AVS Forum banner

LG 2017 OLED Calibration Thread and Settings

29 reading
1.1M views 3.4K replies 355 participants last post by  patoberli  
#1 · (Edited)
LG 2017 OLED Calibration and Settings

I could not find a LG 2017 calibration thread so here is one.
Since all 2017 LG's use the same chipsets, one thread should do it.

My 65 C7 came yesterday and looked pretty good out of the box.
Without burn-in, I know I will be doing it again but it's good practice.

Gear: I1Ddisplay 3, I1Pro2, Lumagen 2020 (for patterns), and Calman 2017 beta.

Display Setup


- All enhancements off.
- C7 Gamma BT.1886
- Set the black level on a full back screen
- WB Warm 2

Calman Setup:

- Selected a profile on hand
- Gamma Target BT.1886
- Black Level set to .0034 (from @chadb B6 recommendations)
- Adjust 2 point High with 100% white, no adjustment needed to Low.
- Run a sweep and offline adjust luminance only
- Run a sweep and adjust luminance where all are even and RGB using 2-clicks from each 1% off.
- Repeat sweep a few times with minor adjustments offline
- Live TV content is displayed between sweeps

The results looked very nice.
However, overnight I ran Clear panel noise and things changed a bit.

I did not capture the calibration but as I recall the DE Mas was under 1%.

Today it is a bit different.
This is the profile used last night although I;ll be doing that again as well.

Next time, I will perform a Panel Clear and play content for a while before calibrating.

Attached are the D3 readings this morning with the

- I1D3 in RAW mode no profile
- I1D3 with an older profile (used last night)
- I1D3 Oled no profile

These were taken one after another and they definately vary.
Withtout a spectro to profile, I would be inclinded to choose RAW for my I1D3.

I'll do another run with a new profile against the I1Pro2 soon.

- Rich
 

Attachments

  • Like
Reactions: LenH
#4 ·
Hi, you have to wait about 1 hour for the display electronics to warm up before doing any measurement, gamma is changing a lot until about 50 minute where it will be stabilize.
 
#11 ·
You absolutely do not need to do this on the 2017 sets. It actually makes near black worse.:eek:

- Black Level set to .0034 (from @chadb B6 recommendations)


Yes, I inherited that from the B6 workflow. I'll redo it in a few days.
I put on Cinema mode with BT.1186 and set the same settings. This TV is fantastic OOTB.


- Rich
 
  • Like
Reactions: jrref
#10 ·
As has already been stated, TV powered on time (aka warm up) and room temperature will impact readings. I've measured repeatable results with my OLEDs days, weeks and months apart when taking into account the above. If you are seeing different.... are you taking the above into account? If so, then it is your meter.
 
#13 ·
You really need to scan and read this thread -> http://www.avsforum.com/forum/139-d...display-calibration/1423111-calman-5-release-notes-discussion.html#post22276312

All your answers are there.

Bottom line, the OLEDs are less stable than LCDs for many reasons. ABL, ASBL, etc.. which all effect your calibration. If you are looking to calibrate the set then go back the next day and get the same exact readings, you would have to warm up all your equipment the same amount of time, make sure your room is the same exact temperature, use the same profile for the meter, and even then the panel may shift slightly, so it's not "realistic". You are "always" going to see some slight drift.

You should just do all of the things that everyone has recommended above and then calibrate in one session and that's it, then go back periodically and re-calibrate if you want. This is not an "exact" science.
 
#15 ·
Are using the JUDD OLED white or D65?
 
#19 ·
JUDD OLED white doesn't work well on these WRGB type OLED (both LG and Sony are trying the mimic this with their default Pro modes). The best way is perceptual matching until the industry figures out how to resolve metameric failure with narrow spectral band displays. If you can't do a perceptual match, just use D65 as the white point.
 
#16 · (Edited)
oh, i posted this on another thread, but here is a calibration i did for gaming.

Gaming mode Calibration:
LG 65" E7 (but all the 2017 screens are the same)

Do note, that all TVs are a little different so this might look different on your TV than mine. But when going into PC mode and using one of the ISF presets the colors are off (still much better than using the gaming preset)
Here is what I got after a few hours of work: (There is still a slight blue push) But it should be better than the default.

First, make sure your input is named PC and the icon is set to the PC.
second, make sure Ultra Deep colors are on. (it is General settings on the TV, there is a toggle for each input)
Third, for non-HDR content use 8bit, RGB, Full

This was calibrated using Calman 5 and a SpectraCal C6-HDR Colorimeter. 10% window.

Settings all play off each other, just taking some isn't very helpful.

