Unity Bypass, LUT Uploads and LUT Generation Size - new information added

A recent thread here on the LG G5 ( https://lightillusion.com/forums/display-calibration-8/lg-g5-confused-about-hdr-3d-lut-calibration-1585.html ) raised good questions about what actually happens inside the hardware when a LUT is uploaded, and what a Unity Bypass upload resets.

That made us realise this information wasn't documented clearly enough, so we went through every LUT-capable hardware integration in ColourSpace in detail, and have added more specific information to each hardware page of the website ( https://lightillusion.com/guides_hardware.html ).

In general terms:

A Unity Bypass upload has no source LUT. It uses the hardware's own commands to return the selected LUT, or LUTs, to null. On some hardware other settings are also reset to provide a valid calibration starting point, such as disabling internal gamut matrices, which provide no benefit with 3D LUT calibration and can compromise accuracy.
A real LUT upload always starts from the ColourSpace 3D LUT.
Depending on the Load option, and the hardware's internal structure, the 3D LUT may be uploaded as-is, or split into 1D and 3D components. When ColourSpace performs the split, the LUT order within the hardware is accounted for automatically.
Hardware without a combined 1D & 3D Load option is not necessarily a single 3D LUT. Some hardware always receives a 1D & 3D split.

We have also updated the LUT Generation Size page (https://lightillusion.com/lut_size.html).
It now explains why a large LUT Generation Size improves accuracy for all hardware, whatever the hardware LUT size or structure, and especially where a 1D is extracted from the 3D LUT.

Thanks to everyone who asked the questions that led to this.
If anything is still unclear for your specific hardware, please ask here.

Steve