HCFR - Open source projector and display calibration software - Page 360 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 1444Likes
Reply
 
Thread Tools
post #10771 of 12074 Old 06-01-2018, 04:16 PM
 
skschatzman's Avatar
 
Join Date: Oct 2016
Posts: 3,750
Mentioned: 80 Post(s)
Tagged: 1 Thread(s)
Quoted: 2838 Post(s)
Liked: 1655
Quote:
Originally Posted by CraftyClown View Post
Thank you, that makes sense that the bt.1886 curve is relative to my displays black level.

So essentially I just need to go in and tighten up my calibration and bring everything a little closer to 100% and in theory I should be closer to a gamma of 2.4.
No. Bt.1886 only becomes 2.4 with a black reading of 0. Basically only OLEDs right now.
skschatzman is offline  
Sponsored Links
Advertisement
 
post #10772 of 12074 Old 06-01-2018, 04:33 PM
Member
 
Join Date: Nov 2017
Posts: 83
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 75 Post(s)
Liked: 0
Quote:
Originally Posted by skschatzman View Post
Quote:
Originally Posted by CraftyClown View Post
Thank you, that makes sense that the bt.1886 curve is relative to my displays black level.

So essentially I just need to go in and tighten up my calibration and bring everything a little closer to 100% and in theory I should be closer to a gamma of 2.4.
No. Bt.1886 only becomes 2.4 with a black reading of 0. Basically only OLEDs right now.
Thanks, yes I understand that without the absolute black that an OLED might provide, a full 2.4 is not possible with bt.1886. I presumed however it might be possible to get something in the high 2.3s though?
CraftyClown is offline  
post #10773 of 12074 Old 06-01-2018, 05:49 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 7,107
Mentioned: 132 Post(s)
Tagged: 0 Thread(s)
Quoted: 5000 Post(s)
Liked: 1922
Quote:
Originally Posted by CraftyClown View Post
Right I see, so with my black level being 0.025 should I hopefully be expecting closer to 2.4 than 2.3 then?
With White at 120 and black at 0.025, the gamma reaches 2.3 from 50% and up; but the average is 2.27.
If you prefer a higher average gamma, set "Effective @50%" to something like 2.37. That would give you a gamma approaching 2.4.

Quote:
It really is a balancing act getting gamma, white balance and colour calibrated correctly
If you use green to adjust gamma and then use red and blue to correct white balance, the latter should have minimal effect on the gamma, especially if you go thorough a second iteration.

Last edited by Dominic Chan; 06-01-2018 at 07:04 PM.
Dominic Chan is online now  
Sponsored Links
Advertisement
 
post #10774 of 12074 Old 06-01-2018, 10:41 PM
AVS Forum Special Member
 
Matt L's Avatar
 
Join Date: Aug 2000
Location: Flushing, MI
Posts: 5,968
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 705 Post(s)
Liked: 583
Quote:
Originally Posted by jt7272 View Post
So the issue seems to be that the RT-AC68U (running the latest firmware) is causing issues with mDNS and the CC. If the CC is connected to the RT-AC68U (either 2.4 GHz or 5.0 GHz) then mDNS discovery does not work. After some googling it seems that there are known issues with mDNS and Asus routers/AP's.
Hmm, I'm running HCFR and Chromecast on my Asus RT-AC87U with zero issues... Both my laptop and my chromecast are wireless on 2.4 band.
Matt L is offline  
post #10775 of 12074 Old 06-02-2018, 12:40 AM
Senior Member
 
Kamikaze_Ice's Avatar
 
Join Date: Jan 2007
Posts: 366
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 211 Post(s)
Liked: 115
Quote:
Originally Posted by webdove View Post
2. When running dense (1% or 2% spacing) automatic gray sweeps at low brightness some kind of LG dimmer seems to activate causing anomalously low readings. Pressing the roller/ok button every few seconds on the remote to pop open the selected input window in the upper left appears to defeat this without disrupting measurements. Similarly when I am tuning adjustments to white balance or brightness with a low level pattern if I backup up from the bottom of the screen adjustment menu to the half screen list of adjustments window every few seconds it seems to prevent that dimmer from kicking in. When the displayed pattern is above about 50% brightness in a 5% of screensize window that signal alone appears to prevent the dimmer from activating. It is odd that they designed in a dimmer for when the screen is dark.
I have an E6. This may be the ABSL "feature" which dims screen, saves power, helps reduce image burn in but not retention. You can disable that in the service menu for 2016 models.
I don't know if this happens on yours, but whenever I open the menu a little "loading" cursor appears at the center of the screen. I place my meters just to the right of it to keep it from messing with measurements.
There is some other silly behavior regarding the menus, brightness and contrast (see below).

Quote:
Originally Posted by webdove View Post
3. As someone mentioned a while back, heating the pixels causes increased luminance for the same input code. Automatic scans of high brightness values will cause this, so you have to redo automatic readings above 50 by hand to correct them. It seems to take about 20 seconds for them to cool. Using a 40% gray wallpaper helps keep everything stable.
This is how I've been doing it, but with a 30% grey (or Masciola's "ABL reset pattern"). This is the first thing I balance because it too will change measurements slightly.
If I used a black screen, I found red would measure higher and show more fluctuation than the others <50%
I also manually measure exactly for that luminance drifting, which is really just green skyrocketing much faster than red/blue. My green for 100% white can go from 650 nits to 730 nits in about 30 seconds while red/blue get up to 675-680 or so.

