Suggestion: Crowdfunding the Creation of a Traditonal DSLR style Full Frame SD2 Merrill

jande9 wrote:

tagscuderia wrote:

RobLI wrote:

... the problems with the R layer being buried deep in the silicone and ending up very noisy.

Just to clarify, the R layer is NOT "very noisy," check SNR with RawDigger; the 3x3 matrices used for colour conversion generates/amplifies noise. A genuine issue with Merrill is that if you expose for the B layer, then the R layer is underexposed. Expose for R and B is clipped

N.B, and it's SPP's that introduces the colour blotching, it's absent from the X3F and neither Kalpanika nor Iridient suffer from it.

Also Quattro's B layer has a greater effect on IQ than Merrill's R, just to add some perspective.

Does this imply that one should take the colour of the incident light into consideration when calculating exposure?

Is it not done automatically, Jan?

Exposure (of the sensor) has units of lux-seconds (lx.s).

Lux, or it's buddy cd/m^2, is used in the camera's metering calculation which ends up in camera exposure EV or any of the combinations of f-number and shutter speed for that EV.

The unit of lux is based on human vision:

The metering sensor should have a response like the above, same as separate lux-meters do.

So it can be seen that blue light (say 450nm) coming in at some radiometric power has much less lux than green light (say 550nm) at the same radiometric power, meaning that the camera metering should recommend much more exposure for a blue scene than for a green scene.

Is the Merrill more forgiving of over exposure under tungsten light than daylight because the risk of blowing the top layer is lessened?

Hmmm . . I'll let Tom answer that . .