disc. His stuff should go through a similar path as Netflix and Amazon. But, it will not be the same path for disc, especially for someone using LL mode.

[DisplayCalNoob]

Originally Posted by Stacey Spears

David tested an LG player during authoring and said it had a lot of banding in the player. Not sure if it sends LL or tunneled to the display. I have personally never seen an LG player. It is Mediatek based I believe.

I don't have any banding issues in DV content, some crop in HDR10 content. The posterization issue was only visible in a DV stream of Guardians of the Galaxy Vol.2, yellow had some annoying posterization.

[Stacey Spears]

Originally Posted by Penton-Man

wrt to the stone, I presume the smiley face indicates it’s been successfully passed, zapped (litho) or scoped

Yes.

[Stacey Spears]

Originally Posted by Geoff D

The profile he uses is for streaming DV, not for disc.

L1 refers to the Level 1 automated metadata pass carried out when first analysing content in DV. Level 2 (L2) is the manual trim pass whereby the metadata is adjusted to achieve the best possible downconversion. Resolve refers to a suite of grading software which actually has this L1 automated metadata pass included for free with certain packages.

For DV 4.0, L8 is the new L2. L8 has all of the new trim pass controls including secondaries.

[Wendell R. Breland]

Originally Posted by Stacey Spears

I wanted to use superscript for the 2, but did not see it as an option in the forum tools. Even my courier, which is a fixed width, did not format correctly. I used that to align the values. On the disc, we do use superscript for the 2.

I will ask the mods about who has the privilege of posting with html codes turned on. I do find many notations written in WordPerfect will translate correctly when posted here, as in:

600 cd/m² – 674
a ≥ b
600Ω load

Have never tried this using Word or other programs.

[Stacey Spears]

Originally Posted by Wendell R. Breland

I will ask the mods about who has the privilege of posting with html codes turned on. I do find many notations written in WordPerfect will translate correctly when posted here, as in:

600 cd/m² – 674
a ≥ b
600Ω load

Have never tried this using Word or other programs.

Word did not work, but copy and pasting from your post did, so thank you! BTW, WordPefect still exists or are you just running a really old copy?

[Vilya]

How easy will this test disc be to use for those of us that are not extremely knowledgeable about all of these fine points? Will this disc be of benefit to a less informed 4K TV owner or will it instead make them feel woefully ignorant?

[Geoff D]

Why not just hold down Alt then press 253, it gets me one of these ² just fine. 0179 is for a ³

[Robert Zohn]

Just got notification that our 1st batch of Spears & Munsil HD Benchmark 2nd Edition discs will be shipping by Monday.

[Geoff D]

Originally Posted by Stacey Spears

Of course, this will be for the version that we call 10,000. Here is a quick graph that Mathias (madVR) ran for me back in December. This uses the same technique as maxCLL uses, which is not a true nit value.

You can use the frame numbers in the xls to align with the frame numbers along the graph below.

Jesus, it regularly peaks at well over 4k nits. That montage is a beast.