| Forums | Register | Reply | Polls | Search | Statistics |
News, Gossip and Rumours Light Illusion Forums / News, Gossip and Rumours /  
 

Dado LG CX – Inaccuracies and Bollox

 
Author Steve

INF
Male
#1 | Posted: 3 Jan 2021 17:27 | Edited by: Steve 
I really didn't want to do this, but as I was asked by a rather large number of individuals do do an appraisal of the bollox in Dado's LG CX video, I have just wasted another hour or so of my life that I won't get back... basically a lot of sales based crap, as the video is just an attempt at an advertorial.
I have to assume Dado was paid in some way to do this, but if I was Portrait/Calman, I be asking for whatever the payment was back...

So, the assessment.
Based on the way the video was presented I've simply highlighted comments/sections in order they are presented, and tried to provided feedback on them.
As the video is basically just PR, some of the comments take the same approach. It's the only was I could get through the crap without going mad...
So this was a relatively quick review, and I may have missed points - but I really did struggle having to watch the crap yet again.

1) An Advertorial.
Not really a major inaccuracy, but Dado really should have stated it was an advertorial for Portrait/Calman, and not pretended otherwise.
(Free software anyone?)

2) Sony BVM X300 as best monitor?
No, as Sony do not provide for end user calibration, other than grey scale. In this day and age that is just not acceptable, as all displays drift an age.
That is a major reason Sony have lost their position as the best displays for grading.
The majority of our users with Sony displays employ external LUT boxes to enable accurate calibration.
(There are plenty of better display options available, for a lot less cost - do some research!)

3) WOLED, not OLED.
When Dado first came across the LG glass in a Panasonic he had no idea what WOLED was, and accused me of lying when I pointed out the issues with the additional white pixel, and gamut desaturation.
(Discussion still visible on the CML website.)
The issue with desaturation can be see here: https://www.lightspace.lightillusion.com/error.html#hdr_volume – select 'HDR Volumetric Accuracy'. RGB Separation shows the issue very well.

4) Human eye and saturation.
The concept that we (humans) cannot see a saturation change above 800 nits is incorrect.
The human eye can perceive saturation very well up into high brightness levels.
The eye's iris helps this further, as it adapts the response of the eye by limiting oversaturation.
As an average midday outdoors brightness is well over 800 nits (FFS, well over!), it is obvious what Dado says is just wrong.
Where we cannot perceive saturation as well is in low lights, as the eye swaps from cones to rods.

Additionally, should the concept of the eye not seeing saturation well as brightness increases have actually been correct, to concept of a display also desaturating does not make it ok – in fact the opposite. Having both the display desaturate, as well as the eye's response, would only make issues worse...

Further, no LG's can reach 800 nits – they max out between 600 to 700.
And even more importantly the LG actually suffers gamut desaturation from around 350 nits!
(This really was a stupid comment from Dado, and shows a real lack of understanding.)

5) The eye's response to brightness.
Brightness is a Log scale, so 1000 nits is just a doubling from 100 nits.
But, while technically true, that is not how our perception works.
Place a 1000 nits display next to a 100 nits displays, and the 1000 nits will look way brighter than the 100 nits display – not just twice as bright.
That has a real impact on how well different displays work as different brightness levels, especially WOLEDs with their gamut desaturation as brightness increases.
(This has a lot to do with the surround illumination, which Dado really should understand.)

6) Netflix and HDR.
No WOLED meets Netflix HDR grading display specifications, for quite a few reasons reasons...
The gamut desaturation is one.
As is failure of an L32 patch profile test.
And obviously not making the required min peak luma.
If you are thinking of using an LG for real grading, HDR is NOT possible, regardless what is said by Dado.
(But the LG can be calibrated well for SDR, using internal 3D LUTs)

7) The LG CX is not the only LG that can be calibrated.
2020 BX/CX/GX/WX/ZX, 2019 B9/C9/E9/W9/Z9/R9, and 2018 B8/C8/E8/G8/W8, and 2017 B7/C7/E7/G7/W7 models with Technicolor 'Studio' Firmware, can all be calibrated.
https://www.lightillusion.com/lg.html
Steve Shaw
Mob Boss at Light Illusion

Author Steve

INF
Male
#2 | Posted: 3 Jan 2021 17:37 
8) Calman is NOT the 'only' calibration software that works.
In fact, it is not regarded as the best, as many independent tests have proven that better results are attained via ColourSpace and LightSpace.
(A selection of independent feedback can be found here: https://www.lightillusion.com/forums/index.php?action=vthread&forum=13&topic=374 - as well as searching the likes of AVSforum - https://www.avsforum.com/threads/colourspace-cms-next-generation-calibration-thread.3049142/page-23#post-59708598 - etc.)

