AVS Forum banner

701 - 720 of 720 Posts

·
Registered
Joined
·
5,545 Posts
Your model is the entry level QLED set. Samsung changes the feature set and chassis from model to model, and this very well may mean that the proper entry point in memory or the same CMS control scheme available in top end sets is not in yours. CalMAN can only work for autocalibration with the internal settings Samsung "exposes" to software control.
 

·
Registered
Joined
·
4 Posts
Your model is the entry level QLED set. Samsung changes the feature set and chassis from model to model, and this very well may mean that the proper entry point in memory or the same CMS control scheme available in top end sets is not in yours. CalMAN can only work for autocalibration with the internal settings Samsung "exposes" to software control.
Thank you! it looks like that is the case and i could understand that if I didn't ask Calman customer service that my Q60 is fully compatible with samsung autocal before i bought the software. They sad my tv is fully compatible with the autocal not just partially (CMS autocal not supported with this model) so bought the software just to find out its not fully compatible with my entry level tv. If i knew that I'am not sure i would bought the software.
 

·
Registered
Joined
·
5,545 Posts
Thank you! it looks like that is the case and i could understand that if I didn't ask Calman customer service that my Q60 is fully compatible with samsung autocal before i bought the software. They sad my tv is fully compatible with the autocal not just partially (CMS autocal not supported with this model) so bought the software just to find out its not fully compatible with my entry level tv. If i knew that I'am not sure i would bought the software.
You can still do manual calibration for CMS. Just be advised that many manufacturers' CMS implementations are very poor and that even small changes can introduce visible image artifacts. In my experience, grayscale and gamma are far more important to overall image quality than CMS.
 

·
Registered
Joined
·
4 Posts
Tv calibration is new to me, when i first tried to do manual calibration for CMS after grayscale my impression was that the end result worse than without CMS. Autocal probably would do a better job but I will try again manually and if its not better i leave on default, but without CMS calibration i feel some color are oversaturated.
 

·
Registered
Joined
·
5,545 Posts
Saturation and luminance are crosslinked on emissive displays and hard to separate. It may well be that what you consider oversaturated is just too bright (luminance targets for primaries are different between SDR and HDR). Running the TV at "Native" gamut in SDR, however, will indeed result in excess luminance and oversaturation.
 

·
Registered
Joined
·
651 Posts
Tv calibration is new to me, when i first tried to do manual calibration for CMS after grayscale my impression was that the end result worse than without CMS. Autocal probably would do a better job but I will try again manually and if its not better i leave on default, but without CMS calibration i feel some color are oversaturated.
Hi,

As Rolls-Royce alluded to, sometimes you can make things worse via the CMS. I own a few Samsungs and their CMS implementation is pretty good but even with that said, you will probably be fine with Movie mode and the Color Space setting set to Auto for SDR content. Rtings' precalibration measures shows that only fully saturated red and blue had dE2000 errors over 3 and after calibration things only modestly improved from a dE perspective. I would agree that getting grayscale and gamma dialed in are most important (and should be done before tinkering with the CMS).


cheers,


--tom
 

·
Registered
Joined
·
1,605 Posts
I made an interesting find about HDR on a Q90R: using the CMS controls for the Red primary makes flesh tones worse. Might be worth checking to see if that's true for other model/year QLEDs.

I've been doing fairly extensive HDR testing on my Q90R (using C6 HDR2000 profiled with i1Pro3, VFPro), testing various Contrast levels using 20pt plus CMS, just 20pt, and just CMS (always 2pt of course). I've done the 20pt calibrations manually using adjustment points I've mapped for each Contrast level. For CMS, I've used Autocal as well as manual calibration using DCI-P3 points. For flesh tone analysis I've been using 19 colors: Dark Skin and Light Skin from Color Checker Classic plus the SG Fleshtones subset of Color Checker SG.

As a concrete comparison, using Contrast 43, manual 20pt, and Autocal CMS, here are flesh tone metrics for the full Autocal CMS versus resetting Red to defaults after the Autocal:

DEITP
Avg
DEITP
Min
DEITP
Max
DEITP
Std Dev
de2000
Avg
de2000
Min
de2000
Max
de2000
Std Dev
Autocal CMS9.990.7218.993.591.931.063.430.69
Reset Red6.020.609.272.081.170.262.150.45

(I pay more attention to the DEITP errors.) For each color, the errors are larger with the full Autocal CMS.

For broader impact evaluation, I've used two patch sets: the combined Color Checkers Classic+SG+Video, and a 216 color HSB-derived set. The HSB set focuses on the center of the CIE chart (see attached), and is generated using 20 degree steps for H, 10%/15%/20%/25% for S, and 25%/40%/55% for B. For the same settings as above, not every color got better with Red reset, but the average errors and standard deviations went down for both patch sets, so I consider leaving Red at default to be a net win.

HSB.png
 

·
Registered
Joined
·
2,879 Posts

·
Registered
Joined
·
1,605 Posts
Autocal uses the following SDR grayscale percentage adjustment points for a Q90R:
7, 10.5, 15.5, 19.6, 24.7, 29.7, 35, 39.7, 44.7, 49.8, 55, 60.3, 65, 70, 75.3, 81, 86, 90.4, 97.7, 100
But when I manually map them out on my 75", I get different results:
7, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55.3, 60.3, 65, 70, 75.3, 80, 85, 89.5, 95, 98.6
 

·
Registered
Joined
·
2,879 Posts
The Q9FN has the same Autocal grayscale, but the TV menu shows 5, 10, 15, 20 ... you get the drift.

Let's face it, Samsung didn't put much of an effort on making these TVs calibratable :rolleyes:.
 

·
Registered
Joined
·
1,605 Posts
The TV menu state is well known; the Autocal points not lining up is not, AFAIK. PD sets the Autocal points, not Samsung.
 

·
Registered
Joined
·
1,605 Posts
Beware if you've been using DDC over IP for CMS adjustments. Here's what happens on my Q90R; I don't know what other models are affected. After initially connecting to the TV, or after doing a Reload CMS, when I look at the DDC Gamut Adjustments pane, the B values are all duplicates of the G values, not the actual B values stored in the TV. When using DDC to manually update an R value, the R value does get updated, but in addition the B value in the TV gets replaced by the current G value in the TV, and this is not reflected in the DDC pane. When using DDC to manually update a G value, the G value does get updated, but in addition the B value in the TV gets replaced by the previous G value in the TV, and this is not reflected in the DDC pane. It appears to me that these same things happen when Autocal is running. As a result, Autocal CMS measurements and hence final result are pretty completely borked.

(My previous post on HDR flesh tones is not affected, as I was using a serial connection, having become suspicious of the IP connection.)
 

·
Registered
Joined
·
1,605 Posts
Beware if you've been using DDC over a serial connection, at least on 2019 sets. Here's what happens on my Q90R.

Make a DDC serial connection. TV mode can be either SDR or HDR.

Bring up the DDC panel and go to the RGB Adjust pane, and change one value (R, G, or B) in the first row (what the TV labels as the 5% Interval) to be non-zero.

Now go to the Gamut Adjustments pane and change any one value (R, G, or B, any primary or secondary).

Now do a DDC Disconnect.

Now use the TV remote to look at 20 Point Settings, and discover that for the 5% Interval, all three values are now zero.

I think this happens routinely when using Autocal, and I think this bug has existed ever since CalMAN first supported 2019 sets. I've seen an Interval other than 5% get zeroed when manually using DDC, but I haven't been able to come up with a way to reproduce it.
 
701 - 720 of 720 Posts
Top