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

[SeeMoreDigital]

Originally Posted by EddieLarkin

Just bought my upgrade copy via the UK site (https://www.biaslighting.co.uk/produ...benchmark-2023), came to £27.93 with shipping. You have to get your discount code via the form linked to above on the US site (https://www.biaslighting.com/pages/s...pgrade-program).

I'm a bit nervous buying from them. If you click on their 'Contact Us' link and look up their address (ie: 7 Union Street, Carmarthen, SA31 3DE, UK), 'Street View' reveals a small lime green painted terraced house with a red front door!

[Ben_R]

So despite this;

Originally Posted by Stacey Spears

That is the official distributor of the disc. All dealers acquire through them.

You post this;

Originally Posted by SeeMoreDigital

I'm a bit nervous buying from them. If you click on their 'Contact Us' link and look up their address (ie: 7 Union Street, Carmarthen, SA31 3DE, UK), 'Street View' reveals a small lime green painted terraced house with a red front door!

Never heard of a dodgy site accepting PayPal

[panasonicst60]

Originally Posted by SeeMoreDigital

I'm a bit nervous buying from them. If you click on their 'Contact Us' link and look up their address (ie: 7 Union Street, Carmarthen, SA31 3DE, UK), 'Street View' reveals a small lime green painted terraced house with a red front door!

Not sure what's up with the address, but they are legit. Excellent customer service too!

[David M]

Originally Posted by rickardl

Re source direct: have you seen any artefacts or non-defeatable processing taking place with all relevant settings at default or 0?

Not Stacey, but Panasonic's approach to chroma upsampling has been pretty widely documented. It can cause minor colored ringing around sudden chroma transitions. I've never actually seen the effects of it on real live action content but I imagine there'll be something that makes it visible... and I think their approach to chroma upsampling still has benefits of its own despite that side effect.

[Stacey Spears]

Originally Posted by David M

Not Stacey, but Panasonic's approach to chroma upsampling has been pretty widely documented. It can cause minor colored ringing around sudden chroma transitions. I've never actually seen the effects of it on real live action content but I imagine there'll be something that makes it visible... and I think their approach to chroma upsampling still has benefits of its own despite that side effect.

They appear to use something similar to bicubic while everyone else uses bilinear.

[Stacey Spears]

Originally Posted by SeeMoreDigital

I'm a bit nervous buying from them. If you click on their 'Contact Us' link and look up their address (ie: 7 Union Street, Carmarthen, SA31 3DE, UK), 'Street View' reveals a small lime green painted terraced house with a red front door!

The US site is the distributor of the disc. All discs get replicated and flow through them. They took over from OPPO eight or nine years ago. Maybe longer, hard to track time before C19 as it is all just a blur now. Not sure the full relationship with the UK site.

[TbeRw01]

Originally Posted by Stacey Spears

I believe I have shared this before. .

Hi, two questions..
I guess you did DV in CM v4.0 and then delivered in CM v2.9?

Do you know why all the CM v4.0 movie/tv show (mostly from streaming) only have a 100 nits L8 trim but have a 600 and 1000nits L2 trim?

Since L2 has no effect at all on CM v4.0 hardware, I don't understand why no L8 600/1000nits trims, maybe you know?


example:

[Show spoiler] {
"Level1": {
"min_pq": 7,
"max_pq": 2675,
"avg_pq": 1229
}
},
{
"Level2": {
"target_max_pq": 2081,
"trim_slope": 1663,
"trim_offset": 2096,
"trim_power": 1656,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 1628,
"ms_weight": 2048
}
},
{
"Level2": {
"target_max_pq": 2851,
"trim_slope": 1989,
"trim_offset": 2077,
"trim_power": 1775,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 1949,
"ms_weight": 2048
}
},
{
"Level2": {
"target_max_pq": 3079,
"trim_slope": 2009,
"trim_offset": 2072,
"trim_power": 1849,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 1987,
"ms_weight": 2048
}
},
{
"Level4": {
"anchor_pq": 691,
"anchor_power": 539
}
},
{
"Level5": {
"active_area_left_offset": 0,
"active_area_right_offset": 0,
"active_area_top_offset": 0,
"active_area_bottom_offset": 0
}
},
{
"Level6": {
"max_display_mastering_luminance": 4000,
"min_display_mastering_luminance": 50,
"max_content_light_level": 0,
"max_frame_average_light_level": 0
}
}
]
},
"cmv40_metadata": {
"num_ext_blocks": 4,
"ext_metadata_blocks": [
{
"Level3": {
"min_pq_offset": 2048,
"max_pq_offset": 2048,
"avg_pq_offset": 1945
}
},
{
"Level8": {
"length": 13,
"target_display_index": 1,
"trim_slope": 1954,
"trim_offset": 2095,
"trim_power": 2018,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 1628,
"ms_weight": 2048,
"target_mid_contrast": 2048,
"clip_trim": 1964
}
},
{
"Level9": {
"length": 1,
"source_primary_index": 0
}
},
{
"Level254": {
"dm_mode": 0,
"dm_version_index": 2
}
}
]
}
},
"rpu_data_crc32": 1664391000

