Trying to find a continuous wave model for something that is a quantum phenomenon is always going to be problematic. Remember that exposure was formalised before the development of QM. More interestingly, studying the effect of very low light levels on photographic plates was part of the process that settled that QM applies to light.
May 31, 2023
181
-
So light meters can not be calibrated to the CIE luminous efficacy curve, it would appear.
-
They certainly can. The problem is, how do they react to sources that don't conform the the CIE efficacy curve. There's no reason that all meters calibrated to the curve will behave the same under different conditions. Most problematic, you can devise light sources that would provide the same set of colours to the standard observer, but would meter quite differently.
-
How about they certainly are?
No, there's no reason that they should. Don't make me insult you by claiming that a source of blue light, nay UV light, would make the light meter recommend more exposure or make a lux-meter say "nope, not much light there" ...
That is a Straw Man, sorry!
-
As Danno says, "hitting the sensor", so I'll agree with him and say that different CFAs and different hot mirrors will cause different exposuires all else being equal.
It comes as no surprise to me that when I remove the UV/IR filter from a Sigma SD1/14/15, the metering lies horribly ...
-
So you're saying that the sensor in the definition is only the silicon part? Does that mean that for the purpose of exposure calculation, you would like to apply the CIE Y spectral sensitivity curve to the energy that passes through the CFA and hot mirror? How then would you measure the exposure for an IR-only camera? It would have response with zero exposure. That doesn't make a lot of sense.
-
No.
For the average punter, there is no such thing. But if we are talking about a camera with a permanent IR filter and only for use with IR it makes sense to me that the metering method would be for IR, i.e. not photometric.
I know of no radiation measurement device that "would have response with zero exposure". Please don't explain, it should make no sense to anyone
-
If light is measured in lux, then you can get finite sensor response with zero light, if the camera's spectral sensitivity is sufficiently different from CIE Y. Consider a camera with the hot mirror removed in the presence of narrowband 1000 nm radiation.
-
Then what do you mean?
-
Ah ,zero light - not "zero exposure" ... now we are saying the same thing.
-
If exposure is defined in terms of lux, then it would be zero exposure for the 1000nm narrowband light source.
-
You said:
"In your way of thinking, does the CFA spectral response affect the exposure? Does the hot mirror affect the exposure?"Since "the exposure" can mean almost anything these days - here's what meant:
If Hm is constant upon a sensor then a) replacing the existing CFA with a different one will affect the photo-sensor's response or b) if the UV/IR filter is removed, will affect the photo-sensor's response.
Where Hm=0.65Lt/N^2 and L and t/N^2 are constant
-
As you like.
I'm unaware of putting any compulsion on you to insult me. Consider this, the CIE efficacy curve is a perceptual one, not a physical one. The physical ones are the standard illuminants. A really well designed exposure meter should be weighting its response according to the luminosity curve, but very few do. I know of no camera's response that does either. So you might calibrate the meter using a light source (best use the appropriate standard illuminant), which the meter will weight appropriately. Take a photo setting the exposure according to this meter. Now use a different light source with a different spectral distribution. Again meter for the same exposure. Because the camera's sensor response is not the same as the meter's then the apparent exposure for the camera will be different.
Just how is it a straw man?
-
No comment.
-
No comment.
-
It sounds like you are taking the "sensor" to be the combination of silicon chip, CFA, microlenses, and the rest of the sensor stack. Right?
-
No comment.
-
Hard to have a discussion with no agreement on definitions...
-
Erm, OK (backs off slowly and quietly).
-
It's whatever exposure that I prefer, on an individual image. It holds the proper "key" for the subject and mood of light, with ample detail in both shadows and highlights. An old Fuji EXR camera gets it right for me most of the time from a jpeg. When not, I shoot again after a minor adjustment to EV.