Quote:
Originally Posted by webdove View Post
8. I found that white balance adjustments in HDR (std) change which digital codes (and brightness percentages) that they affect if you alter either brightness or contrast, but those adjustments do not shift the codes that they affect if you alter OLED. Thus the brightness and conrast controls in software must be ahead of the look-up table and the OLED control must be after the look-up table. To learn this I set all adjustments to zero, then I boosted green by 20 counts for codes 637 and 482. I then adjusted Contrast and Brightness over a span of value pairs that resulted in a maximum of “greenness” at one or the other of those two input codes (e.g I aligned the adjustment numbered 637 so that its affect was centered around input data point 637).
For adjustment code 637 the contrast/brightness pairs were: 76/54, 77/53, 78/51, 79/48, 86/40.
For adjustment code 482 the contrast/brightness pairs were: 87/54, 89/53, 91/52, 93/51, 94/50, 97/48.
Sadly, there is no combination of contrast and brightness settings on my E6 that causes all the white balance adjustment points to actually land at the input codes that LG says they affect (what a pain!!). This means all the effort that people like Ryan put into creating LG OLED calibration patterns is useless for me for adjust my HDR white balance.
In order to adjust white balance I picked contrast 100 since that was the “reset” value, I used OLED 67 to get a diffuse white value at 50% brightness of 94.38 as Zoyd recommended and I set brightness to 52 which gave me the best fit to BT.2390 min.002, max 467 at 2% brightness steps between 0 and 20%. To allow precision measurements I set measures/parameters/grayscale_levels to 100 (I would have preferred 200 for more precision, but HCFR won’t let you type in a number more than 100. Zoyd, allowing that to be 200 would be helpful). I then went through every one of the 20 white balance adjustments individually, set green up 20, red and blue down 20 and stepped through the 1% brighness levels comparing with an unadjusted reference and noted which brightness values were shifted towards green from the lowest to the highest brightness and what the centerpoint of the brightness range was. This gave me a table of “range of brightness levels affected” for each “white balance adjustment code”. This is the only way that I found would allow me to properly adjust my white balance.
First, I don't know what or if any procedures LG has published for your model, but the one for 2016s are literally impossible to follow.
I try to follow their pattern targets heavily assuming that because they're unique there is underlying tone mapping behavior (formula?) and are based on these targets.

This is something I spent a lot of time doing for my display (hdr control alignment). Do note that any of this may not apply to your display.
But first I want to talk about that silly behavior I mentioned above...
You can see this behavior I'm about to describe with a gradient ramp. Just set something like control point 591 to +50 red.
Lower contrast from 100 to say, 85. You can see the red portion move. Make note of where it is on the screen (I used a sticky note). Now change display preset to GAME and back to standard. For me the red portion would NOT be where my sticky note was, every time. Now if you go back into the 20-point controls and change any control point by +/-1 the red portion would now shift and go back to my sticky note.
This "shifting" also happens in SDR.

My conclusion is that white balance controls are locks in place until something makes the display update it's formula(s) for whatever internal functions or calculations it is doing.
Basically any time I change brightness or contrast I need to either switch inputs, enter/leave game mode, or toggle dynamic contrast to ensure measurements are taken with the "current" alignment.
This shifting can happen at any point though, so I like to measure 50% before and after every measurement and remeasure when different.
Hell I swap in/out of game mode after making any changes even unrelated ones, and periodically when doing manual sweeps of <50% patterns.

That was something I discovered while trying to align the HDR 20-point controls. Also, each control point needs a different combination of brightness/contrast to align the pattern to the control. Matching one will throw off another. The most troublesome ones for my E6 are 622, 637, 653 and 668. Remember, this may not apply to your display.
I choose to best match 637 and 653.

I also make use of the Service Menu controls.
I think your 2017 model has 2-point controls in HDR modes, but the 2016's don't so I use the 2-point service menu controls to set to best distribute error everything below control point 254. 127 controls are way to strong, a +/-1 is like +/-50 for any other control). I use the "warm" set for SDR and the "Neutral" set for HDR (only warm/cool/neutral in my service menu, Warm2 still uses service menu "warm" as it's base).
Sub-Brightness 128 (default)
Sub-Contrast 127 (128 is default)
OLED 99
Contrast 92
Brightness 52
This is a pretty good alignment, but for control points 419-653 they are all shifted up one. What I mean is 482 controls are used to adjust a 498 pattern, 498 controls for 513, and so on.
A 668 pattern is not the peak of my E6, but IMO this is where the tone mapping rolloff is intended to begin. And with the controls being shifted up once, I can now use 668 controls to balance everything above 668 and remains linear as long as 100% matches the balance of a 668 pattern. Takes a bit of going back and forth between 653 and 668 controls.
I also have red and green slightly higher due to how fast green luminance skyrockets. They're set between a manual measure and what they reach after 60 seconds of continuous measurements (aka splitting the difference so tinting is not an issue).


I can't stress enough that this is what I do due to how MY DISPLAY behaves, and that this is far from typical procedures, so take all of this with a grain of salt lol.

----

Quote:
Originally Posted by CraftyClown View Post
Haha, just when I think I've got it, I realise I have more to learn &#x1f601;

So when you say adjust the green to set the correct gamma is it as simple as ensuring it is 100% on the grayscale or should I be looking somewhere else to do this?

I'm also curious as to what average gamma I should be expecting with BT.1886?
I don't know what your controls look like, but most displays I've seen let you adjust red green and blue for a particular point (2/10/20 point white balance).

