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

[Stacey Spears]

I will take a look at the motion on the CX and see what type of difference I see with Real Cinema on vs. off.

My plan is to send in UHD 120 (119.88) from a MadVR Envy into a G2 in PC mode / 4:4:4 mode. Mathias has confirmed that we also get a natural 5:5 in that scenario. This allows me to bypass Dolby Vision and get good tone mapping for all content. Real Cinema is disabled when 120 Hz is the input frame rate. At least on the G1. However, Mathias did confirm it was still showing at a 5:5 rate. Vincent also confirmed this I believe.

[Stacey Spears]

Originally Posted by DisplayCalNoob

Wouldn't those frames be in a purely gamma based form?

Can you clarify what you mean by that? it is just repeating the frames an equal number of time. I don't think the EOTF matters. By that I mean it should repeat the frames in their native space.

Originally Posted by DisplayCalNoob

If I disable the feature on my display, there does appear to be an up tick in overall detail(static/motion). When I increase Reduce Judder to 5 while displaying the Stock Ticker on the UHD HDR Benchmark disc, the top scrolling pattern goes from having a ghosting affect to a stable smooth motion. The difference in image quality jumped significantly.

Any of the motion comp controls will make the ticker, motion, xXx and Sarah look better in the detail / smoothness department. Sarah does break every algorithm as the ropes move over the gravel in the cement though.

A note about the new Sarah clip. We used the 60p version for 24p. This gave us the speed we wanted. For the 60p version, we simply repeated every frame twice and then encoded as 60p. Again, gives us the speed we like. When we play the original 60 at 60, it moves way to fast. I am surprised she did not fly out of the hammock.

[Stacey Spears]

We have a new color temp pattern. I suspect everyone will think it is backwards by default because you see red at the top (+1500k over D65) and blue at the bottom (-1500k below D65).

The pattern is NOT intended to show you what +1500 and -1500 looks like relative to D65. The way the pattern works is you find the row closest to neutral white (D65) and then you look at the label. The label tells you how far away from D65 you are.

e.g. If the +1000 row looks closest to D65, then that means the current color temp is 1000 above D65 or ~7500.

Here is the pop-up help text for the new color temp pattern:
This pattern is used to choose a basic Color Temperature mode, for displays that have selectable color temperature.

There are multiple rows of gray steps. The row labeled D65 should look neutral white if the color temperature is correct, and the other rows will have a slight blue or orange tint.

If you have a D65 optical comparator you can use it to verify correct color temperature. You should see that the row labeled D65 matches the color of the comparator. If a different row is closest, the number next to that row tells you how far the temperature is off in degrees Kelvin.

If you don’t have a D65 comparator, you can visually look at the pattern and get a rough idea of whether the color temperature is correct. When D65 is correct, the top row will look notably pale orange and the bottom row will look notably pale blue.

If you have a control that allows you to select a specific temperature in degrees Kelvin, turn it up or down until the D65 row looks correct. If your display just has a selection of named color temperatures, go through them one by one, and select the one that makes the D65 row look correct.

So far everyone that has seen that pattern thinks it is backwards and Don even pointed out that people will think it is backwards before writing one line of code to generate the pattern.

The top is +1500 and will look warm in color (orange-ish) if D65 is correct. If you then turn up the color temp to its coolest setting, all rows leading up to that row will now be blue-ish and that row will look the most neutral white. It is interesting to watch the color temp shift up and down as you move color temp from warm-to-cool and vice versa.

We also created patterns specifically for using an optical comparator. I would never in a million years want to use them to calibrate grayscale. But they are there for those that can't / won't spend the money on a colorimeter and software to use it. You do need an optical comparator. We spent $5k on a prototype for one to bundle with the disc, but we were not happy with the results in the end.

[Stacey Spears]

Our goal is to finish disc 2 (SDR) around 4/15 and get back to the HDR disc. We have an open issue with DTS:X and it may just be a limitation of DTS:X since DTS:X Pro is a decoder side feature that uses an up mixer to create more than 7.1.4 channels. The issue is certain top channels actually produce audio in other speakers. DTS is looking at it now. If it can't be fixed we may either disable that option or just warn the user they need to disconnect the other speakers if trying to calibrate that specific channel.

