LCD desaturates primaries at low luminance levels?

Evil_Husky said:

I've looked at that lightspace article and the cie graph representation of the issue and it pretty much looks as bad as I thought. 😱

I wish I knew this year ago when I was shopping for an oled tv, I would get an lcd instead. With all the constant complaints about mostly irrelevant stuff about these tvs such as low level grey uniformity (vertical lines and blobs, who even sees these in real life content), motion (which looks completely normal to me), burn in (that doesn't happen with normal use) etc, you would think that people on this website would talk more about things such as badly oversaturated darker colors that are clearly visible in a side by side comparison, bad white point shift at an angle and requiring custom white point and perceptual matching to a reference display due to wide gamut and metameric errors.

So you need ColourSpace LGN license for lut editing?

Click to expand...

Hi,

Post-production houses have many different monitors in the same room, so all have to display the same colors as close as possible.

If the main reference monitor has desaturated primaries, then others without that problem will display the content differently.

In that case, you can use ColourSpace and apply a filter to desaturate the secondary RGB OLED or WRGB OLED available to the same room.

Usually, Sony X300 is commonly used as a reference, and it's not suffering from that issue with de-saturation in lower luminance levels.

EIZO CG319 (~7K$), which is LCD, doesn't have a problem also, as you can see below, it can retain saturation to lower luminance levels; only at the lowest (8% to the pictures below), it will have a de-saturation.

1000p SDR REC.709 Verification:






EIZO CG3146 (31K$), has no problem also:






LCD in the consumer market are not good displays, as they are suffering from much more issues compared to WOLEDs.

So WOLED's, especially LG, which has the richest calibration capabilities for calibration in the global TV market, is the best option as a consumer TV, compared to other TVs available to the market.

Panasonic WOLED will require an external LUT box to able to reach the performance of LG in SDR, and 4K LUT Box is very expensive (see the price of Lumagen PRO, more expensive from the TV).

Sony WOLED can't reach the performance of LG/Panasonic, and if you will buy an external LUT Box because you can't disable the Sony processing to the signal, it doesn't provide a native gamut as an option, any colorspace option is including Sony processing enabled.

The ideal display is the one that has pixel-per-pixel control (+1 to OLED), no ABL/APL (+1 to OLED up to 150 nits), no local dimming (+1 to OLED), no significant shifting if you don't watch dead-centered (+1 to OLED).

Consumer LCD has all these problems.

Professional LCD doesn't suffer from local dimming.

It's not available a perfect TV/monitor; even Sony X300 has many problems... any display tech/models have positive/negative stuff.

It's not possible to buy a perfect monitor; even if you search for a 20-30K$ 31-inch post-production monitor, you will find issues also.

LG WRGB OLEDs in the consumer world are not perfect also.

Still, for the price and for all these stuff a consumer can get, it's insane that they are available all these features at this price range in the consumer market.

Dolby Monitor (50K$) had a more significant white point problem than the issue you see with WRGB OLED.

The new RGB JOLED has a more significant issue than WRGB OLED with metameric failure.

As you can see below, LG WRGB OLED with ColourSpace and 3D LUT for SDR, you can get better measurements compared to a very expensive professional monitor (which is only 31 inches max, so you can't use it in a living room)

LG WRGB OLED SDR REC.709 1000p verification:



ColourSpace user with LG OLED have posted a lot of times 1000p verification of SDR REC.709:

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17

The de-saturation LUT manipulation filter is not required for consumers, its for post-production.

Consumer need ColourSpace HTL (for X-Rite) meters or HTP (for hign-end meters). HTP can manipulate the LUT.

Using the unique 3D CIE graphs of ColourSpace, this issue can be easily displayed.

Regular CIE Charts are 2D, the presentation of Luminance to able to see luminance errors is missing.

The following graphs show the de-saturation of an LCD , compared to the lack of de-saturation in a (W)OLED looking at Red Primary only.

The top dot is 100% Red luminance, the bottom dot is 5% Red luminance.

The gray tangent line shows the distance of measured vs. target location of each measurement, as you can see in LCD, they are going towards to white.

Red @ LCD:



Red @ WOLED:



The color of the dot represent the dE range.