On the right side of your table, and above the measure buttons you will find some radio buttons. Set it to RGB, so the first three rows of the table (xyY) will be replaced with RGB luminance.
Then just make the three of them match your target Y.
Kamikaze_Ice is offline  
post #10776 of 12074 Old 06-02-2018, 09:47 AM
Member
 
jt7272's Avatar
 
Join Date: Jan 2012
Location: Pittsburgh, PA
Posts: 80
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 50 Post(s)
Liked: 46
Quote:
Originally Posted by Matt L View Post
Hmm, I'm running HCFR and Chromecast on my Asus RT-AC87U with zero issues... Both my laptop and my chromecast are wireless on 2.4 band.
OTE=Matt L;56279048]Hmm, I'm running HCFR and Chromecast on my Asus RT-AC87U with zero issues... Both my laptop and my chromecast are wireless on 2.4 band.[/QUOTE]

Eh, that doesn't mean a whole lot as they are different HW, FW, etc. The 66u and 68u are quite similar and are known to have mDNS/Bonjour issues. The 87u is not reported to have these issues.



Some have reported that the issue goes away when using Merlin FW but others still have it. You can find a lot of info on the SmallNetBuilder forums

Epson 6040ub
Denon AVR-X3400H + Mirage OMD-5 surrounds
Home-made Hybrid PCB Electrostatic Floorstanding L+R Speakers (with built-in DSP + amps)
Home-made sub with 15" Ultimax woofer, Dayton SA1000 Amp
Sony PS4, Sony UBP-X800, Panasonic DP-UB820
jt7272 is online now  
post #10777 of 12074 Old 06-02-2018, 10:51 AM
Member
 
webdove's Avatar
 
Join Date: Jul 2006
Location: NH
Posts: 59
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 6
Thanks K. Actually mine is a 2016 E6 like yours should they should behave similarly.

I will have to try some sweep experiments on one of the adjustments to see if Brightness and Contrast are "sticky" like you have noticed.

I have a few questions:
1. Where can I find documentation on using and the effects of the service menu adjustments?

2. Does your TV have the same dramatic green luminance falloff with saturation that mine does?

3. Have you also noticed that HCFR version 3.5.1.6 reports argyll driver errors on any attempt to open a new adjustment window after the first one and version 3.5.1.4 does not?

Attached is my most recent attempt at HDR cal using Zoyd's recommended Diffuse white target of 94.
Attached Thumbnails
Click image for larger version

Name:	Sat Sweep Lum 2018-05-11a.png
Views:	31
Size:	38.1 KB
ID:	2410768  
Attached Files
File Type: zip LG E6 2018-05-11a HDR smooth cal O67, c100, b52, c55ext GDI(overlay) RGB8 i1pro2-d3 all.zip (24.5 KB, 7 views)
webdove is offline  
post #10778 of 12074 Old 06-02-2018, 09:13 PM
Senior Member
 
Kamikaze_Ice's Avatar
 
Join Date: Jan 2007
Posts: 366
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 211 Post(s)
Liked: 115
Quote:
Originally Posted by webdove View Post
Thanks K. Actually mine is a 2016 E6 like yours should they should behave similarly.

I will have to try some sweep experiments on one of the adjustments to see if Brightness and Contrast are "sticky" like you have noticed.

I have a few questions:
1. Where can I find documentation on using and the effects of the service menu adjustments?

2. Does your TV have the same dramatic green luminance falloff with saturation that mine does?

3. Have you also noticed that HCFR version 3.5.1.6 reports argyll driver errors on any attempt to open a new adjustment window after the first one and version 3.5.1.4 does not?

Attached is my most recent attempt at HDR cal using Zoyd's recommended Diffuse white target of 94.
Are you using the most recent firmware and the "PC Icon mode"?
This restricts ONLY HDR for this icon mode to a 709 gamut within bt.2020
I had to downgrade firmware to get a proper coverage back.

I was stupid and forgot to write down my HDR mode settings, so I'm still re-creating them.
I'm not at the color stage yet, working on grayscales for HDR on multiple devices (pc, built in apps, anything going through home theater receiver, game consoles (ps4/xb1x).

I've found most of my trouble is being caused by the PC icon's 4:4:4 chroma, which is made worse thanks to the state of bit-depth issues of current Nvidia drivers and Windows. [email protected] bit is far cleaner and smoother than any other combination, and ycbcr conversions make this it slightly more visible in secondary colors, especially magenta. Not to mention the differences between Nvidia's HDR api and Windows HDR (and/or full screen exclusive mode to output 10/12 bit depth instead of the forced 8 bit in full screen windowed that Microsoft has yet to address...)

I'd really like for HCFR to allow MadTPG to use exclusive full screen mode for HDR measurements. I can't confirm, but I've heard CalMAN has an option on their end which does this so I am assuming this would need to be implemented on HCFR's end. Could anyone who has both software please confirm/deny this?

Non-PC Icon modes are far easier to work with IMO. Chroma conversions are cleaner and more linear for these modes, even if the content is the same (4:2:0), but aside from movies I want the 4:4:4 chroma. Clean text and other fine details are far more important in this case, IMO.

Also, I looked at your .chc, any particular reason you're using 0.002 as minimum target level and a max of 467?
Either way your panel looks like it behaves much different than mine.
I don't use GDI for measuring, as this is controlled entirely by Windows. It has it's own issues (bit depth, rgb<-->ycbcr conversions, bad dithering) and changes nearly every update. I use MadVR for movies which is another reason I don't like GDI.
I'll just use a 3D LUT for SDR stuff anyways (eeColor, or via Reshade if I want a resolution larger than 1920x1080 for SDR. Usually I downsample to 1920x1080 and route PC through eeColor)

Re: #1
Please be aware that any changes in the service menu are your choice. If you screw something up, it is completely your fault and consequences are yours alone to bear.
That said, there's not much documentation (at least for us non-LG techs or engineers, see service manual link below) but there are only a few things worth changing anyways.
There are two service menus by the way, IN_START and EZ_ADJUST. Passcode to enter them is 0413
Don't do anything with IN_START, except for toggling ASBL on/off. Don't play with anything else even if an option sounds obviously safe to change.
EZ_ADJUST has White Balance and Sub-B/C sections, the only useful ones.
Don't touch ADC calibration.
Don't hit the "reset" button on the white balance page (different "defaults" than what mine started with and causes a red flagged status for ADC calibration).

