I looked into the value of light meters again, The ASC DP's in fact did conclude that relying on light meters led to exposures that were 1-2 stops off. But these were tests with RED, Genesis, Viper, and Arri digital cameras, ie top shelf cinema cams. The light meters are designed to emulate the near-logarithmic response of silver halide film, and thus useless for setting exposure of a digital sensor. (adjusting lighting ratios on the other hand, that's a different story) They recommended to ALWAYS set exposure for a digital imaging system based on waveform, and histogram if you have access to such. But the tests weren't done with the types of cams most people here use, such as the Canon 5d. Barry Green, dunno what you think of that guy, would extend this thinking down the food chain, saying The waveform reflects what all the camera processing has done to the signal, and is the only truly accurate way to set exposure. It takes everything into account. If lacking a waveform, use zebras. A lightmeter can only be used as the vaguest tool for getting somewhere in the ballpark, but not for accurate setting of exposure.
Lightmeters are for film, or for scouting sets, or for setting a distinct ratio of lights, but exposure (on a video or digital cine camera) should be done using the appropriate mathematical digital tools, the waveform monitor and the zebras. But that's Barry Green's take, he's no ASC DP.