[Stacey Spears]

Originally Posted by TbeRw01

Hi, two questions..
I guess you did DV in CM v4.0 and then delivered in CM v2.9?

Do you know why all the CM v4.0 movie/tv show (mostly from streaming) only have a 100 nits L8 trim but have a 600 and 1000nits L2 trim?

Since L2 has no effect at all on CM v4.0 hardware, I don't understand why no L8 600/1000nits trims, maybe you know?


example:

[Show spoiler] {
"Level1": {
"min_pq": 7,
"max_pq": 2675,
"avg_pq": 1229
}
},
{
"Level2": {
"target_max_pq": 2081,
"trim_slope": 1663,
"trim_offset": 2096,
"trim_power": 1656,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 1628,
"ms_weight": 2048
}
},
{
"Level2": {
"target_max_pq": 2851,
"trim_slope": 1989,
"trim_offset": 2077,
"trim_power": 1775,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 1949,
"ms_weight": 2048
}
},
{
"Level2": {
"target_max_pq": 3079,
"trim_slope": 2009,
"trim_offset": 2072,
"trim_power": 1849,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 1987,
"ms_weight": 2048
}
},
{
"Level4": {
"anchor_pq": 691,
"anchor_power": 539
}
},
{
"Level5": {
"active_area_left_offset": 0,
"active_area_right_offset": 0,
"active_area_top_offset": 0,
"active_area_bottom_offset": 0
}
},
{
"Level6": {
"max_display_mastering_luminance": 4000,
"min_display_mastering_luminance": 50,
"max_content_light_level": 0,
"max_frame_average_light_level": 0
}
}
]
},
"cmv40_metadata": {
"num_ext_blocks": 4,
"ext_metadata_blocks": [
{
"Level3": {
"min_pq_offset": 2048,
"max_pq_offset": 2048,
"avg_pq_offset": 1945
}
},
{
"Level8": {
"length": 13,
"target_display_index": 1,
"trim_slope": 1954,
"trim_offset": 2095,
"trim_power": 2018,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 1628,
"ms_weight": 2048,
"target_mid_contrast": 2048,
"clip_trim": 1964
}
},
{
"Level9": {
"length": 1,
"source_primary_index": 0
}
},
{
"Level254": {
"dm_mode": 0,
"dm_version_index": 2
}
}
]
}
},
"rpu_data_crc32": 1664391000

Everything was done in 2.9.

As far as L8 at 100 and L2 at 600 and 1000, sounds like they were either lazy or messed up. I recall hearing that for one of the Star Wars films they did not want to do any trims, it got rejected and they had to go back and do it.

We did a 100, 600, 1000 and 2000 trim on the montage. At least I recall that, the new montage was re-graded in in 2020, so its all fuzzy at this point. The test patterns have no trim per the DV calibration spec.

[Geoff D]

Originally Posted by Stacey Spears

Everything was done in 2.9.

As far as L8 at 100 and L2 at 600 and 1000, sounds like they were either lazy or messed up. I recall hearing that for one of the Star Wars films they did not want to do any trims, it got rejected and they had to go back and do it.

We did a 100, 600, 1000 and 2000 trim on the montage. At least I recall that, the new montage was re-graded in in 2020, so its all fuzzy at this point. The test patterns have no trim per the DV calibration spec.

Most likely Last Jedi as colourist Walter Volpatto doesn’t like setting trims, or indeed metadata in general, because his view is that the grade is the grade is the grade and it should not be set to different levels. I heard the same story about QC rejecting their first attempt, though I’ve read on John Daro’s blog that Dobly have since changed their practices so you’re not forced to set trims if you don’t want to?