Do note that these controls will effect EVERYTHING. SDR, HDR, all icon modes, all inputs, all apps.

20-point controls don't work with HDR, but will stack with the user menu 20-point controls. Use them if you want to avoid the ISF modes (like using Game for slightly lower-than-PC-icon-mode input latency). Dosen't help color though (no CMS alternative in service menus so game modes are limited to "preferred color" and I have no clue what this actually does to the gamut)

The "White Balance" controls can be used as a global 2-point high (gain)/low (cut) controls. Because they work for HDR I use them as a replacement for HDR control point 127, as +/-1 adjustment to 127 are way too strong. This requires stronger adjustments to control points 254-482 to compensate for this method.
I use the "Neutral" set to keep HDR's balance separate from the "Warm" set I use for HDR (With SDR being set to Warm2)

I just set my Gains based on color clipping. Red clips first at 192 (Warm default) and I balance around that.
The Sub-B/C controls adjust the strength of what Brightness and Contrast (in user menu, not service menu) are set to.
Basically sub-B/C are fine/small adjustments while Brightness/Contrast are coarse/large adjustments.
Setting Sub-Contrast to 127 has been the best for me (helps aligns HDR 20-point controls with minimal impact on peak brightness as Contrast can stay higher than it would have been if Sub-C was 128 with a similar control alignment.


E6 Service Manual: https://lg.encompass.com/shop/model_...OLED55E6PU.pdf
*You can change the "model.pdf" portion to get service manuals for other displays. Not all of them will match (ex: OLED65E6PU.pdf = service manual for 65UH7700)
The service menu doesn't have much about calibration, and some of their instructions I find are questionable.
Almost every LG panel has instructions that basically state that setting one of the rgb gains to 192 and lower others to keep will keep "full dynamic range". This has been the case for my three LG displays (E6, LM7600 and some first gen webOS model I forget.
In the B6 service manual they say to set white while using VIVID and that any other way would be an error... (wtf? I mean I can understand why for luminance but there's more to setting white than just being bright...)
In the E6 service manual they say to adjust balance without effecting saturations to set to 129-255 (can't make saturations worse if they're already clipped...)

Re:#2:
I've never noticed anything like this. If the "O67" in the file name means OLED Light is set to 67, this might be why.
For SDR try measuring smaller windows (like 5-8%).
I leave OLED at 100 and lower Contrast to 52, and use 2-point low and service menu "warm" CUTS to get black where it belongs while having a 100% white at ~130 nits.
I do it this way as I've found the number value for Contrast is being used to determine the trigger point for screen dimming (that we can't disable, separate from ASBL).
Somewhere in the forum listed what I measured for a 100% white at various window sizes (5% increments, IIRC), and the difference between High OLED/Low contrast vs low OLED/High contrast. The two setup were practically identical in all regards except those two settings. The difference between the two with a 100% white full screen field (100% window) was a difference of ~35-40 nits the former setup was brighter than the latter, and being a gamer there was much less distracting dimming fluctuations.

I don't think any of that is related to what you're seeing with green though.
If you're using HCFR's saturation measurement pages, try using the newer "color checker" patterns instead (change them via bottom of References tab in Preferences).
You will get different results than you do from the dedicated saturation pages. I'm sure someone here could explain why much better than I could.


Re:#3:
I did have this issue. I believe Zoyd said something about different agryll drivers not fully integrated yet or something like that, which is probably why.
I just went back to 3.5.1.4.
I was getting random bogus pattern display errors with MadTPG though. Checking the "OSD" box in the generator options made them stop (OSD does NOT get shown despite being checked, and I'm fine with that). This is probably in madshi's court as I found this from someone who was having this issue with DisplayCAL. I wouldn't have thought to "enable" the OSD would do anything for this because I didn't want it on lol.
Kamikaze_Ice is offline  
post #10779 of 12074 Old 06-03-2018, 03:34 AM
Newbie
 
Join Date: Sep 2017
Posts: 11
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 10 Post(s)
Liked: 0
Hi,

I want to measure and calibrate HDR10 on my LG B7 - to be precise I want to to 2-pt WB only and do a reference measurement before! Unfortunately, I am not sure on how-to do it.

1) 2-pt WB calibration: Do I only aim for RGB balance (= each 100%) or should I also try to adjust the luminance response?

2) Which patterns should I use from the "Mehanik HDR10 calibration and test patterns set" disc: "HCFR\01. Grayscale\02. 5% steps" or "01. Grayscale\04. LG OLED 2017 steps - 1000nit". Both are specified with MaxCLL: 1000 cd/m2, MaxFALL: 400 cd/m2

3) Which are the correct preference setting of HCFR for these HDR patterns and my TV?
- Rec2020 or Rec2020/P3?
- BT.2390 on/off?
- Values for Master MinL, Master MaxL, Content MaxL, Frame Avg. MaxL?

Thank you in advanced!
maxOLED is offline  
post #10780 of 12074 Old 06-03-2018, 05:28 AM
Member
 
stama's Avatar
 
Join Date: Feb 2013
Posts: 75
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 71 Post(s)
Liked: 43
@zoyd

I have a feature request and a bug report, which would help with HDR calibration:

- feature request: enter HDR mode when starting the measurements, and exit HDR mode when finishing the measurements.