Input named PC
ISF Expert (Dark Room)
OLED Light: 49
Contrast: 84
Bright: 52
Color: 55
Expert -
Color Gamut: Wide (Auto can we wacky so better for force it to something)
Gama: 2.4

White Balance-
Color temperature: Warm3
Method: 2 point (i'm lazy)
Pattern: Outer
Low-
Red: 2
Green: 2
Blue: 2
High-
Red: -10
Green: 0
Blue: 31

Color Management System (CMS)
(Color, Saturation, Tint, Luminance)
RED: -12, 15, -3
Green: -20, -30, -1
Blue: -5, 11, -2
Cyan: -30, 4, 1
 
#22 ·
Bought a new 65B7 this week. My last TV was a 55EC930V.

I have to say that the B7 is a much better TV. But I can see a more pronounced banding on this set than what I had in the EC930. I hope it improves with time.

So far what I found more interesting was the black tuning control. The B7 allows for a much more fine brightness tuning than the EC930. Yesterday, I had to decrease the brightness to 51 in order to really have 100% blacks. Level 52 seemed black until I stuck my nose in front of the panel and saw a very, very dim light.

I'm letting my panel to have some hours of use before I try any calibration because it really is changing each day. I started with a brightness of 54 for instance, now is 51. Colors are different too.

Does anyone know how I can check the number of hours of use and the number of panel recalibrations?
 
#24 ·
You don't need to adjust the brightness on these 2017 sets like you did on the 2016 sets. I've "played" with it and you kill the black when you raise the brightness. The only thing you can do if you want is raise it 1 click to 51 but not more than that or you will not get true blacks when watching good content.
 
#23 ·
  • Like
Reactions: raaj
#30 ·
It depends on how far off it is. Some of the review show quite a bit of drift in the C7 grayscale. Once that is fixed, the colors are very close. Much better then the B6.
I can calibrate SDR but my meters don't stack up against the pro's. I have not had the opportunity to calibrate HDR. The tone-mapping is a bit more aggressive than I would like, without Dynamic Contrast=Low, there are many films that appear dark in the mid-tones.

I'd like the option to chose tone-mapping approaches on the C7. Dynamic Contrast=Low produces a big change but thus far, this is my choice.

Back to your original question, @ChadB ;calibrated my B6 and had the experience to correct many of its issues that I simply did not know nor have the experience to fix. He also calibrated HDR. You are also paying for this experience. Once completed, I copied his SDR settings to Cinema mode and used my meters to get the DE vanishing low. My meters are not as accurate as a pro's, but in the end, I could switch between the two settings and could not tell the difference.

- Rich
 
  • Like
Reactions: bmcn
#36 ·
It's got a couple of issues so as D-Nice says, it's close enough so best to just leave it at default and concentrate in getting the gray scale perfect. :)
 
#40 · (Edited)
Here are some measurements taken of the nit levels using the Sony HDR 10 10000 nit patterns (available from the menu via 7669).

Three modes were used with the default Contrast, OLED Light, and Color settings. All enhancements Off.

The following modes were measured using the HDR-10 patterns. The pattern with the 50 and 100 nit strips also has a gradient bar which raises the maximum nit value shown which may affect the Active HDR processing (Dynamic Contrast = Low).

These modes were measured:

  • Standard (User) with Dynamic Contrast: Off
  • Cinema Home (User) with Dynamic Contrast: Off
  • Cinema (User) with Dynamic Contrast: Low (Active HDR)

The strips measurements are:

  • 50 nits on 0 to 100 pattern
  • 100 nits on 0 to 100 pattern
  • 1000 nits (on 100 to 1000 pattern)
  • 2000 nits (on 1000 to 2000 pattern)
  • 10000 nits (on 2000 to 10000 nit pattern)

ABL clearly is affecting the 1000 nits and above patterns.

The 50 nit strip measurements:

  • Standard DC: Off = 31 nits
  • Cinema Home DC: Off = 17 nits
  • Cinema Home DC: Low = 8 nits

The 100 nit strip measurements:

  • Standard DC: Off = 82 nits
  • Cinema Home DC: Off = 34 nits
  • Cinema Home DC: Low = 17 nits

What is clear is the Cinema mode tone-mapping on HDR content mastered at 10000 nits dramatically affects the average picture level (100 nits and below).

In content, Cinema DC: Low can make this worse when there is high-not content on the screen. Therefore, Active HDR/Dynamic Contrast: Low brightens scenes (by observation) but when bright content is present darkens 100 nit and below.

