2019-01-14 at 21:08 #15234
You have been so helpful to us in the past, so we did contribute a little something. I will contribute more in the future to DisplayCal this year, I promise.
We have a new 4K editing system using Blackmagic-Design’s 4K Extreme card with SDI output and a BMD Teranex SDI to HDMI box which converts SDI to HDMI 2.0. This interface allows you to load two 33 point 3D LUT profiles for greater accuracy with color, manipulation and supposedly can be used to accurately color calibrate consumer televisions so they can be used for critical color correction work. It has buttons on the front to switch between the two 3D LUT profiles or to entirely bypass them. This is very straight forward, but we are using the bypassed mode right now because we don’t have a good 3D LUT to load yet.
We are using Davinci Resolve 15 to calibrate a LG OLED 55″ TV (model# LG OLED55B8PUA). Here is a link to a Dropbox download of the technical manual for this TV. It was too large to include in your system’s attachment. https://www.dropbox.com/s/6drd0bpxq2xzrly/LG%20OLED55B8PUA_Manual.zip?dl=0 We are using an X-Rite i1 Display Pro as our calibration instrument.
I have included a ZIP file containing screen shots of our DisplayCAL 22.214.171.124 setup, the calibration and Profiling Complete Dialog, a spreadsheet containing the current TV settings we are using, the RGB and Gray Balance report output from DisplayCAL, and a compressed archive of our settings in DisplayCAL. I didn’t know exactly what you would need to help us, so I gathered up as much info as I could provide.
We are not getting good calibration with our current settings. We have tried DisplayCAL’s Full Field pattern insertion in advanced settings as well because we are using an OLED TV. We would be happy to pay you for your time to optimize our display for Davinci Resolve and understand what we are doing wrong. We are wanting to color grade in Davinci Resolve using this monitor.
i1Display Pro on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.2019-01-15 at 15:34 #15243
judging from the results and TV settings, I’d recommend the following:
On the TV:
- Disable dynamic contrast (off)
- Set the TV to its native gamut mode (whichever that is, probably wide? so should be good already from your settings)
- OLED light around 19 should give you 100 cd/m2
- Set gamma to 2.4
- Can logo luminance adjustment be turned off?
- Disable ASBL in the service menu:
- Due to OLEDs temperature stability problems, I’d reduce the number of profiling patches
- Set profiling patch sequence to “maximize lightness difference”
- Use the smallest possible patch size
It’s been suggested by others to play some content on the TV during the warm-up period before starting measurements.2019-01-15 at 18:06 #15249
I’ve just posted a question about primaries advertised in the EDID vs actual primaries. LG OLEDs advertise only sRGB primaries but must have a much wider gamut.
I wonder if this skews the results.2019-01-15 at 18:14 #15250
EDID has nothing to do with this. No thread hi-jacking please.2019-01-15 at 18:43 #15252
Just trying to help.2019-01-17 at 22:36 #15290
Good news your expertise improved things dramatically, but was wondering if I can even make things better. I followed your instructions pretty close:
On the TV:
– Set dynamic contrast to “Off”
– Set the TV gamut to “Wide” (native gamut mode is auto)
– Set OLED light to “26” to achieve 100 cd/m2 in DisplayCAL (19 gave me 80 cd/m2)
– Set gamma to “2.4”
– Set logo luminance adjustment to “Off”
– Set ASBL to “Off” in the TVs Service Menu using instruction from the link you provided: https://www.avsforum.com/forum/40-oled-technology-flat-panels-general/2440714-how-turn-off-asbl-lg-oled-tv.html I ended up purchasing a Harmony Remote 350 and was able to access the Service menu using those instructions.
– Reduced the number of profiling patches to the default 1553
– Set profiling patch sequence to “maximize lightness difference”
– As recommended I played some content on the TV for 30 minutes prior to starting measurements.
I have enclosed the results in a zip file called AVCOM Cal_2018-01-16.zip. The results from this calibration were much better than anything that I tried before.