At the moment, the HDR mode is exited not when the measurements stop, but when closing HCFR. The HCFR user interface is burned out and reading what's written on it is almost impossible while HDR mode is on. And if you have an OLED, then you also get image retention in the rectangle where the HCFR window is.

- bug: start a new measurement project for HDR, make sure on the measurements instrument settings page the "write debug info" is disabled, save it, close it. Now open it again. Go to the measurements instrument settings page, and you will see the "write debug info" is enabled.

The side-effect of this is that when the measurements process is started in fullscreen mode, there will be a popup dialog telling you the integration time has changed and which waits for you to press the OK button. But immediately after, the fullscreen pattern is displayed, hiding the popup dialog. And you wait, and wait, and wait, and don't understand why the pattern does not change on the screen. Well, that's because you did not notice the small popup dialog the few miliseconds it was visible, and don't know you had to press the OK button on it.


All of these issues make things difficult for those that use the TV as the display where the HCFR user interface is displayed, and it would be great if they would be fixed.
stama is offline  
post #10781 of 12074 Old 06-03-2018, 06:21 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 7,107
Mentioned: 132 Post(s)
Tagged: 0 Thread(s)
Quoted: 5000 Post(s)
Liked: 1922
Quote:
Originally Posted by maxOLED View Post
Hi,

I want to measure and calibrate HDR10 on my LG B7 - to be precise I want to to 2-pt WB only and do a reference measurement before! Unfortunately, I am not sure on how-to do it.

1) 2-pt WB calibration: Do I only aim for RGB balance (= each 100%) or should I also try to adjust the luminance response?
For 2-pt WB only aim for RGB balance. Make sure the "w/Gamma" option is turned off (it is by default).

Quote:
2) Which patterns should I use from the "Mehanik HDR10 calibration and test patterns set" disc: "HCFR\01. Grayscale\02. 5% steps" or "01. Grayscale\04. LG OLED 2017 steps - 1000nit". Both are specified with MaxCLL: 1000 cd/m2, MaxFALL: 400 cd/m2
Better ask Mehanik for this one. My guess is the latter.

Quote:
3) Which are the correct preference setting of HCFR for these HDR patterns and my TV?
- Rec2020 or Rec2020/P3?
- BT.2390 on/off?
Use Rec2020/P3, BT.2390 on.
See also:
https://www.avsforum.com/forum/139-di...l#post55853852

Last edited by Dominic Chan; 06-03-2018 at 06:51 AM.
Dominic Chan is online now  
post #10782 of 12074 Old 06-03-2018, 11:52 AM
Newbie
 
Join Date: Jun 2018
Posts: 12
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 0
Hi,

this is the first time I write. First of all I want to thank you for your great support.

I need your help with HCFR, I would like to calibrate my VPR Epson HC3500.

Are this settings right?

Display Type: non-refresh

Reading Type: Telephoto

Can I set Image Area equal to 100%?

I use ColorMunki Display

Thank you
Silver12 is offline  
post #10783 of 12074 Old 06-03-2018, 12:21 PM
Newbie
 
Join Date: Sep 2017
Posts: 11
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 10 Post(s)
Liked: 0
@Dominic Chan
Thank for your answers!

---------------------

I calibrated 2-pt WB of my LG B7 for HDR10.

I used the patterns "HCFR\01. Grayscale\02. 10% steps" from "Mehanik HDR10 calibration and test patterns set" as the "04. LG OLED 2017 steps - 1000nit" patterns did not matched the "target" luminance response" at all. "10% steps" instead of "5% steps" simply to reduce time.

I used following HCFR settings (still not quiet sure if they are 100% correct):

- Use 10-bit targets
- Rec2020/P3
- SMPTE 2084 HDR with BT.2359, Maxer MinL = 0, Maxer MaxL = 1000, Content MaxL = 1000, Frame Avg. MaxL = 400

I attached before and after screenshots ans the HCFR projects. I used 2-pt WB only.

My settings for my "calibration v4" are: low = 0,0,-1; high = -1,-3,0.

I was able to reduce the overall dE value. 40% - 60% have still the largest error. I was not not able to improve these values without increasing dE significant for the others. The overall slop of my color temperature curve is also in principle uncanged. However, after calibration is is a little bit closer to 6500K.

Does someone know if a better results should be principle possible using 2-pt wb only for the LG B7 - perhaps by using more "extreme values" for low and high? Any other comments on my calibration result?

Do you recommend to do 20-pt as well?
Attached Thumbnails
Click image for larger version

Name:	HDR 10 - without calibration.JPG
Views:	56
Size:	221.8 KB
ID:	2411274   Click image for larger version

Name:	HDR 10 - with calibration v4.JPG
Views:	49
Size:	216.3 KB
ID:	2411276  
Attached Files
File Type: zip HDR10 calibration.zip (2.5 KB, 10 views)
maxOLED is offline  
post #10784 of 12074 Old 06-03-2018, 01:36 PM
AVS Forum Special Member
 
Stuntman_Mike's Avatar
 
Join Date: Feb 2009
Location: NYC
Posts: 1,820
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 791 Post(s)
Liked: 626
I'm going to attempt an HDR10 grayscale calibration. I have a 65" OLED E6P. I have HDR patterns that I downloaded from another thread. My question is, does anyone know if running the patterns from the USB of the OLED has any adverse effect on accuracy? Do I need to run them from my PC instead? If so, what would be the best player to run them in?

My plan was to just use a thumb drive with the patterns, which successfully triggers HDR, make white balance adjustments, and then apply changes to all inputs. Just occurred to me that this may not be optimal for whatever reason, so just wanted to check and make sure.