On the plus side, the wides in a 9.1.n scenario seem to work correctly. This is not one of the normal DTS:X speaker layouts. I have mentioned before, but will again.

We let the user select the number of base layer speakers: 5.1, 7.1 or 9.1.
We let the user select the number of top layer speakers: 2, 4 or 6.

Then when you select levels or bass management you then select base or top and those are the only speakers you see. The option is Dolby and DTS. These are all built on Atmos and DTS:X.

For A/V Sync, we offer in True HD 7.1, Atmos, DTS:X and DTS-HD MA 7.1. Audio is in center channel only. We also offer in 23.976, 59.94 and HD, UHD resolutions. And of course in SDR, HDR10, HDR10+ and Dolby Vision. We added a second Sync-One2 pattern for calibration of the Sync-One2 and we made additional changes to the Sync-One2 based on feedback from them. In my quick test, DTS:X requires a different A/V sync offset than Atmos in SDR. Have not tested the other modes yet.

The panning tests are not based on the speaker layout selections. They just travel in 3D space.

We have two rattle tests. 500-200 Hz and 200-15 Hz. We only offer in the center channel since it is the channel most likely to be set to small.

[DisplayCalNoob]

Originally Posted by Stacey Spears

Can you clarify what you mean by that? it is just repeating the frames an equal number of time. I don't think the EOTF matters. By that I mean it should repeat the frames in their native space.



Any of the motion comp controls will make the ticker, motion, xXx and Sarah look better in the detail / smoothness department. Sarah does break every algorithm as the ropes move over the gravel in the cement though.

A note about the new Sarah clip. We used the 60p version for 24p. This gave us the speed we wanted. For the 60p version, we simply repeated every frame twice and then encoded as 60p. Again, gives us the speed we like. When we play the original 60 at 60, it moves way to fast. I am surprised she did not fly out of the hammock.

My apologies, its more about what exist in that frame. DV dynamic metadata, pixel brightness, gamut(color volume) mapping. Does that make up the frame that is being repeated?

I'm guessing that the idea of 3:2 and 5:5 pulldown came about before consumer HDR(metadata and tone mapping), so the frames may not reflect the content accurately. I may be way off, but looking at the content, I see a difference in color, detail and how small highlights appear. This is on a 2016 Vizio P Series.

[Geoff D]

Why would it not be repeating that exact same frame? DV metadata is there to serve the frames it was encoded for, it's not designed to link to pulldown frames on an ad hoc basis because there's no need to. The display is just duplicating the final frames it's been given by the DV decoder.

[Stacey Spears]

Originally Posted by DisplayCalNoob

My apologies, its more about what exist in that frame. DV dynamic metadata, pixel brightness, gamut(color volume) mapping. Does that make up the frame that is being repeated?

I'm guessing that the idea of 3:2 and 5:5 pulldown came about before consumer HDR(metadata and tone mapping), so the frames may not reflect the content accurately. I may be way off, but looking at the content, I see a difference in color, detail and how small highlights appear. This is on a 2016 Vizio P Series.

DV metadata is based on a source start and end frame for each cut/edit/shot. The 5:5 should not have any impact on tone mapping.

[DisplayCalNoob]

Originally Posted by Stacey Spears

DV metadata is based on a source start and end frame for each cut/edit/shot. The 5:5 should not have any impact on tone mapping.

If there were impact, that would be due to the display manufacturer implementation of motion processing? My display seperates reducing Judder and Motion Blur and then has Film Mode for 5:5.

I'm just trying to understand what I'm seeing, I remember the smallest changes in color, where a object is, faces. My memory is weird, I can remember faces of people I haven't seen in 20 years.

I'll move on and wait for the next release.

[Balling]

5:5 pulldown does not have any impact, but the OLED is not the same as CRT, it has insanely low latency (0.005 ms), and input lag too (5 ms on 120 fps game mode), so some movies may look unnatural. That is why you need to use at least some smoothers. Wow, are you seriously not aware of that? (The end of Spider-man: No way home Blu-ray where rips to other universes happens is such an example.) Now there are two 60/1.001 fps movies out there and second of them, gemini man was done correctly! So 60 fps motion is preserved and nowhere there is too much motion, so there is no "artifacts". Really, almost all youtube is 60 fps nowadays and hollywood does not know how to not introudce opera effect.

