Originally Posted by threephi
As a professional lighting technician, I can affirm that this is incorrect. Incident meters are designed to measure the intensity of light at the point of the sensor, whether your source is broad and diffused or narrowly focussed.
What you really need to judge a headlight's effectiveness (electrical and durability considerations aside) are two pieces of data: beam characteristics (shape, falloff, etc) and intensity. Beam pattern photos at a standard distance and all shot at identical exposures are a good way to judge the former, and an incident lighting meter that gives you readings in footcandles is an excellent way to judge the latter. A good set of data might be to measure the distance from the fixture at which you get 2 footcandles in the brightest part of the beam.
I am going to disagree that beam photos tell you much of anything useful at all. All it tell you is that some guy at a agreed upon distance with an unknown camera, an unknown sensor/film, and an unknown lense used an aperture / shutter speed that may or may not be in agreement with what everyone else is using took a picture against a wall that is an unknown color. Most of the people taking the pictures may well be using software/camera firmware/scanner firmware that makes it "look nice" with out actually telling them anything.
That tells me nothing other than lights sure are pretty. It is nice to see beam patterns that is certain.
but if I can take a spot reading against a known object (like a grey card) at various parts of the beam it tells me a hell of alot more than a picture some dude took after pointing a light at his wall.
wether it tells anything to anyone else.. perhaps that is a different matter.
If you want I can go back and color correct all of those light pictures and make them look like totally awesome lights.