If I need to use a PC, if someone could tell me what was the correct setting for RGB level low/high for the display and PC, I'd appreciate it. I remember reading in this thread that there was some confusion as what needed to be set to what, but I can't find the posts.


Thanks

Stuntman_Mike is online now  
post #10785 of 12074 Old 06-03-2018, 01:56 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 7,107
Mentioned: 132 Post(s)
Tagged: 0 Thread(s)
Quoted: 5000 Post(s)
Liked: 1922
Quote:
Originally Posted by maxOLED View Post
@Dominic Chan
Thank for your answers!

---------------------

I calibrated 2-pt WB of my LG B7 for HDR10.

I used the patterns "HCFR\01. Grayscale\02. 10% steps" from "Mehanik HDR10 calibration and test patterns set" as the "04. LG OLED 2017 steps - 1000nit" patterns did not matched the "target" luminance response" at all. "10% steps" instead of "5% steps" simply to reduce time.

I used following HCFR settings (still not quiet sure if they are 100% correct):

- Use 10-bit targets
- Rec2020/P3
- SMPTE 2084 HDR with BT.2359, Maxer MinL = 0, Maxer MaxL = 1000, Content MaxL = 1000, Frame Avg. MaxL = 400

I attached before and after screenshots ans the HCFR projects. I used 2-pt WB only.

My settings for my "calibration v4" are: low = 0,0,-1; high = -1,-3,0.

I was able to reduce the overall dE value. 40% - 60% have still the largest error. I was not not able to improve these values without increasing dE significant for the others. The overall slop of my color temperature curve is also in principle uncanged. However, after calibration is is a little bit closer to 6500K.

Does someone know if a better results should be principle possible using 2-pt wb only for the LG B7 - perhaps by using more "extreme values" for low and high? Any other comments on my calibration result?

Do you recommend to do 20-pt as well?
To me a dE of 0.4 is totally negligible. Having said that, it also looks like using low = -1,0,-1 should improve the low end.
Dominic Chan is online now  
post #10786 of 12074 Old 06-03-2018, 03:45 PM
Newbie
 
Join Date: Sep 2017
Posts: 11
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 10 Post(s)
Liked: 0
Quote:
Originally Posted by Dominic Chan View Post
To me a dE of 0.4 is totally negligible. Having said that, it also looks like using low = -1,0,-1 should improve the low end.
low = -1,0,-1 was actually my v3 settings. I attached the measurement. I actually went with v4 as the color temperature is closer to 6500K for 10%, while 20% seems a little bit better on v3. What do you about the v3 settings, should I prefer these and not emphasizing to much color temp @ 10?

If I remember correctly I tried also low = -1,0,-2. But I think the dEs were not as good. Unfortunately, I did not saved the measurement.
Attached Thumbnails
Click image for larger version

Name:	HDR 10 - with calibration v3.JPG
Views:	33
Size:	221.4 KB
ID:	2411388  
Attached Files
File Type: zip HDR10 grayscale 10 steps - 2pt WB v3.zip (1.3 KB, 4 views)
maxOLED is offline  
post #10787 of 12074 Old 06-03-2018, 04:48 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 7,107
Mentioned: 132 Post(s)
Tagged: 0 Thread(s)
Quoted: 5000 Post(s)
Liked: 1922
Quote:
Originally Posted by maxOLED View Post
low = -1,0,-1 was actually my v3 settings. I attached the measurement. I actually went with v4 as the color temperature is closer to 6500K for 10%, while 20% seems a little bit better on v3. What do you about the v3 settings, should I prefer these and not emphasizing to much color temp @ 10?
Yes, as expected (-1,0,-1) is slightly better than (0,0,-1) - average dE of 0.41 vs 0.42.
I would look at the RGB graph in addition to dE. At low levels a relatively low dE can still mean a relatively high RGB imbalance, something I was able to see when viewing dim scenes.
Attached Thumbnails
Click image for larger version

Name:	v3_v4.jpg
Views:	42
Size:	297.1 KB
ID:	2411468  

Last edited by Dominic Chan; 06-07-2018 at 09:52 AM.
Dominic Chan is online now  
post #10788 of 12074 Old 06-04-2018, 10:41 PM
Member
 
webdove's Avatar
 
Join Date: Jul 2006
Location: NH
Posts: 59
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 6
Quote:
Originally Posted by zoyd View Post
I've updated the test version link to 3.5.1.6

Changes relative to 3.5.1.5

- Added initial support for Murideo 6G TPG (SDR only). Prefs->Advanced->External Pattern Generator
- Added Dolby Vision support for LG OLEDs when using the HDFury Vertex (grayscale only). Prefs->References->Vertex
Hi Zoyd,

I bought a Vertex and was trying it today. I have an nvidia GTX 1050 talking through it to an LG E6P.

1. I like the fact that it gets controlled by the reference color space settings. That seems logical.

2. I had previously been using GDI HDR10 mode with RGB 8bit full range and setting 16-235 in the generator. When I turned on the vertex I noticed that my blacks were crushed and I had to set the screen to RGB 8bit limited to correct them. It appeared that the Vertex had been programmed to map RGB Full input to 8 bit output if that makes sense (I will take a picture of the Vertex GUI tomorrow). Was that mapping choice on purpose and should I just change to using limited range in windows to work with it?

3. I tried dolby (which was the real reason that I bought the Vertex). The screen blanked out and the dolby vision logo appeared, but I never got a screen image (I have been using 1080P RGB 8bit to keep bandwidth down). Should that have worked? Do I need to do something else to get a visible Dolby image?

