[thoth]

yes my auto cal brings up the eotf. How does the work? It seems that my curve flat lines at the top of curve?

That's how you want it to look, it hard clips at the peak luminance supported by your display. This page, selecting the HDR Contrast setting, is important for HDR, and you have to determine the best setting yourself. I'd start with Contrast at max 50, and measure the whole curve. If it's like the Q90R, the measured values (white line) will be a fair bit higher than the target values (yellow line) in the right half of the curve. You want to incrementally lower Contrast and re-measure the whole curve as you go. It's a good idea to use the History tabs for this; for each new Contrast value, click the + sign next to the current History tab to get a new one, so that you can retain measurements for each Contrast value you try. (You can right-click on a History tab to rename it, so you can keep track of which Contrast value was used for it.) Keep lowering Contrast until you start seeing measured values (white line) that are noticeably less than the target values (yellow line) somewhere on the curve, probably in the 5-15% range. Probably you want to measure from 50 down to 40 or so. Then the hard part is to pick the Contrast value that has the best fit of measured curve to target curve. If there are a few Contrast values that all seem they might be good, then you might want to complete separate Autocal runs for each one of them, to see which one ends up with a better overall calibration at the end. Once you've settled on a good Constrast value, you probably don't have to go through this whole thing again, you can just use that value for future calibrations.

 

[FLPnotc]

I keep coming up with questions. Here’s another one.

Is this a proper chain to my pattern generator to my hdmi in tv this my setup - raspberry pi with tpg hdmi out to hdmi in av/receiver then hdmi out av receiver to hdmi in Tv

or does my raspberry pi w tpg hdmi out have to to directly into hdmi in tv?

would this interfere any kind of way the calibration that is being done?

 

[thoth]

I keep coming up with questions. Here’s another one.

Is this a proper chain to my pattern generator to my hdmi in tv this my setup - raspberry pi with tpg hdmi out to hdmi in av/receiver then hdmi out av receiver to hdmi in Tv

or does my raspberry pi w tpg hdmi out have to to directly into hdmi in tv?

would this interfere any kind of way the calibration that is being done?

It's better to have the RPi go directly to the TV, as it avoids any possibility that the AVR alters the patterns.

 

[FLPnotc]

Dang it. One more question. For HDR You gave recommendations to change settings to "5% steps rather than 10% in the Pre-Calibration Measurements and Post-Calibration Verification pages" which helped a lot.

Can this be done for both SDR and HDR? or was that only recommended for HDR. If not would there be recommended for SDR setting for pre cal and post cal verification?

 

[thoth]

Dang it. One more question. For HDR You gave recommendations to change settings to "5% steps rather than 10% in the Pre-Calibration Measurements and Post-Calibration Verification pages" which helped a lot.

Can this be done for both SDR and HDR? or was that only recommended for HDR. If not would there be recommended for SDR setting for pre cal and post cal verification?

Yes, it can be done for both SDR and HDR. Just to be clear, that doesn't alter what Autocal does, it just gives you a more accurate picture of the pre and post results.

Two more things you might want to change. One is to go into the CalMAN settings, in Application Measurement Options, under AutoCal Targets, and change DeltaE Formula to de_ITP if it isn't that already. Second, whenever you initiate an AutoCal operation using an AutoCal button, if a dialog box pops up, and it has a DeltaE Target option, probably with a default value of 0.5, try changing that to 0.1 before you click OK. These may make Autocal runs take a little longer, but they might also improve accuracy.

 

[FLPnotc]

what a rabbit hole i just gone through. I am pleased with the outcome!

One more thing as usual. Where is the detlaE formula thing. I am planning on redoing all of the cailbration tomorrow with this setting on.
I looked every where. I can seem to locate it.

 

[BigGuybrush]

Is it taking too long because you're doing manual adjustments while an HDR pattern is displaying, or an individual measurement is taking too long, or something else?

The problems already appear in the "Autocal" function, 2P WB, while he regulates the 70%, the nits already fall.

Have you tried the Full Field Pattern Insertion option, under Application Measurement Options?
You may have to experiment with the frequency/duration/level settings to get the desired results.

Yes I have, but maybe not long enough. Is there perhaps a guideline for the settings?

I don't own that set, but I'm surprised CalMAN is disabling Smart LED for HDR calibration, I would have expected it should be on High.