[TbeRw01]

Originally Posted by Geoff D

Most likely Last Jedi as colourist Walter Volpatto doesn’t like setting trims, or indeed metadata in general, because his view is that the grade is the grade is the grade and it should not be set to different levels. I heard the same story about QC rejecting their first attempt, though I’ve read on John Daro’s blog that Dobly have since changed their practices so you’re not forced to set trims if you don’t want to?

AFAIK, trims always been optional and plenty of movies don't have any or only have a 100-nits trim.

I just checked all the DV metadata for the Star Wars movies and:

Star.Wars.Episode.VIII.The.Last.Jedi.2017
100/600/1000nits trims (they did the trims static though)

Star.Wars.Episode.VII.The.Force.Awakens.2015
100/1000nits trims

Star.Wars.Episode.IX.The.Rise.of.Skywalker.2019
only a 100nits trim

Rogue.One.A.Star.Wars.Story.2016(disney plus)
only a 100nits trim

Solo.A.Star.Wars.Story.2018(disney plus)
only a 100nits trim

Star.Wars.Episode.I.The.Phantom.Menace.1999
no trims, not even 100nits

Star.Wars.Episode.II.Attack.of.the.Clones.2002
no trims, not even 100nits

Star.Wars.Episode.V.The.Empire.Strikes.Back.1980
no trims, not even 100nits

Star.Wars.Episode.VI.Return.of.the.Jedi.1983
no trims, not even 100nits

Star.Wars.Episode.IV.A.New.Hope.1977
no trims, not even 100nits


L1-L2 plot of The Last Jedi BD:

[Show spoiler]





**** I converted my rip to Profile 8 but that doesn't change anything related to the metadata itself.

[Stacey Spears]

I consider the trim pass to be the secret weapon of DV. It allows the colorist to control the tone mapping on lower nit displays. The tone mapping built into displays are all over the place.

But, I suspect it comes down to time and money. It takes more time, which costs more money, to do trims.

ICtCp is the second best part of DV, too bad it is not supported on BD. I would have loved to use it. It fixes artifacts in chroma caused that YCbCr introduces that are not in the RGB master.

[RocShemp]

This whole trim talk is flying over my head. Anyone kind enough to point me towards some recommended reading on the subject?

[David M]

Originally Posted by RocShemp

This whole trim talk is flying over my head. Anyone kind enough to point me towards some recommended reading on the subject?

Don't have any good links, but it just refers to the Dolby Vision metadata allowing the colorist to adjust how the encoded pixel data is tone mapped for different displays.

For example, if a scene containing highly saturated, high-luminance flames (for example) looks correct on the 4000 nit studio monitor in the grading suite, but looks unnaturally colored on the 1000 nit monitor, the colorist can add a correction in the DoVi metadata to change how the 4000-to-1000 tone mapping process is done for that particular shot.

[RocShemp]

Originally Posted by David M

Don't have any good links, but it just refers to the Dolby Vision metadata allowing the colorist to adjust how the encoded pixel data is tone mapped for different displays.

For example, if a scene containing highly saturated, high-luminance flames (for example) looks correct on the 4000 nit studio monitor in the grading suite, but looks unnaturally colored on the 1000 nit monitor, the colorist can add a correction in the DoVi metadata to change how the 4000-to-1000 tone mapping process is done for that particular shot.

Okay, I get that. Thanks for the reply.

[TbeRw01]

Originally Posted by RocShemp

This whole trim talk is flying over my head. Anyone kind enough to point me towards some recommended reading on the subject?

if you want to visualize how the L2 trims work, here I did a test file using the eagle scene from Spears and Munsil montage that removes each L2 metadata one by one.

https://drive.google.com/file/d/1qdH...usp=share_link

[rickardl]

Originally Posted by David M

Not Stacey, but Panasonic's approach to chroma upsampling has been pretty widely documented. It can cause minor colored ringing around sudden chroma transitions. I've never actually seen the effects of it on real live action content but I imagine there'll be something that makes it visible... and I think their approach to chroma upsampling still has benefits of its own despite that side effect.

I found your old comment (edited) from 2018 below, have you anything to add or amend now?