Thanks

Web
webdove is offline  
post #10789 of 12074 Old 06-06-2018, 03:31 AM
Newbie
 
Join Date: Jun 2018
Posts: 12
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 0
Hi,

I'm trying to calibrate my VPR with HCFR.

My video chain is entirely set to 16-235. Is it correct that I should choose limited in HCFR settings? I ask you because I have also read in a post that it should be set to Full instead.

I admit I am a little confused

Thank you very much
Silver12 is offline  
post #10790 of 12074 Old 06-06-2018, 06:42 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 7,107
Mentioned: 132 Post(s)
Tagged: 0 Thread(s)
Quoted: 5000 Post(s)
Liked: 1922
Quote:
Originally Posted by Silver12 View Post
My video chain is entirely set to 16-235. Is it correct that I should choose limited in HCFR settings? I ask you because I have also read in a post that it should be set to Full instead.
Simple answer: If your video chain is set to 16-235 then your computer should also output 16-235. However, this happens automatically as the computer's graphics card "sees" the EDID of the TV and will scale its output to 16-236. Thus in HCFR the GDI option should be set to 0-255 to avoid double-scaling.
Answer for the more experienced user: The graphics card usually does not do a very good job in scaling. Thus, for best accuracy, you should have the scaling done by HCFR by setting it to 16-235, but then you will need to make sure the graphics card does not do the scaling. This is a little tricky as you need to override the EDID. You can find the details by Googling the madLevelsTweaker app.

Option 2 is “preferred” but may be harder to set up.

Note: you only need to worry about this when using HCFR's internal patterns. If you're using external patterns (disc, chromecast etc) then the external player will take care of that.

Last edited by Dominic Chan; 06-06-2018 at 10:56 AM.
Dominic Chan is online now  
post #10791 of 12074 Old 06-06-2018, 08:00 AM
Newbie
 
Join Date: Jun 2018
Posts: 12
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 0
So I can set:

Potplayer: limited
MadVR: Limited
NVidia GTX: limited
VPR: Limited

HCFR: Full if I use HCFR Patterns

Is this right?

Thank you
Silver12 is offline  
post #10792 of 12074 Old 06-06-2018, 08:37 AM
Newbie
 
Join Date: Jun 2018
Posts: 12
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 0
This is getting even more complicated. Reading about MadVR, it says that the best option for playing videos on PC is:

Potplayer 16-235
MadVR 16-235
GPU 0-255
Epson 16-235

I don't care about the desktop so I could choose this option. What I don't understand is what I should then use for calibration with HCFR, 16-235 or 0-255?

thank you all
Silver12 is offline  
post #10793 of 12074 Old 06-06-2018, 08:54 AM - Thread Starter
AVS Forum Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 6,804
Mentioned: 150 Post(s)
Tagged: 0 Thread(s)
Quoted: 1783 Post(s)
Liked: 1350
^Recommend for calibration -> 16-235(HCFR):0-255(Video Card):16-235(TV)

Also ok to use 0-255(HCFR):0-255(Video Card):0-255(TV)

Not recommended: 16-235(Video Card)
CFC and ConnecTEDDD like this.
zoyd is offline  
post #10794 of 12074 Old 06-06-2018, 09:27 AM
Newbie
 
Join Date: Jun 2018
Posts: 12
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 0
Thank you a lot
Silver12 is offline  
post #10795 of 12074 Old 06-06-2018, 10:14 AM - Thread Starter
AVS Forum Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 6,804
Mentioned: 150 Post(s)
Tagged: 0 Thread(s)
Quoted: 1783 Post(s)
Liked: 1350
Quote:
Originally Posted by webdove View Post
Hi Zoyd,

I bought a Vertex and was trying it today. I have an nvidia GTX 1050 talking through it to an LG E6P.

1. I like the fact that it gets controlled by the reference color space settings. That seems logical.

2. I had previously been using GDI HDR10 mode with RGB 8bit full range and setting 16-235 in the generator. When I turned on the vertex I noticed that my blacks were crushed and I had to set the screen to RGB 8bit limited to correct them. It appeared that the Vertex had been programmed to map RGB Full input to 8 bit output if that makes sense (I will take a picture of the Vertex GUI tomorrow). Was that mapping choice on purpose and should I just change to using limited range in windows to work with it?

3. I tried dolby (which was the real reason that I bought the Vertex). The screen blanked out and the dolby vision logo appeared, but I never got a screen image (I have been using 1080P RGB 8bit to keep bandwidth down). Should that have worked? Do I need to do something else to get a visible Dolby image?

Thanks

Web
Thanks for testing, are you displaying desktop on the LG? DV mode will only work when the TV is the 2nd extended monitor so only the pattern is displayed. Also, only grayscale patterns will display properly at the moment.
zoyd is offline  
post #10796 of 12074 Old 06-06-2018, 12:58 PM
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,642
Mentioned: 211 Post(s)
Tagged: 1 Thread(s)
Quoted: 3396 Post(s)
Liked: 4218
Quote:
Originally Posted by zoyd View Post
^Recommend for calibration -> 16-235(HCFR):0-255(Video Card):16-235(TV)
+1

Quote:
Originally Posted by Silver12 View Post
I don't care about the desktop so I could choose this option. What I don't understand is what I should then use for calibration with HCFR, 16-235 or 0-255?
...and you will have headroom for all levels above 235, do if you display a contrast pattern (for video levels) it will not clip to 235, but it can show to you all the levels up to 254 (WTW) after making proper adjustment of your display contrast slider. You will be able to check your Brightness also, incorrect adjustment (16 or below has to be invisible) will be easy to spot using a Brightness pattern, when you will try to make visible the 17 bar.