There is no Cinema mode that preserves the brightness of 100 nit content and below.
Since only Cinema has white balance controls, there is no mode that tracks EOTF below 100 nits than can be calibrated in the user menu.

Sony's claims to better dynamic tone-mapping may be well founded. Hopefully, LG will add tone-mapping options to Cinema mode that do a better job tracking the EOTF.

- Rich
 

Attachments

#43 ·
What is clear is the Cinema mode tone-mapping on HDR content mastered at 10000 nits dramatically affects the average picture level (100 nits and below).
Hi, they haven't released movies mastered for 10000nits, so it's not so important to know how LG tone mapping affected with 10000nit content.

You need to measure with 10% Window HDR patterns (1000/1100/4000nit) to have better idea because Contrast pattern has large APL so it will enable ABL which will affect the panel output.
 
#41 ·
Robert at VE let me work on one of the C7s in the showroom today to try the Calman DV calibration. You need to have firmware 3.60.09 which is only available by the sets update feature since it's not on the LG web site. Overall, the way I understand everything, when you run the Calman DV Custom workflow it puts the TV's Dolby Vision engine into pass-though mode via special Dolby Vision "Relative Mode" metadata as per the instructions. That allows access to the 2 and 20 pt gray scale controls. When I ran the 20 pt after the 2 pt., the lower IRE's were targeting close to 2.4 gamma up to about 35-40% then the curve settled down closer to the 2.2 target. I'm not sure if this is the way it's supposed to be since I was expecting everything to be very close to 2.2 and only have to make a couple of adjustments which was not the case. Because you are dealing with very high luminance levels, it is very tedious to get the 20 pt. flat. Once I got the 20 pt reasonable, and everything as close to D65 as possible with gamma 2.2 as per the workflow instructions, I ran the program in the workflow that creates the "special file" and copies it to a thumb drive. It essentially runs white and all the CMS colors, similar to when you profile your meter with a couple of additional measurements. I was then able to connect the thumb drive to the LG, run DV content, and a message asked me if I wanted to update the DV settings which I did and it loaded just fine. I then did the final step of running the last DV HDR layout in the workflow to verify the work and the grayscale was pretty flat and it tracked fairly well to the 2084 EOTF curve just like you would see when calibrating in HDR10 mode. Now all this said I needed to see how good or bad the calibration was with real content so I ran the same Netflix content on another un-calibrated C7 in the showroom and the set I calibrated, which were fairly close together, with the same firmware, both in DV Cinema mode, all the user settings the same, at default, except for the 2 and 20 pt settings on the calibrated set. The result was that I immediately saw that the skin tones on the DV calibrated set were more accurate, the overall "balance" of the picture was better and i could easily see more shadow details as well. I didn't have a lot of time to try and evaluate specular highlights, etc, but you could easily see that the DV calibration seemed to improve the PQ overall. As I stated, I was expecting the 20 pt to be pretty close the way it is when you calibrate HDR10 but on the set I worked on, it wasn't and it takes a while to do the full 20 pt at these high luminance levels. Overall, interesting process and it works but I wished it was less work and quicker to execute. :)
 
#45 ·
Robert at VE let me work on one of the C7s in the showroom today to try the Calman DV calibration. :)

By any chance did you try to revert back to LG's default settings for DV.??

Great to read that LG/Calman have a good way to calibrate DV.:)
Thanks for posting and thanks to Robert for his part in this.

ss
 
#42 ·
Has anyone watched Edge of Tomorrow on Bluray or TNT on these OLED's? I'm loving my E7, but the dark scenes in this movie are totally unbearable. You can't see anything that is going on towards the last parts of the movie on standard or cinema. Will calibration correct this, or is this just a bad content/OLED combo?
 
#48 ·
I love this movie but I agree it was unwatchable on TNT. I just chalked it up to it being on FIOS. It looked terrible in the dark scenes.
 
#49 ·
metameric failure of displays



Hi, using HCFR you can do SDR and HDR10 calibration, for Dolby Vision you will need CalMAN Enthousiast.

A professional will save you a lot of time until to find out how to prevent wrong calibration adjustments decisions which will appear (in HDR10 calibration for sure), he will use better meters or approaches (to prevent metameric failure of displays), so you will have more free time to spoend for watching movies than from searching forums from solutions.
 
#50 ·
I got my C7 a few days back...what settings are people using for HDR Picture Mode?...I see articles online listing SDR calibration settings but nothing for HDR (except to use Cinema Mode)...can't be that I'm supposed to use the same settings for Contrast, OLED light, sharpness etc that I use in SDR right?...or do I leave all the HDR settings at their default (except changing it to Cinema mode)?
 
