HDR calibration for digital Mike take 2

Hi all,

Here is a take 2 of the lighting validation where I assemble the HDR shot on-set and grade it so that the lit results of digital Mike (based on current WIP shading) can be matched to photographed reference as close as possible.

Since I have not got to the point where I use OLAT gray ball and Macbeth chart footage, I opt to expose and neutralize to the Macbeth chart as its gray chips values are more accessible.

zero direction Subject image: I pick _MG_1381.cr2. Use dcraw to convert it to linear TIFF. Exposure and neutralize it so the gray chip is ~0.198 using multipliers.

BG of Subject image and HDR is matched as good as I can in terms of the orientation of the room.

For exposing and neutralizing HDR, I lit a digital Macbeth chart with an HDR that is merged/stitched right out of PTGui. It's really underexposed and warm. I had to expose it up 6 stops. The Macbeth chart lit by 6-stop-up HDR looks like in Pic 1. Then, expose and neutralize the pic 1 with multipliers in Nuke so the gray chip is ~0.198 (in pic 2).

Pic 1

Pic 2

The same multiplier is then cloned in Nuke to expose and neutralize the HDR. I then check again the digital Macbeth chart's gray chip is ~0.198 when lit by neutralized HDR.

The exposed and neutralized HDR is then used to light digital Mike and a gray ball of 32% diffuse.

pic3. digital Mike lit by exposed and neutralized HDR.

_MG_1387.cr2 with the center of gray ball graded to be ~0.32.

As a comparison, digital Mike lit by Darren's HDR as the ground truth.

The skin in pic 3. render is redder than the neutralized ref Mike, although the center of gray ball in all images (ref image and two cg renders) is ~0.32 luminance and ~0.01 saturation.

Not satisfied with the results. To be continued....