79.6% Adobe RGB
88.9% DCI P3
91.2% Adobe RGB
93.8 DCI P3
But, I looked at the content from a video project that I knew was good. The picture looks way too dark and lacking pop. Also, the darker areas were crushed to black. I put up a 10 grey bar scale the black and next lightest black bar were indistiguishable. So, I made a few changes and WOW I think the results are much better. Here is what I changed on the TV
1. Set OLED light to “50” to achieve 133 cd/m2 in DisplayCAL
2. Changed Contrast from “85” to “90”
3. Changed Black Level from “Low” to “High”
I have enclosed the results in a zip file called AVCOM Cal_2018-01-17.zip.
Here is a summary from these new settings:
84.0% Adobe RGB
94.5% DCI P3
100.1% Adobe RGB
102.9 DCI P3
What can I do to make the results even better? Thanks in advance. I have learned alot!
Bjarne2019-01-20 at 2:05 #15319
Make sure the output black level from the DeckLink matches what the TV expects as input (black level). It seems the 2019-01-17 result has raised blacks because Resolve sends 16..235 through the DeckLink but the TV expects 0..255 (thus the reduced contrast of 1681:1).
Re the black crush issue, use a pluge pattern before calibration to set correct black brightness so that no levels are clipped (might also help to set the TV gamma back to 2.2, the 3D LUT will ensure the correct target gamma regardless).
Some discussion over at AVSforum in the LG OLED thread seems to suggest that contrast should be set to 100, thus leaving OLED light as the only control to adjust peak luminance.2019-01-24 at 0:32 #15438
I checked to make sure that the Black Magic Decklink is outputting 0.255. At least I think so. I am running in 4:4:4:4 environment with no SDI to HDMI. clipping. I set the TV to 2.2 gamma and set the contrast to 100%. I tried adjusting the OLED back light and brightness to set black levels using a black level chart. I then ran calibration again the contrast level seems better now, but I have some errors. I would appreciate it if you could give me a few more suggestions.
B2019-01-24 at 15:35 #15449
Display stability is still an issue. Levels 88% and up measured the same luminance as white, meaning anything above was dimmed down by the TV. Do you have to use a 200cd/m2 target? It could be that the TV might not be stable at that luminance level. You can also try to enable black background on the measurement window instead of using APL patterns and see if that helps. It also looks like you’re not using the smallest possible patch size (it is generally a good idea to make the patch area as small as possible).2019-01-24 at 20:44 #15459
I wish I understood more basics about calibration, I thought that a full screen area was best and didn’t even understand that this area is called a Patch. Now I do. I also selected the Black background option. I think that fixed a great deal. In doing so, I had to change the RGB white balance color settings on the TV, which makes more sense because now those values are not off from center like before. Please let me know if my results are good and if there is anything that I can do to make them even better.
As I said before, I am looking to learn how to color grade in both the Rec. 709 and 2020 color spaces. From my understanding, you have helped me setup my LG OLED monitor in the Rec 709 color space. What would be the next steps to configure the LG OLED monitor to work in the Rec. 2020 color space?
Again, thank you for your help and patience.
Bjarne2019-01-24 at 20:54 #15462
Alright, this profile looks good. You can re-use it for the BT.2020 3D LUT. On the 3D LUT tab, disable the “create 3D LUT after profiling” checkbox. Then, change source colorspace to BT.2020. You can now create the LUT without taking new display readings.2019-06-21 at 5:15 #18237
Marcelo Cosme Carlos dos SantosParticipant
Alright, this profile looks good. You can re-use it for the BT.2020 3D LUT. On the 3D LUT tab, disable the “create 3D LUT after profiling” checkbox. Then, change source colorspace to BT.2020. You can now create the LUT without taking new display readings.
Which tone curve?2019-06-21 at 11:57 #18241
SMPTE 2084 roll-off. Recommend set peak luminance to > 300 cd/m2 if display peak is lower.2019-06-29 at 15:27 #18428
I also try to calibrate with DiplayCal my LG Oled TV thru Resolve I’m on Mac with the Spectracal C6 and the output to monitor is the UltraStudio Mini Monitor. I’m not able to have the square patch target on oled monitor it stays on my primary monitor, the IP address is ok but I must set the target full screen to get the patches on the TV can you tell me why. Thank you so much in advance2019-06-29 at 16:58 #18430
I’m not able to have the square patch target on oled monitor it stays on my primary monitor,
Known Resolve or DeckLink bug.