See there why you need headroom: LG 2017 OLED Calibration Thread and Settings

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is offline  
post #10797 of 12074 Old 06-06-2018, 03:14 PM
Member
 
webdove's Avatar
 
Join Date: Jul 2006
Location: NH
Posts: 59
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 6
Quote:
Originally Posted by zoyd View Post
Thanks for testing, are you displaying desktop on the LG? DV mode will only work when the TV is the 2nd extended monitor so only the pattern is displayed. Also, only grayscale patterns will display properly at the moment.
I had two windows setup. HCFR was on the first with the desktop icons and an icon bar at the bottom. The LG was the second window with no icons or icon bar. Is that the correct configuration?
Attached Thumbnails
Click image for larger version

Name:	IMG_0471.JPG
Views:	26
Size:	2.47 MB
ID:	2412812  

Last edited by webdove; 06-06-2018 at 03:29 PM.
webdove is offline  
post #10798 of 12074 Old 06-06-2018, 03:17 PM
Member
 
webdove's Avatar
 
Join Date: Jul 2006
Location: NH
Posts: 59
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 6
Windows 10 GTX 1050 incorrectly reports RGB full mode as limited

Here is a fun fact.
I initially setup my Vertex to “follow” the colorspace and Deep Color of the input.
I set my Nvidia GTX 1050 for RGB Full 8bit.
However, the vertex reports an input signal that is RGB Limited 8bit. Which it transmits unchanged as RGB Limited 8bit to the TV.
If I do a sweep, the first several codes are all 0 luminance as you would expect by sending full 8bit codes to a TV has been told to expect limited 8bit. Thus both the Vertex and the TV are behaving as though the PC is generating limited 8bit when I set the display to full 8bit.
(BTW, it is not calibrated at the moment. I just did a reset. I have HCFR set to generate 65 levels in RGB full starting with codes 3, 8, 12, 16, 20, 24, 28. They should be every 4, but HCFR is probably truncating to integer percent brightness then truncating back to RGB code).
If I set my Nvidia to RGB Limited 8 bit, the Vertex still reports RGB Limited 8bits input and output, but now a sweep looks normal for an uncalibrated condition.
Thus the evidence indicates that the Nvidia when switched to RGB Full is indeed generating RGB full data, but is incorrectly reporting with metadata that the data is RGB limited according to both the Vertex and the LG E6.
Attached Thumbnails
Click image for larger version

Name:	IMG_0460.JPG
Views:	25
Size:	2.34 MB
ID:	2412794   Click image for larger version

Name:	IMG_0461.JPG
Views:	30
Size:	1.33 MB
ID:	2412796   Click image for larger version

Name:	IMG_0467.JPG
Views:	25
Size:	2.35 MB
ID:	2412798   Click image for larger version

Name:	IMG_0468.JPG
Views:	22
Size:	2.23 MB
ID:	2412800   Click image for larger version

Name:	IMG_0470.JPG
Views:	22
Size:	2.22 MB
ID:	2412802  


Last edited by webdove; 06-06-2018 at 03:26 PM.
webdove is offline  
post #10799 of 12074 Old 06-06-2018, 04:55 PM
Newbie
 
Join Date: Jun 2018
Posts: 12
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 0
Quote:
Originally Posted by zoyd View Post
^Recommend for calibration -> 16-235(HCFR):0-255(Video Card):16-235(TV)

Also ok to use 0-255(HCFR):0-255(Video Card):0-255(TV)

Not recommended: 16-235(Video Card)
Just now I figure out that I only have a 1.4 HDMI VPR with its limitations.

So I end with only two options for playing 2160p 24fps videos:
1) GPU: RGB Full 8 Bit
2) GPU: YCbCr Limited 12 Bit

Now I have understood that GPU Limited is not good for calibration but maybe 12 bit is truly important, most for UHD 10 Bit Videos. So, if you were to decide, would you choose Gpu 16-235 12 bit or RGB 0-255 8 Bit? Or you have an even better solution?
Silver12 is offline  
post #10800 of 12074 Old 06-06-2018, 05:20 PM - Thread Starter
AVS Forum Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 6,804
Mentioned: 150 Post(s)
Tagged: 0 Thread(s)
Quoted: 1783 Post(s)
Liked: 1350
Quote:
Originally Posted by webdove View Post
Here is a fun fact.
I initially setup my Vertex to “follow” the colorspace and Deep Color of the input.
I set my Nvidia GTX 1050 for RGB Full 8bit.
However, the vertex reports an input signal that is RGB Limited 8bit. Which it transmits unchanged as RGB Limited 8bit to the TV.
I'll have to fire mine up this weekend and check your screenshots. DV has to be done in full range RGB end-to-end but I don't remember if you can actually change the range on the TV side, if so it needs to be full range. Also, the program will automatically set the Vertex to the right input/output scheme and override however you have it set.



Quote:
Originally Posted by Silver12 View Post
Just now I figure out that I only have a 1.4 HDMI VPR with its limitations.

So I end with only two options for playing 2160p 24fps videos:
1) GPU: RGB Full 8 Bit
2) GPU: YCbCr Limited 12 Bit

Now I have understood that GPU Limited is not good for calibration but maybe 12 bit is truly important, most for UHD 10 Bit Videos. So, if you were to decide, would you choose Gpu 16-235 12 bit or RGB 0-255 8 Bit? Or you have an even better solution?
I'd still choose 1. The higher bit depth for the limited conversion doesn't guarantee it's any better. You can always check the magnitude of the differences and then if they are small, use whichever is more convenient.
zoyd is offline  
Sponsored Links
Advertisement
 
Reply Display Calibration

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off