This is exactly why I don't understand how it should work with the "Smart LED" switched off. By the way, it is switched off with SDR and HDR.
I already asked Calman, here is the answer from support:
We disable it as we found that having it turned on during calibration caused some odd side effects to greyscale test patterns. In other words, with Smart LED being a complication process, it doesn't necessarily have the same affect on test patterns as it would with normal viewing content, so it can't really be calibrated or equated for.

 

[thoth]

Where is the detlaE formula thing.

Click on Open Workflow and Application Settings Panel in the upper right, then click on the Application Measurement Options tab, then look in the third major grouping, AutoCal Targets.

 

[thoth]

The problems already appear in the "Autocal" function, 2P WB, while he regulates the 70%, the nits already fall.

Yes I have, but maybe not long enough. Is there perhaps a guideline for the settings?

This is exactly why I don't understand how it should work with the "Smart LED" switched off. By the way, it is switched off with SDR and HDR.

Hmm, I see in Samsung 2016 SUHD TVs it says only SDR calibration is supported, don't know if that's still the case. Turning local dimming off (or as low as possible) for SDR calibration is normal advice. I assume you're using a 10% Window for HDR. If you want to try pattern insertion further, I don't know of any guidance to point at, I'd start with an extreme, like Frequency=0, Duration=10, Level=0%, to see if consistent stable readings are possible.

 

[Rolls-Royce]

Hmm, I see in Samsung 2016 SUHD TVs it says only SDR calibration is supported, don't know if that's still the case. Turning local dimming off (or as low as possible) for SDR calibration is normal advice. I assume you're using a 10% Window for HDR. If you want to try pattern insertion further, I don't know of any guidance to point at, I'd start with an extreme, like Frequency=0, Duration=10, Level=0%, to see if consistent stable readings are possible.

True. On our 2016 KS 8000, only manual HDR calibration is supported. Autocal works with SDR only.

 

[FLPnotc]

Click on Open Workflow and Application Settings Panel in the upper right, then click on the Application Measurement Options tab, then look in the third major grouping, AutoCal Targets.

Would I set both to delta and cube delta to - DeltaE Formula to de_ITP ?

Another thing reading up on the last post.
Should Local dimming zones be set to high in HDR calibrations?
and should Local dimming zones be set to off in SDR calibration?

 

[thoth]

Would I set both to delta and cube delta to - DeltaE Formula to de_ITP ?

Another thing reading up on the last post.
Should Local dimming zones be set to high in HDR calibrations?
and should Local dimming zones be set to off in SDR calibration?

The Cube setting won't apply for you, but sure, set it to de_ITP as well. For your Q90T, what does the Full DDC Reset change Local Dimming to for SDR and HDR?

 

[FLPnotc]

The Cube setting won't apply for you, but sure, set it to de_ITP as well. For your Q90T, what does the Full DDC Reset change Local Dimming to for SDR and HDR?

When doing Full DDC reset Local dimming for HDR is on high. Local dimming for SDR is on Standard.
Should I leave it at those levels when calibrating?

 

[thoth]

When doing Full DDC reset Local dimming for HDR is on high. Local dimming for SDR is on Standard.
Should I leave it at those levels when calibrating?

You should definitely leave it on High for HDR. For SDR, on my Q90R, I don't measure any difference (using calibration patterns) between Low and Standard. If you want to check your Q90T, use the Manual Calibration workflow, configure for SDR, set Local Dimming to Low, go to the Pre-Calibration Measurements page, change the option to 5% steps, and do a Read Series. Then skip forward to the Post-Calibration Verification page, change the option to 5% steps, change Local Dimming to Standard, and do a Read Series. Then go to the Pre/Post Comparison page and see if you notice any real difference. If you don't then it's fine to leave it at Standard. If you want to be conservative, change it to Low for calibration, and then back to Standard for normal viewing.

 

[FLPnotc]

I applied Delta_ITP. It shows the post cal with a bit higher error avg and max error.

Does that mean the calibration is worst or will Detl_ITP always give you a higher number of errors but overall better calibrated tv?

 

[thoth]

I applied Delta_ITP. It shows the post cal with a bit higher error avg and max error.

Does that mean the calibration is worst or will Detl_ITP always give you a higher number of errors but overall better calibrated tv?

Samsung QLEDs aren't super stable, especially for HDR, so minor variation from calibration to calibration is inevitable. So it could just be that. But it could also be because of the DeltaE Formula change.