And for a lot of the described workflows, Calman Studio is needed, which is very expensive. Plus, all Calman versions have yearly maintenance fees – with the 'Home' versions having to be re-purchased every year in full.
(But, what do you expect from an Advertorial that is promoting sales bollox..?)

9) C6 Probe
The C6 probe is just an X-Rite OEM probe in drag, and far more expensive.
There is nothing special about the C6 probe, and even the supplied 'certification certificate' is inaccurate, as it is not NIST compliant.
(And it is just a certification, not a calibration certificate, performed with an expired meter, as shown here: https://www.avsforum.com/threads/c6-hdr2000-colorimeter-from-spectracal.2896137/page-7#post-60124571 )
Go with the standard X-Rite OEM i1D3, or even the i1D3 Plus, as both are identical, and a lot cheaper.
Additionally, LightSpace/ColourSpace has an EDR for WOLEDs that will work with any X-Rite i1D3 probe, and is not restricted as with the C6 options.

10) Calman and LG connection.
Rather than using Calman for the display connection, use the Free DeviceControl.
(To work with LG TVs DeviceControl is free to ColourSpace users.)
See https://displaycalibrations.com/lg_templates_for_device_control.html
and
https://www.avsforum.com/threads/colourspace-cms-next-generation-calibration-thread.3049142/page-48#post-60106862

11) WiFi
While Wifi can be used with LG Tvs, we do not not recommended its use for calibration.
There are far too many issues with interrupts, packet failures, and more, that may be missed during calibration, but will cause errors and artefacts in the final calibration.

12) Pattern generator.
While the CX does have an internal TPG (but older models do not), it has been proven to not be bit-accurate, and randomly flashes inaccurate colours, mainly magenta.
It is far better to use an external TPG so that the whole signal path is profiled. Profiling the whole signal path really is recommended.
Davinci can be used, or the Free PGenerator: https://www.lightillusion.com/pgenerator.html - with a HD Fury as needed for HDR metadata.

13) TPC and GSR
Disabling TPC and GSR will not totally stop ABL type issues, so be aware...
(TPC is for static images as well.)
The LG will draw the same current even without disabling TCP/GSR, just it will not 'dim' the picture with paused frames

14) Connecting to the LG TV.
As mentioned before, the Free DeviceControl (for ColourSpace users) is better for LG management.

15) Calman Requirements
There are also a number of inaccuracies with the requirements for Calman use.
With some of the suggested approaches, an Ultralink card is needed, and that cannot be used with Calman LG Home, and needs the very expensing Calman Studio version, as well as paying yearly maintenance fees, as mentioned previously – be aware of overall costs, as they are not as suggested.

16) HDR Settings
The HDR settings within the TV should really be set to values that match the display's capabilities, not idealised values. But, as the LG is not a mastering display this is likely to be problematic... Probably sticking to fixed values, even if they are not actually correct will minimise issues – but this has not been explained well at all. The obvious error being the use of 1000 nits, when the display cannot make that.

17) C6 Probe again...
More bollox – it is JUST an i1D3 OEM in drag, with a very high ticket price.

18) Next, Next, Next...
Yeah, very restrictive and means you do not understand when things go wrong.
But, that is a user's choice – personally a less restrictive approach always generates better end results, as independent testing has proven time and again.
(It's also obvious Dado has no real idea what he is actually doing/saying, as he has connected to the TV as an 'LCD 2019'... the probe in use changes from a C6 to the later C6 HDR2000... and the 'Post Calibration' graphs show no calibration data at all, as well as randomly showing SDR Calman screens, when talking about HDR... Just more sales Bollox slide show images, with no valid information/data...)

19) Judd Vos Offsets.
No! The Judd Vos offset is for RGB OLEDs only NOT WOLEDs!
(See https://www.lightspace.lightillusion.com/perceptual_colour_match.html )
And trying to use it will never match the Sony Dado is so keen on!

20) Autocal dE Formula
No calibration system should EVER use a dE formula for calibration...
It really is that simple.
It doesn't matter what the dE formula is, it is just wrong to use it for calibration – ever.
The use of dE is just for an assessment of calibration accuracy, after the calibration has been performed, and should never be the actual basis for calibration.
This goes a long way to helping understand why Calman often has issues with calibration accuracy...

21) Legal Range during calibration
The TV actually uses 16-255 range for SDR, so be aware of that, as if that is not used the results will not be as expected. HDR is 16-235.

