But Were Afraid To Ask!
I pose the question: Does a homemade lightprobe need to be white or 18 % gray to be effective?
So the lightprobe standard is :
one - side chrome
one - side 17% grayCorrect?
I have purchased a garden globe (10-inch glass, chrome sphere) at Michael's craft store (San Francisco bay area crafts store) and bought some primer gray spray paint. Does anyone know if you can buy 18% gray spray paint. Could you use a white side of the sphere to capture the color of the reflected light? In your book, Digital Lighting and Rendering (on page 273) the sphere looks rather "white" instead of gray.
Jeremy Birn replies...
This isn't a "standard," and everyone does things a little differently. Your lighting reference doesn't even need to be based on a sphere - a little model of your 3D object, or even just a white card that you point in several directions, could be just as good reference. Some people make a "sun-dial" type thing (a vertical stick with a white disc of cardboard running through it) to see the different shadows and light directions that way.The idea of the 18% gray sphere is just that it is likely that you can shoot a picture of it with the camera adjusted to the same exposure settings as it's likely to be already for filming the shot. (Based on the assumption that an "average" surface reflects back 18% of the light that hits it.) If you have a light meter, you can test your paint job until a reflected light reading is 2 1/2 stops lower than an incident reading. (2.5 f-stops actually comes to 100 / 2 ^ 2.5 = 17.677% gray, but who the hell would mix paint that way I don't know!)
It's great to collect whatever reference you can, but please remember that this is an art, not a science, and don't be too rigid about how you collect your reference (there's little tolerance for slowing down a shoot for this stuff) and don't put too much faith in your process of collecting and matching it (no director is going to accept your scientific judgment that the lighting is "correct" if it doesn't look good to him yet.)
The best that these things do is help guide you in roughing-in a starting point for your lighting, and since big factors like the shaders and texture maps on your objects can radically darken or brighten your objects in your final renders (some shaders can look twice as bright in the same lighting as others), a small difference in how you create your initial lighting match is not always going to matter much.
Read more information on 3D & CG lighting at Jeremy's website: http://www.3drender.com