Originally Posted by David M

I like their take on chroma upsampling in theory, although especially on TV sized screens, the effects will be very hard to discern with most live action content on BD. It's potentially much more useful for upscaling DVD's 360x240 or 360x288 chroma channels than it is for BD. It's been noted that what they're doing to Cb and Cr can leave colored ringing, most visible on test charts but I've found lowering the Chroma sharpness control basically gets rid of it, but it'll almost never be visible in content anyway. I need to do more tests with that to see if that introduces any other collateral damage. It's mild either way.

One weird thing is that it appears to have CUE when playing 480i (and presumably 576i) content from BD (not DVD). All that really affects is some old SD bonus features that are encoded as SD on certain BDs. SD content is rare as hell on BD in the West so I'm not losing sleep over it. I've tested some 25fps content encoded on BD as 1080i/50 with 2:2 and it doesn't have CUE there, but will need to see if it affects pure 1080i/50 or 1080i/60 video - hopefully not.

If it does, the European firmware has a "Bandwidth Limitation" menu in the picture settings which lets you apply vertical and horizontal lowpass filters to the luma and chroma channels. This is applied after deinterlacing but before scaling, so doesn't mangle interlaced content. If seeing CUE with SD on BD bothers you, you can just turn up the vertical chroma filter and essentially hide the problem. As I said when I reviewed the European UB900, I'm not sure what the intended use of this menu is - this is the first use I've found for it. If I recall rightly, it isn't featured on the US version.

As usual I'm impressed that Panasonic are identifying potential system performance gaps and making actual improvements (HDR Optimizer essentially performing tone mapping in the player rather than the display). Looking forward to doing more testing with it. As is typical with their products, there is no behind-your-back guff like noise reduction going on (the last Samsung players I looked at applied noise reduction to the chroma channels without having any off switch).

I also really, really like the Edge Correction feature which can be clicked up a notch to remove ringing caused by the scaler.

Very happy with the UB820 so far. I'll do more testing later and pass any feedback on.

[Geoff D]

Originally Posted by TbeRw01

AFAIK, trims always been optional and plenty of movies don't have any or only have a 100-nits trim.

I just checked all the DV metadata for the Star Wars movies and:

Star.Wars.Episode.VIII.The.Last.Jedi.2017
100/600/1000nits trims (they did the trims static though)

Star.Wars.Episode.VII.The.Force.Awakens.2015
100/1000nits trims

Star.Wars.Episode.IX.The.Rise.of.Skywalker.2019
only a 100nits trim

Rogue.One.A.Star.Wars.Story.2016(disney plus)
only a 100nits trim

Solo.A.Star.Wars.Story.2018(disney plus)
only a 100nits trim

Star.Wars.Episode.I.The.Phantom.Menace.1999
no trims, not even 100nits

Star.Wars.Episode.II.Attack.of.the.Clones.2002
no trims, not even 100nits

Star.Wars.Episode.V.The.Empire.Strikes.Back.1980
no trims, not even 100nits

Star.Wars.Episode.VI.Return.of.the.Jedi.1983
no trims, not even 100nits

Star.Wars.Episode.IV.A.New.Hope.1977
no trims, not even 100nits


L1-L2 plot of The Last Jedi BD:

[Show spoiler]





**** I converted my rip to Profile 8 but that doesn't change anything related to the metadata itself.

You’re right, I’m misremembering, it was the dynamic metadata itself as Volpatto wanted to set a uniform transform for the whole film rather than have it be adjusted shot by shot. His method was initially rejected by Dobly as explained in Daro’s blog here: https://www.johndaro.com/blog/2022/1...ion?format=amp

Edit: and you can see that the metadata plot for Last Jedi is virtually static

[Geoff D]

Got my copy today, thanks Stacey and thanks to all involved for creating such a comprehensive set of test goodies.



I’ve already found the hidden Dolby Vision test pattern and as expected it’s got a great big red X owing to the ZD9’s low latency Dolby.

[Scarriere]

I don't understand much of what is discussed here, but definitely know that I want this disc for one of my tvs.
There was something with Sony that I was waiting for to be sorted out.
Is that done now? Can I order the disc for my X90K?

[TbeRw01]

Originally Posted by Geoff D

as expected it’s got a great big red X owing to the ZD9’s low latency Dolby.

The LLDV bug is not actually a big issue. It's the L2 trim Chroma_Weight that doesn't work but It's almost NEVER used in Movies and TV Shows by colorists.

The Spears and Munsil montage just happens to use it a lot.