Hi!

I'm calibrating an ASUS PA32UCDM (QD-OLED) and an LG C2 (WRGB OLED) using ColourSpace + i1D3 with DaVinci Resolve as TPG.

Current signal chain:

Resolve
↓
DeckLink

HDMI OUT
→ ASUS PA32UCDM (Reference)

12G-SDI OUT
→ Blackmagic Micro Converter SDI→HDMI 12G
│ → LG OLED C2 (Client)

Both displays are calibrated to internal hardware 3D LUTs.
I'm trying to avoid unnecessary full 17³ re-profiles due to profiling time and OLED wear, so I'd appreciate guidance on best practice.

1. Black level before profiling
For OLEDs (LG WRGB OLED, Samsung/Sony QD-OLED, and professional QD-OLEDs like the ProArt), and profiling it with the limitations on black measurements from i1d3, is it generally correct to leave the brightness / black level control at native default before profiling?

My reasoning is that these panels already achieve true black natively and raising black level introduces an artificial floor
The LUT would then encode compensation for an adjustment that was not actually needed.

So in practice, should black level simply remain at default unless there is a verified clipping or near-black issue?
Does the answer differ between consumer OLEDs (LG C2) and a professional QD-OLED monitor (PA32UCDM)?

2. Peak white strategy before profiling
Since a 3D LUT often reduces luminance while correcting native white to D65, some headroom above the final 100 nit SDR target seems necessary before profiling.

I see three possible approaches:
Method A
Set display luminance about 10–15% above target
Leave RGB Gain controls at default
Use max luminance cap in ColourSpace LUT generation to constrain output to exactly 100 nits

Method B
Pre-calibrate with OSD RGB Gain to D65 and 100 nits
Reset RGB Gain to defaults before profiling
Profile the panel in its native state and let the LUT handle the full correction

Method C
Pre-calibrate with OSD RGB Gain to D65 and 100 nits
Leave those controls active during profiling
Let the LUT correct only the remaining residual error

Which method is preferred for grading displays?

Would the recommendation differ between:
ASUS PA32UCDM (QD-OLED)
LG C2 (WRGB OLED)

3. ColourSpace settings for grading monitors
Beyond Disable Gamut Mapping, any other settings commonly recommended for reference/grading monitor workflows that im not aware off?

4. Metameric failure - perceptual match without a spectrophotometer
The perceptual match guide states that probe matching against a spectrophotometer is imperative as part of this workflow. Without a spectro, the only available probe correction is the device-specific i1D3 correlation file for each panel.

Does the correlation file satisfy the probe matching requirement in this context - i.e. is a device-specific correlation file a valid substitute for spectro-based probe matching when performing the perceptual match procedure on QD-OLED and WRGB OLED, or does the absence of a spectro introduce an uncertainty that the correlation file cannot compensate for?

5) Current profiling settings
LG C2
SDR: 10% window
HDR: 3% window
Full black stabilization: 0.35s

ASUS PA32UCDM
SDR/HDR: 3% window
Full black stabilization: 1.0s

Regarding the subsequent 10^3 verification profile:
Must the stabilization and patch settings be identical to the profiling settings to accurately verify the LUT, or can stabilization be disabled?

If a 10^3 verification is run without stabilization to observe panel variance, is there a legitimate risk of burn-in or severe temporary image retention on either the WRGB or QD-OLED panel?

How do you reconcile the contradiction between the "calibration state" and "real-world use"? While utilizing stabilization yields the lowest average dE2000 during verification, this artificial, thermally cooled state does not reflect real video content. Is it standard practice to perform a secondary, shorter verification (like a grayscale sweep) without stabilization to better represent real-world thermal drift?

Any feedback on whether these patch sizes and stabilization times for both profiling and verification, look reasonable would also be appreciated.

It's a better idea to use REC.709 for the ASUS, and after the calibration, use perceptual matching to find a custom white point coordinate that will match visually on your LG the calibrated white of your ASUS, then use the modified REC.709 with custom xy for white colorspace to calibrate the LG TV.

For the calibration, it's better to use the VIDEO signal from the calibration; ensure that you have ticked the SDI 4:4:4 in Resolve for both devices to receive the RGB signal.

ASUS input as video and LG input as Limited/Low (not AUTO)

For the ASUS use FULL patchscale, and Resolve output VIDEO.

For LG, extended for pre-cal/profile with Resolve output DATA and LEGAL in CS (Resolve DATA).

After the verification of both, set Resolve output to VIDEO.

For the pre-cal of ASUS, use the RGB controls to pre-cal the 100% White for 100 nits to set the 'backlight' setting, and after finding the backlight value, reset the RGB controls and proceed with 33p 1D post, (1D post with Augment if required to improve further the 1D post) and then the 17p 3D LUT and upload into 1D pre + 3D.

For the LG, pre-cal with SM WB the 109% White for 123.4 nits (for getting 100 nits to your 100% White later), and then follow the processing using the LG Guides.

For near black, you should check and make the required adjustments using pattern for video levels with flashing bars (Ted's Media Files into your timeline) after the 100/109% precal, adjust if required without raising your native black level.

The stabilization patches are useful during the full profile, not for 1D LUT or the verification.

You can verify with primary and secondary if you want a quick evaluation instead using 10-Point verification.

For SDR use 10% area of windows for both.

Load/activate a UNITY 3D LUT into the Biconvertor before taking measurements.

The correct approach is explained in some detail in the 3D LUT Calibration User Guide - Peak White Nits Level & Colour Temperature.

Steve