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

[DJR662]

Originally Posted by HDMan72

I would imagine the PQ curve might be off in some way? We would need a colorimeter to be sure I suspect. I lowered my Pro contrast down to the default 90, the same amount of bars were lit without clipping.

What we need is Geoff coming to our houses with his gear.

[Geoff D]

Originally Posted by DJR662

I've actually been using Pro for both SDR and HDR viewing all this time. Initially I had contrast set to 84 for HDR. I did that by eye using the very beginning of the last chapter of Pacific Rim UHD (that shot of the ocean and the sky). Then I changed that to 86 later on when you said that's what tracks the PQ curve the best according to your calibration. From what I remember back then, contrast set to 86 for SDR didn't really clip that much, so I just took the small tradeoff for SDR (if any at all for real world content?) and used Pro for both SDR and HDR and never looked back.

I now actually prefer trying to use as few different picture modes as possible. If that means a small tradeoff in any of them, then I can live with that. I used to have different picture modes for day and nighttime viewing, for cable TV, for 3D and then for HDR. In the end I got tired of having to switch each time, never mind forgetting to switch back to the appropiate picture mode when done watching in another.

So right now I use Pro for 2D SDR/HDR, Custom for 3D and Home for the occassional 4000 nit discs (though I must confess I hardly use the latter).

Btw when I looked at the montage in 4000 nits with the Panasonic 820, toggling between Pro with the Optimizer on and Home with the Optimizer off, I was reminded again why I don't use the Optimizer with the Z9D.

TBH I always double check my settings before I watch stuff anyway, it doesn't take but 10 seconds to switch to the optimum viewing mode.

So if by setting contrast to 93 in Pro gets you the most bars without clipping, I wonder what this means for the way it follows the PQ curve with that setting then?

He said on a 1000-nit pattern he gets all the bars without clipping on 93 in Pro, which is fine but (if we're talking about a ZD9) it'll make the HDR a fair bit brighter than it'll otherwise be at all points in the luminance curve.

Remember, the ZD9 doesn't apply a different tone map according to what HDR10 content you're playing, it's on the contrast control and where you set it as to where it'll clip & how much luminance is being applied across the curve. Where it clips you can only gauge by eye, how much luminance you get you can only gauge with a meter and software.

Originally Posted by HDMan72

I would imagine the PQ curve might be off in some way? We would need a colorimeter to be sure I suspect. I lowered my Pro contrast down to the default 90, the same amount of bars were lit without clipping.

If they're all lit already in 93 then yes, lowering contrast won't clip them further, you're doing the opposite of clipping them.

[Wendell R. Breland]

Originally Posted by Penton-Man

As an aside, AJA’s FS-HDR was used in monitoring the production of Top Gun: Maverick –

I may have missed it but they did not mention 3G-SDI for the FS-HDR.

In looking at some pictures of a few post houses it appears Tektronix does not have the presence they once had in waveform monitoring. They do have 3G-SDI as an option on all their series of WFM's.

EDIT: It appears AJA's have 3G-SDI built in, with 12G-SDI as an option. The Tek's do not offer 12G-SDI.

[HDMan72]

Originally Posted by DJR662

What we need is Geoff coming to our houses with his gear.

That would be fantastic!

Are you guys having a hard time logging on here. Last few days, I can’t get in, page loads forever than crashes

[HDMan72]

Originally Posted by Geoff D

TBH I always double check my settings before I watch stuff anyway, it doesn't take but 10 seconds to switch to the optimum viewing mode.


He said on a 1000-nit pattern he gets all the bars without clipping on 93 in Pro, which is fine but (if we're talking about a ZD9) it'll make the HDR a fair bit brighter than it'll otherwise be at all points in the luminance curve.

Remember, the ZD9 doesn't apply a different tone map according to what HDR10 content you're playing, it's on the contrast control and where you set it as to where it'll clip & how much luminance is being applied across the curve. Where it clips you can only gauge by eye, how much luminance you get you can only gauge with a meter and software.


If they're all lit already in 93 then yes, lowering contrast won't clip them further, you're doing the opposite of clipping them.

I finally realized that myself, 1,000 nit pattern at 93. 4,000 nits no way. Is that why 85 is common in Pro (from what Ive read over at AVS Forum), to compensate 1,000-4,000 nits?

[Geoff D]

Pro at 85/86 I have mine set to the latter) is where the TV's luminance curve most closely follows where it should be when watching HDR10, that's the foremost consideration for setting it in that manner, so the amount of highlights it resolves is a by-product from where it's set, if that makes sense?