Green dots have dE2000 below 1, Yellow (1-2.3), and Red (over 2.3)

The RED dot (larger dot) at the top, its the location of the primary color of the selected colorspace.

All these look better when you move/rotate/zoom in-out this chart.

Why do you want to fix that?

Do you watch at the same time the same movie with LCD and OLED? Are you a colorist?

The fix is required when you have multiple monitors, and one of them has a de-saturation problem.

For example, if you have an older reference LCD (cost 5-10K$) as a mastering monitor and you want to match the picture to the client view LG OLED, available in the same room.

It's the reason I posted examples of measurements from 2 professional monitors, which are not suffering from that LCD de-saturation issue.

You have to measure your LCD initially to find out what kind of de-saturation it has if you still need to apply a LUT manipulation.

But all these are required when both monitors have a properly loaded 3D LUT and have a mismatch visually.

So it will require loading a 3D LUT to both monitors.

Then you have to use perceptual matching.

The black level also has to match also.

ColourSpace has a function to limit via a 3D LUT the black/white level to specific nits to able to match black level or white level between 2 monitors.

After performing all the above, you have to verify to see what kind of de-saturation problem you have.

Are you going to perform all these and have the capability to load 3D LUT to both TVs?

CalMAN has no LUT viewer, no LUT manipulation tools, no way to find out which is your correct LUT file as it's creating multiple files during autocal.

If you don't have RGB Separation of 3D charts, it's not so easy to detect if the filter you will apply is matching.

All these need ColourSpace HTL for LUT generation, and you get and install to a different system (as its not possible to have two copies of ColourSpace on the same PC) the ColourSpace LGN, and use LGN for LUT manipulation.

(You can have ColourSpace ZRO and ColourSpace on the same PC as they are different installations).

You will need a PGenerator also to able to have a reference pattern generator for LUT-based calibration.

Evil_Husky said:

Because I'm quite sure all of the content I watch on my tv was made on lcd displays that desaturate gamut at low luminance. I thought this was the point of calibration (to minimize the difference between displays in terms of color reproduction)? But looking at rec2020, I guess this is not the case anymore, since there's zero displays available to consumers that can even come close, yet somehow it's still the standard.

Also, would be great if we could see the volumetric gamut measurements in all those professional tv reviews instead of just 2d gamut triangle.

Maybe I just don't understand how any of this works though, so correct me if I'm wrong.

Click to expand...

Hi,

In post-production, they are not using consumer LCD TVs.

The expensive post-production reference LCD from 5$K until ~30K$ is not suffering from de-saturation in the low end ...in comparison with consumer TV/ PC monitors.

The EIZO 3146 has the desaturation at 8% Luminance only.

All these monitors have been calibrated via 3D LUTs.

Your Home TV, according to your posts, has not calibrated with 3D LUT.

Have you verified the primaries/secondaries with 20-Point Luminance ramps to see the performance?

REC2020 is a container; all post-production HDR monitors can cover P3, so all movies are mastered for P3 and then mapped inside to REC2020 (as a container) at export.

Professional TV reviewers have to buy ColourSpace to see these charts because LightIllusion has never sent a single free license to reviewers; ....they have asked many times.

Professional TV reviews are sponsored (get free software) from other calibration company; it's the reason you see only regular charts.

So if one of those reviewers will buy ColourSpace and post reviews with 3D volumetric charts, the other company he is getting free software will complain or stop providing free license, so the professional reviewer will have to buy the other software.

Evil_Husky said:

Ok, I didn't know that, the post by by Rolls-Royce in the beginning of the thread saying that OLED is an outlier really confused me.

Click to expand...

Rolls-Royce is talking about consumer LCD or regular cost LCD PC monitors.

Some OLEDs have a different problem, and they expand saturation at lower luminance levels.

However, all these are not going to be measured when you will use ColorChecker or regular gamut measurements.

Look the Animated PNG of LG E6 for example:

The gray triangle is the gamut coverage, the value I'm changing is the Luminance range (255 is 100%)

The white triangle is the REC.709, the selected as target colorspace.

You can get these charts with gamut triangle with ColourSpaceZRO also.

Sony AF9 OLED was doing the same oversaturation at the low end, but that TV got all awards that year.

See 3D Charts (available in ColourSpace HTL).