There are multiple formulas for determining how close two measurements are to each other, and they have fairly different characteristics. What can add confusion is that the formula CalMAN uses for Autocal doesn't necessarily match the formula used for the various charts you see. If you look at one of your saved sessions, look at any chart that is labeled DeltaE 2000; that's the default formula that CalMAN uses to present results to you in charts. If you right-click on one of those charts and select Properties, in the property sheet that pops up you can see Chart Type, where you can select a different formula to display instead. If you select DEITP, you'll now see values based on the formula that's being used by Autocal when you set the DeltaE Formula setting to de_ITP (don't ask me why CalMAN can't keep naming consistent). Typical DEITP values can be much larger than typical DeltaE 2000 values. One other thing if you change the chart type, though: in places where CalMAN shows you Avg Error and Max Error values, those are hardwired to DeltaE 2000, they unfortunately don't update to match chart type you've actually selected; it's annoying and confusing.

Now to get back to your question: I don't understand the math behind the formulas enough to know if it's possible to have DEITP values go down but DeltaE 2000 values go up, or vice versa. Perhaps someone else can say.

Using de_ITP was just a suggestion; some people believe it gives a more accurate assessment, especially for HDR, and some believe it's flawed. If you don't want to use it, just change the formula back to what it was.

 

[FLPnotc]

@thoth

Is Delta_itp only for HDR?


Can you take a looks at setting notes I have to see if everything seems to line up with what you recommend


SDR - (Delta to use unconfirmed)
For SDR reset movie mode and set local dimming @ low and set black level low
do the pre/post 5% 0-100 21 point instead of 10% 11
do 2 Pt 30/100% when time to choose .1
Skip contrast / brightness page

HDR - Use Delta_itp
For HDR reset Movie mode and set local dimming @ high set black level low
do the pre/post 5% 0-100 21 point instead of 10% 11
2 pt 30%/70% @.1
Skip Contrast / brightness page
Match EOTF best a possible


Let me know if other settings I can change to make it more accurate.

 

[thoth]

@thoth

Is Delta_itp only for HDR?


Can you take a looks at setting notes I have to see if everything seems to line up with what you recommend


SDR - (Delta to use unconfirmed)
For SDR reset movie mode and set local dimming @ low and set black level low
do the pre/post 5% 0-100 21 point instead of 10% 11
do 2 Pt 30/100% when time to choose .1
Skip contrast / brightness page

HDR - Use Delta_itp
For HDR reset Movie mode and set local dimming @ high set black level low
do the pre/post 5% 0-100 21 point instead of 10% 11
2 pt 30%/70% @.1
Skip Contrast / brightness page
Match EOTF best a possible


Let me know if other settings I can change to make it more accurate.

DEITP can be used for both SDR and HDR. It has been stated that CalMAN will eventually switch over to using DEITP as the default everywhere, but who knows when that will actually happen.

For changing the DeltaE Target from .5 to .1, you can do that everywhere it shows up: 2 Pt, Multipoint, and Colorspace.

By the way, if you want to get some sense of how stable your display is, try using the Manual Calibration workflow on several different days, just taking both SDR and HDR measurements, don't change any settings. Then compare the results.

 

[FLPnotc]

That is great insight. I will put that on my notes.

Here is something maybe you can answer also.
I noticed when changing my 4k firestick and Bluray player to YCbCr instead of RGB It seems like I get a better picture. Along with changing fires stick to 10bit instead of 8bit

Is it just placebo or is there something that is really happening.
Is samsung q90t better suited for YCbCr and 10bit settings?

Sorry I know its not a calman samsung question. But it might help Samsung users after calibrating to get better hardware settings from sources the play from. I don't know why but it just looks better smoother darker cleaner vibrant. All the above.

 

[thoth]

I noticed when changing my 4k firestick and Bluray player to YCbCr instead of RGB It seems like I get a better picture. Along with changing fires stick to 10bit instead of 8bit

There are three different possible problems, one or more of which could be contributing to that. HDR10 video content is 10-bit, so if you configure the source device to output 8-bit, you're forcing it to downscale the color, which is definitely bad for picture quality. Pretty much all SDR and HDR video content is also YCbCr, so if you configure the source device to output RGB, you're forcing a colorspace conversion that can introduce errors. Finally, if you configure the source device to output RGB, there could be a levels mismatch somewhere in the chain, at your AVR or the Q90T, confusing RGB 0-255 vs 16-235, which causes all colors to be misinterpreted. Ideally your AVR should be configured for pass-through video output, so that it doesn't introduce any conversion errors. One other thing, assuming the Q90T is like the Q90R, if you want to watch HDR10+ content, the source device needs to output 10-bit, not 12-bit.