#51 ·
I ran a quick Calibration,, using a Jeti 1211 as reference meter and a Klein K10-A as the read meter
Settings.
ISF Dark Room
Oled Light 42
Contrast 78
Brightness 51
H/V Sharpness 0
Color 50
Color Gamut-Auto
Gamma 2.4
Color Temp-Warm 2
Point High-Red +7, Green 0, Blue +5
Point Low-Red 0, Green 0, Blue 0
All enhancements Off.

As you will see from the Screen shots, all readings are under dE2000 of 3. That's good, considered to be with in excepted tolerance for a ISF calibration. The changes made to the default Gray-scale are very minimal and the Gamut there are no changes from default settings, that is how close the LG OLED 65C7 is to excepted tolerance for a ISF calibration. Can I improve on these readings and have better PQ by doing a 3DLUT calibration, of-course I can. But non the less this LG OLED C7 comes from the factory close to excepted tolerances. :)

ss
 

Attachments

  • Like
Reactions: bigapp
#52 ·
Thanks for doing this. Even though I own Calman and a meter I still have Chad B headed this way in Sept to cal mine. I'm getting mine done by a pro because I know I can only get so far with the limited experience I have. So I should be able to get close and Chad will bring it all the way:D
 
#64 ·
I haven't had time to do much on my C7, so I will default to D-Nice and Chad.

I see a disconnect between measured color gamut and color checker skin tone patch results and how the 7 series looks with real content. With color gamut at auto, bumping color up and tint toward red helps skin tones look more natural. This is the opposite of the 6 series.
When using Wide color Gamut (not auto for 3DLUT) and running a large profile/3DLUT, I am seeing the apposite adjustment with Red for real life skin tones. Yes I agree about the "disconnect", and imo its not just with the ColorChecker but with Lum and Sat sweeps as well.
Also it seems that Calman and the Lumagen Pro 444x, don't play well together for the time being. That is what I thought was the issue until I saw your post.

Last night I used LightSpace for my pre settings and of-course my profile, but I haven't had a chance to check the outcome.

The biggest issue I am seeing that is caused by the C7 is white point drift. I have less than 200 hours on my C7 so I am hoping that the C7 will become more stable.

ss
 
#55 ·
Technical info:

Break in:
The C7 was broken in for a total of 200 hours, with multiple power and compensation cycles throughout. Cinema mode was chosen to track the C7’s break in since it is one of the more accurately measuring modes, while the ISF Expert modes would be reserved for calibration and other testing. Fresh out of the box with just a normal warm up time, after turning the power save feature off the C7’s Cinema mode measured extremely well. The previous 6 series’ familiar greenish emphasis in the mid tones was significantly reduced, and color tracking at different brightness levels and saturation points was excellent and greatly improved over the 6 series. Even the skin tone simulations measured with the correct tint and saturation, a feat that eluded the 6 series. Initial picture impressions were somewhat surprising; unlike the previous 6 series OLEDs which tended to have overly rich color and dark shadow detail, the C7’s color was polite, with slightly pale skin tones, and most content did not seem to be lacking in dark textures.

After 97 hours, the C7 was remeasured under identical conditions and settings. As the break in process progressed, the gamma shifted lower which tends to slightly reduce the depth perspective and wash out the image. There was also a tiny bit more green in the white balance.

Finally, at 200 hours the C7 was measured again, with the same settings and under the same conditions. Between 97 and 200 hours, the gamma lowered only a tiny bit more, but the green emphasis in the white balance increased to the point where it would become more visibly intrusive.

The resulting pre calibration picture was good, but watching Blu Ray movies within the initial 150 hour period showed a mixture of both enjoyable and annoying characteristics. First, the motion had the unnatural Soap Opera Effect along with fleeting jagged edges and general weirdness around the borders of moving objects. Blacks looked superb, but depth and dimensionality of the image was not as great as hoped; and whites often seemed compressed. Dark picture content just above black was easy to see and very well delineated. Skin tones, while generally natural and inoffensive, sometimes had a jaundice tinge. The image had excellent pop and a great photo realistic quality.

Windows and fields:
Some other surprising departures from previous LG OLED models included differences with various size and style measurement windows. Previously, there had been no significant change in the way different standard window sizes and styles measured. With the C7, shadow detail measured a tiny bit stronger with small windows than with medium to large windows, and significantly stronger than with APL style windows. This is shown as an increase in gamma at the low end with larger and APL windows. Measurements attached for 2, 10, and 25% size conventional windows and 10% size/25% APL windows.