[kfbkfb]

I have a Magnavox MBP6700P UHD Blu-ray player which I use with an HDMI to NTSC downconverter to supply a picture to my 19" CRT TV.

Considering that the UHD Blu-ray format allows for a wider color gamut than NTSC, will the color adjustment tests from this test disc be useful for setting up an NTSC color picture?


Kirk Bayne

[rickardl]

I was looking at this contrast pattern http://spearsandmunsil.com/wp-conten...st_Correct.png
and noticed some black tick-marks. I don't recall them from my disc but anyway,
at 940, is that to indicate whiter-than-white above that? similar to 235 video level?
And thin ticks in the ramp between 993-994 and 995-996 to indicate whiter-than-white between them, right?
How thick/thin should the middle "stripe" of max white be in the ramp?

[Geoff D]

Originally Posted by rickardl

I was looking at this contrast pattern http://spearsandmunsil.com/wp-conten...st_Correct.png
and noticed some black tick-marks. I don't recall them from my disc but anyway,
at 940, is that to indicate whiter-than-white above that? similar to 235 video level?
And thin ticks in the ramp between 993-994 and 995-996 to indicate whiter-than-white between them, right?
How thick/thin should the middle "line" of max white be in the ramp?

Yes, 940 is the limit of a 'legal' 10-bit signal while full range is 1023.

I'm just guessing but the thin ticks in the ramp might be to indicate the ideal thickness of the white bar in the middle of the ramp, i.e. if the thickness exceeds the ticks then it's clipping <940 information.

[rickardl]

Originally Posted by Geoff D

I'm just guessing but the thin ticks in the ramp might be to indicate the ideal thickness of the white bar in the middle of the ramp, i.e. if the thickness exceeds the ticks then it's clipping <940 information.

Here is what the help text says for the HD version:
"The ramp in the center that goes from black to white and back should have a very thin white stripe in the center, but no other streaks or bands."

[Geoff D]

Originally Posted by rickardl

Here is what the help text says for the HD version:
"The ramp in the center that goes from black to white and back should have a very thin white stripe in the center, but no other streaks or bands."

It won't be as thin as it looks in the image you linked though as that's showing all the range, at least it is on my iPad. So the line won't be ultra thin by default, and as long as you're showing up to 940 then you're showing the legal video range.

[rickardl]

Originally Posted by Geoff D

It won't be as thin as it looks in the image you linked though as that's showing all the range, at least it is on my iPad. So the line won't be ultra thin by default, and as long as you're showing up to 940 then you're showing the legal video range.

But you don't want it this thick, right? (Contrast_Incorrect_Clipped)
http://spearsandmunsil.com/wp-conten...ct_Clipped.png

[Geoff D]

Originally Posted by rickardl

But you don't want it this thick, right? (Contrast_Incorrect_Clipped)
http://spearsandmunsil.com/wp-conten...ct_Clipped.png

I guess not! And yet up to 940 is still resolved in the other part of the pattern, maybe the rest of it is for not clipping whiter-than-white in the 10-bit space.

[rickardl]

Originally Posted by Geoff D

I guess not! And yet up to 940 is still resolved in the other part of the pattern, maybe the rest of it is for not clipping whiter-than-white in the 10-bit space.

I posted some comments about the HD version of the ramp here:
https://forum.blu-ray.com/showpost.p...&postcount=722

[INdetectableMAN]

Is there any news about the release date for the new album?

Is something known? Is the new album ready? Can it be reserved?

[JurassicBD]

Originally Posted by Stacey Spears

DV metadata is based on a source start and end frame for each cut/edit/shot. The 5:5 should not have any impact on tone mapping.

I'm also looking for any kind of rough eta on the new version. I noticed you raised your amazon price by 5 bucks, but there is no indication that the edition has changed, so I assume your replication costs have increased instead of it reflecting a switchover to the new edition, even though your site still lists the SRP at 39.99.

[Wendell R. Breland]

Anyone here have the capability of monitoring video levels and other video parameters via a HDMI 2.0, HDCP 2.2 source? If true what hardware/software are you using?

I am looking into the HDRScope by Drastic Technologies (here) and the Blackmagic Design Intensity Pro 4K (here). A HDRScope video below:

HDRScope AJA Kona 5 Live Demo - YouTube