22) 3D LUT
No, a Matrix is NOT a 3D LUT, as it uses just 5 patches, and no, the patch sequence is not in the slightest bit important for a matrix – you have to use the same 5 patches regardless... That really is just 'sales bollox' to suggest otherwise.
(See https://www.lightspace.lightillusion.com/error.html#matrix_lut )

Also, with the LG no 3D LUT can be used for HDR (as the instability of the display means the required profiling just does not not work with Calman).

A 3D LUT calibration would work with ColourSpace, as it has the required capabilities to overcome the inherent instability issues...
This is proven by the FSI WOLED displays, which use the same LG glass, and can be 3D LUT calibrated with ColourSpace/LightSpace, but not with Calman.
(See https://flandersscientific.com/calibration/, where it states *Note: CalMAN can only be used for SDR calibration. The same is true with ASUS Proart HDR displays, with Calman incapable of performing calibration.)

For SDR a 3D LUT can be used, but still Calman can only work acceptable/reliably with limited patch sequences (Lightning LUT), while the best SDR results are generated with ColourSpace, and large profiles sets.
(See the previous links that compare the issues with Calman calibration vs. ColourSpace/LightSpace.)

23) Tone mapping
There is no mention of Tone Mapping at all, which again shows the lack of understanding Dado actually has with regard to calibration.
For any attempt as using any display for HDR work (rather than home viewing) any tone mapping must be disabled.
https://displaycalibrations.com/lg_templates_for_device_control.html
and
https://www.avsforum.com/threads/2019-lg-oled-calibration-and-user-settings-no-price-talk.3042154/page-27#post-58518000
and
https://www.avsforum.com/threads/lg-oled-2019-pq-curve-upload-free-template-for-devicecontrol-interface.3088376/#post-58608526
(Ted helped Portrait/Calman correct their inaccurate PQ data.)

So, after wasting another hour or more of my life, it really is just paid-for sales bollox, with a lot of inaccurate statements, and little that would expand anyone's understanding or knowledge.
After the issue we have with https://www.lightillusion.com/forums/index.php?action=vthread&forum=13&topic=352 our work, I guess it is not really a surprise...

Some links for research to further understand the reality of LG calibration, and potential issues associated with different calibration approaches.
The LG TVs that can be 'calibrated' are: 2020 BX/CX/GX/WX/ZX, 2019 B9/C9/E9/W9/Z9/R9, and 2018 B8/C8/E8/G8/W8, and 2017 B7/C7/E7/G7/W7 models with Technicolor 'Studio' Firmware.

https://www.lightillusion.com/lg.html
https://www.avsforum.com/threads/colourspace-cms-next-generation-calibration-thread.3049142/page-23#post-59708598
https://www.avsforum.com/threads/2020-lg-oled-calibration-and-user-settings-no-price-talk.3113174/page-42#post-60162967
https://www.avsforum.com/threads/dedicated-raspberry-pi-pgenerator-thread-set-up-configuration-updates-special-features-general-usage-tips.3167475/page-21#post-60364650
https://www.avsforum.com/threads/colourspace-cms-next-generation-calibration-thread.3049142/page-48#post-60106862
Steve Shaw
Mob Boss at Light Illusion

Author Steve

INF
Male
#3 | Posted: 4 Jan 2021 13:39 
Actually, Ted at DisplayCalibrations.com has done a far better report on the bollox in Dado's video - far more involved than my basic review, with follow-up information:

https://liftgammagain.com/forum/index.php?threads/buying-a-new-lg-cx.15380/#post-152742

Steve
Steve Shaw
Mob Boss at Light Illusion

Author editsupport
ZRO
#4 | Posted: 8 Jan 2021 14:43 
Really, that hour of your life was not wasted , at least I again learned a lot from it. Much appreciated.
Martin

Author Steve

INF
Male
#5 | Posted: 8 Jan 2021 18:48 
Thanks Martin - that is much appreciated.
We do try to call-out bollox when it is presented as fact.
(And there has been a fair bit of that...)

Steve
Steve Shaw
Mob Boss at Light Illusion

News, Gossip and Rumours Light Illusion Forums / News, Gossip and Rumours /
 Dado LG CX – Inaccuracies and Bollox

Your Reply Click this icon to move up to the quoted message


Watch this topic / Get updates by Email
 (?)
Only registered users are allowed to post here. Please, enter your username/password details when posting a message, or register first.

 

 
Online now: Guests - 1
Members - 0
Max. ever online: 264 [6 Jun 2020 12:02]
Guests - 264 / Members - 0