But it just so happens that on a ZD9 it will resolve ~1600 nits at this setting (in fact, I've tested mine again and it's reading 1800 nits in the S&M patterns, not 1600) so it's a great balance because it gives you more than enough highlight information to be able to enjoy the full 1000-nit representation of several studios' discs (plus anything else in the 1000-2000 range) and you get very close to where the PQ curve is intended to be - whereas if you pump up Pro to 93 to just resolve up to 1000 nits the curve will be too bright and you'll clip >1000 nit highlights.

Pro on 86 isn't perfect of course, there are several movies out there which have some very extreme >2000 nit highlight detail (or have moderate highlight detail but pumped up into silly brightness levels) and it will heavily clip that stuff, so you then have to make the decision to use something like the Panasonic HDR Optimiser or lower the contrast on the TV to pull down more highlights at the expense of following the intended brightness curve.

Both are a compromise; the Optimiser will hold APL up to 50% of the signal (so up to 100 nits, basically) then rejigs the curve into something more akin to SDR and I don't like the way it affects the dynamism between light and dark, it really flattens out the impact of the highlights even though they're resolving more detail. If you pull down the contrast (as I said, I have my '4000 nit mode' on Cinema Home set to 82) then you will lose brightness across the board, not just from 50% up but the whole range, yet because it maintains the shape of the curve it still retains plenty of HDR 'impact', particularly in a darkened room, so even though it's darker than the Optimiser it's more dynamic IMO.

[HDMan72]

Understand now, thanks! My Cinema Home mode I did with 4,000 nits and contrast at 82 and yup, it looks great in a darkened room from what I see. I thought about getting a Panny when I visited Robert Zohn’s shop and he said not needed with my Z9D, said my Oppo 203 is perfect match. What colorimeter do you use?

[Geoff D]

ColorMunki Display (same basic internals as i1Display 3) which I first profiled to the TV with a ColorMunki Photo spectro (now renamed as the 'i1 Studio').

[DJR662]

Originally Posted by HDMan72

That would be fantastic!

Are you guys having a hard time logging on here. Last few days, I can’t get in, page loads forever than crashes

Not just last few days, this is happening from time to time the past few months actually.

[Geoff D]

There were several problems with bots and whatnot a few months back, but he's right that very recently it's gone to shit again.

[ROSS.T.G.]

Originally Posted by Geoff D

ColorMunki Display (same basic internals as i1Display 3) which I first profiled to the TV with a ColorMunki Photo spectro (now renamed as the 'i1 Studio').

For arguments sake, how much error can I expect if I only use the i1 Display Pro and skip the Munki?

[Geoff D]

It's tough to say as one meter might be closer than the next one that comes off the line, the next one further away and so on. But depending on what profiles your calibration software has you might be able to get it closer, most have an OLED profile of some description.

[koberulz]

How do you do the profiling step? I did my initial calibration with an i1Display Pro, but I've picked up an i1Studio in the meantime.

[ROSS.T.G.]

Originally Posted by Geoff D

It's tough to say as one meter might be closer than the next one that comes off the line, the next one further away and so on. But depending on what profiles your calibration software has you might be able to get it closer, most have an OLED profile of some description.

Thanks, that piece of mind is important.

[Geoff D]

Originally Posted by koberulz

How do you do the profiling step? I did my initial calibration with an i1Display Pro, but I've picked up an i1Studio in the meantime.

Depends on your software, some have a profiling setup all laid out I think? For something like HCFR it's a little bit more work (a tutorial about which you can find in the HCFR thread at AVS).

[edit]

But basically you simply take a reading of red, green, blue and white panels with the spectro, then same again with the colourimeter, and the difference between them is used to calculate an offset which will correct the colourimeter for the most accurate readings on your specific display.

[Geoff D]

Originally Posted by ROSS.T.G.

Thanks, that piece of mind is important.

I've given out so many pieces I'm surprised I've got any left.

[koberulz]

Originally Posted by Geoff D

Depends on your software, some have a profiling setup all laid out I think? For something like HCFR it's a little bit more work (a tutorial about which you can find in the HCFR thread at AVS).

[edit]

But basically you simply take a reading of red, green, blue and white panels with the spectro, then same again with the colourimeter, and the difference between them is used to calculate an offset which will correct the colourimeter for the most accurate readings on your specific display.

If the spectro is so much more accurate why use the colorimeter at all?

[Stacey Spears]

Originally Posted by DJR662

Not just last few days, this is happening from time to time the past few months actually.

Last weekend the site was almost non-responsive for me.

[Stacey Spears]

Originally Posted by koberulz

If the spectro is so much more accurate why use the colorimeter at all?

Speed. A specto can take minutes to read black while a Klein can do it instantly.

[Stacey Spears]

As an experiment this weekend, I copied the EDID out of the LG and used it in front of the Z9D and DV did not clip at factory default. So I don't think the EDID in the Z9D is optimal. I would not use the LG EDID either as it is requesting a lower nit level for Dolby Vision, but Sony is requesting too high a nit level.