We continue to hear suggestions that perhaps, when making a white balance determination with a diffuser-equipped camera, used at the camera position for the shot, aimed at the subject, it would be advantageous for the diffuser to have a fairly narrow directivity pattern.
I wanted to give a little demonstration of this. Unfortunately, I didn't have a diffuser with a really narrow pattern, but I did have my trusty Color Parrot (which has the narrowest pattern of all my serious diffusers), so I cheated a little. I made the measurement from fairly near the subject (Carla) so that the diffuser's "intake" would be have a significant component of light reflected from the subject, as I would if it had an even narrower pattern.
I took a CWB reference frame with my EOS 20D that way, and then another CWB reference frame from a WhiBal gray card in front of the subject. I took two real shots using CWB, one with the "Color Parrot real close" reference and one with the WhiBal gray card reference.
Here we see the shot with "Color Parrot up close" white balance correction:
Here we see the shot with "WhiBal gray card" white balance correction (I forgot to call, "this is picture", so excuse the model's "off the clock" expression):
This is of course the result we should expect.
Measuring the light reflected from the subject is not an object of taking the measurement from the camera position. Rather, our hope is that the overall ambient light falling on the diffuser (at the camera position) will be similar to that falling on the subject.
Thus, making the camera-plus-diffuser concentrate on the subject (by working toward a narrower directivity pattern) is wholly counter-productive.
Linear Mode