Thankfully, the white balance was very consistent with all sizes even up to full fields. Peak light output when calibrated to 50 fL was basically unchanged with 50% windows and an extremely impressive 43.1 fL with full fields. The dynamic brightness pattern on the AVS 709 disc looked similar to past OLEDs with only a slight darkening of 10-30% as the APL changed. These results suggest that, while shadow detail may now suffer from small dynamic shifts, ABL (undesirable brightness limiting with full white screens) is pretty much a thing of the past.

Color:
A second departure from previous models concerns color reproduction and behavior of the color gamut selection and color filter functions. The C7’s color reproduction seemed too relaxed at times, lacking in richness, despite measurements suggesting nearly perfect color accuracy with color gamut at the default Auto setting. While measurements are generally the preferred way of setting and testing color, LG provides a color filter function to aid in color calibration in the absence of a suitable meter. Surprisingly, the C7’s color had to be increased to 72 with a tint of G2 for the result to be correct with the color filter function, with the image looking surprisingly decent but at odds with the measurements. Changing the color gamut selection to wide “fixed” that problem, indicating perfect color alignment at default color and tint settings with the color filter. Reference screencaps looked impressive in Wide, but the measurements showed colors were oversaturated. The same experiment was done on the 6 series, which did not show the same color gamut differences with the color filter function. The 6’s color gamut selection behaves as expected, with colors ranging from merely rich in Normal color gamut to over the top in Wide.

As with the C, E, and G6, the C7’s CMS behavior is somewhat less than desirable. While an improvement over the CMS behavior of the B6 and earlier OLEDs, the C7’s CMS tends to have sharp adjustment transitions which can lead to artifacts on some content. In addition, in between color shades such as skin tones tend not to be touched at all by the C7’s CMS. CMS behavior was tested with color gamut set to both Wide and Auto with slightly modified but ultimately similar results. It is therefore advisable to keep CMS calibration to a minimum.

HDR:
HDR was significantly improved over the 6 series, in more ways than just higher peak light output. The grayscale and gamma can now be calibrated nicely over the entire range, though a modified “code value” adjustment remains as the multipoint control. By running very closely spaced measurement runs, it was possible to determine exactly which brightness level the different code values adjusted. The lower three and uppermost three controls had a wide coverage area (like Q factor in the audio world), while the middle 14 controls adjusted a fairly narrow range. The stimulus level each code value adjusted, with contrast and brightness at default, will be listed in an attached document. I was able to use the levels editor of CalMAN to make a special set of measurements that lined up with these levels, which facilitated calibration.

Uniformity and misc:
There was an inconsistent change in light output when switching the screen shift function off in the OLED panel settings menu. The screen shift caused resolution patterns to garble slightly as the screen shifted. Turning it off caused a puzzling loss of peak light output, going from 50 fL to 42 fL, though cycling the picture modes or power seemed to bring the light output back up.

After break in, dark uniformity was the best I’ve seen on an OLED. Full fields were examined coming out of black in 1% increments. There was no vignetting (though I believe I caught glimpses of it earlier in the break in process), and banding was only slightly visible on a 3% field, though it did not look like it would be visible at all on any real image. With bright 50%-100% fields, uniformity was good, though the right side looked slightly warmer than the left. Measurements of the C7’s near black behavior showed excellent shadow detail with gamma near ideal from .5% to 5%. This is dramatically better than the 6 series, and proof of the C7’s superior shadow detail and near black performance.

Input lag with any picture mode except game and any combination of motion settings was 88 ms. Game mode measured an impressive 21 ms. Relabeling the input to PC was not tested.

Even with sharpness set to 0 and all enhancements switched off, there was some edge enhancement on sharpness patterns with 1080P or lower resolutions. 4K/UHD sharpness and resolution patterns looked superb, with full resolution and no artifacts.
Some later thoughts... I haven't seen a need for an alternative WP on these. D65 on them looks like D65 to me. Interestingly, the sets I have thought needed an alternative WP most were some Vizio and LG edge lit IPS LED LCDs.

IMO, small windows correlate best with the x7's subjective look in regards to shadow detail.

I see a disconnect between measured color gamut and color checker skin tone patch results and how the 7 series looks with real content. With color gamut at auto, bumping color up and tint toward red helps skin tones look more natural. This is the opposite of the 6 series.

Externally upscaling lower resolution sources to 4K can eliminate the slight edge enhancement and exaggeration of graininess I noted in the review.
 
#57 ·
Some later thoughts... I haven't seen a need for an alternative WP on these.
Hi Chad,

Have you had the opportunity to place an OLED right beside a plasma where both were calibrated to x3127 y3290 and observe the color of the x3127 y3290 white light?