Editing Red

[Inkanyezi]

The lightmeter doesn't have something as a Bayer array. It just measures the luminace. And the camera might correc something based on the different fields of the lightmeter when using matrix metering.



Let's be polite. For this moment this sounds as complete nonsense to me.

George

Ok, I will try to be polite.

We assume that the light meter assesses cd/m², whatever it reads. Now if we have a white patch that reflects 120 cd/m² and beside it a red, a blue and a green patch as well as a black one, the energy of the incident light will be unevenly absorbed. The red patch will absorb virtually all green and blue light, the green patch will absorb virtually all blue and red light, and the blue patch will absorb red and green. The black patch will absorb all three colours.

When measuring the luminances from each one of them, the white pach with 90 % reflection will certainly reflect the highest number of candelas, 120 cd/m².

The illuminance for all is equal, and assuming the white incident light has equal energy in all of the three colours, the red patch will reflect 30 cd/m². So will the other coloured patches. The black patch may reflect less than 2 cd/m². The rest of the energy is absorbed by the coloured patches and the black. The coloured ones absorb 70 % of the light, and the black one near 99 %. Absorbed light is turned into heat, making the surface warmer. The white surface that absorbs 10 % will be less heated than the black one that absorbs 99 %. Those other coloured surfaces absorb about 70 % and will also be substantially heated.

When we photograph these different patches, we might want the white reach zone IX and the black zone I, while the coloured ones should be placed in zone VIII. Note that it is only for convenience we assume them to reflect the same amount of light. Depending on colour temperature and balancing, the real figures will certainly be other.

If we measure the red patch as if it were white, and assume that we place it in zone IX, which is the place where it would rightly be if it had been white, its red luminance will be higher than the red channel of the sensor can assess. Our meter, whether completely colour indifferent as most hand held meters, or tri-colour as in a camera, will not assess its correct value to fill the red channel without over-flowing. The white patch reflected 120 cd/m², but the red one only 30 cd/m². The meter will tell us to expose the red patch three times more than the white patch, in order to make it as bright as a white of 120 cd/m², but as it is one single colour, with the two other absent, it will fill up the red sensels at only a third of that exposure. The red will be clipped if we expose it in hope of getting it as bright as the white patch.

The red coloured patch reflects less light than the white one, less cd/m², and it cannot be used for evaluating exposure unless we know how to compensate for the lack of the other two colours. We have to reduce exposure about three times in order to render it below clipping.

I know that this may be difficult to grasp, but nevertheless, that is the way it works.

And that is why exposure will have to be reduced with a saturated red of high luminance, as this image which has been exposed with compensation of minus two full stops:

[george013]

When we photograph these different patches, we might want the white reach zone IX and the black zone I, while the coloured ones should be placed in zone VIII. Note that it is only for convenience we assume them to reflect the same amount of light. Depending on colour temperature and balancing, the real figures will certainly be other.

I would say that the coloured ones with the figures you gave are more on the left site of the histogram.

If we measure the red patch as if it were white, and assume that we place it in zone IX, which is the place where it would rightly be if it had been white, its red luminance will be higher than the red channel of the sensor can assess. Our meter, whether completely colour indifferent as most hand held meters, or tri-colour as in a camera, will not assess its correct value to fill the red channel without over-flowing. The white patch reflected 120 cd/m², but the red one only 30 cd/m². The meter will tell us to expose the red patch three times more than the white patch, in order to make it as bright as a white of 120 cd/m², but as it is one single colour, with the two other absent, it will fill up the red sensels at only a third of that exposure. The red will be clipped if we expose it in hope of getting it as bright as the white patch.

I don't know how you can measure the red patch as if it were white. But I think you mean that you overexpose the red patch so that it comes in zone IX. Then you have a problem. The tolerance you have with overexposing without clipping is determined by the highest channel value. Not by the absence of 2 other colours.

The red coloured patch reflects less light than the white one, less cd/m², and it cannot be used for evaluating exposure unless we know how to compensate for the lack of the other two colours. We have to reduce exposure about three times in order to render it below clipping.

And this is what I think complete nonsense.


George

[Inkanyezi]

I don't know how you can measure the red patch as if it were white.

That is simply what the colour blind light meter does, or we assume middle grey, just the same. It will give a lower value than the real one, as it is only one of the colours in the "grey" for which the meter is calibrated.

But I think you mean that you overexpose the red patch so that it comes in zone IX.

That is where we could place it when shooting black and white, if there is no white present in the subject.

Then you have a problem. The tolerance you have with overexposing without clipping is determined by the highest channel value. Not by the absence of 2 other colours.

That's what the whole discussion is about. The meter assesses a lower value for any colour, as it assumes all colours to be present.

And this is what I think complete nonsense.


George

You were the one suggesting politeness, something you now broke at least twice. To me your ramblings are nonsense, but nevertheless, I try to explain basics of colour photography and usage of the light meter. Should I really care to be polite?

When the red colour reads as a third of what white does when measured for zone IX, it reaches the highest channel value that the sensor can register.

[george013]

Urban,
The lightmeter DOESN'T READ COLOURS. It just reads the value of the reflected lights in cd/m2.
If you have a subject that will give with a good exposure the calculated value of 250,0,0 then you will have no room for overexposing without clipping. But that same counts also for 250,250,250.

Grey is another name for a color with equal parts of R,G and B. So white and black is also a kind of grey.

George

[Inkanyezi]

Urban,
The lightmeter DOESN'T READ COLOURS. It just reads the value of the reflected lights in cd/m2.

I have not suggested that the light meter reads colours. You need not yell that to me, I know it does not read colours. That is the very reason why its readings sometimes are off.

As you say, it reads reflected light in cd/m², although an incident meter does not. Only reflected light meters do.

For black and white photography on film, this will not cause a problem, as B&W film, like the light meter, is not sensitive to colour. However, the digital sensor is sensitive to colour. If a white surface would reflect 120 cd/m² and we take a reading from a saturated red surface that reflects 30 cd/m², it should be rather evident that the red colour will reach a much lower reading on the meter than white. However, although just a thrice of the cd/m² value, it will fill the red sensels just as much as white would.

[george013]

As you say, it reads reflected light in cd/m², although an incident meter does not. Only reflected light meters do.

Why wouldn't an incident lightmeter work the same as a reflected lightmeter? An incident lightmeter is correcting the incident light based on a average reflection of 18%, or 12% or 20%.

For black and white photography on film, this will not cause a problem, as B&W film, like the light meter, is not sensitive to colour.

You see the different colors, only in grayscales.

However, the digital sensor is sensitive to colour. If a white surface would reflect 120 cd/m² and we take a reading from a saturated red surface that reflects 30 cd/m², it should be rather evident that the red colour will reach a much lower reading on the meter than white. However, although just a thrice of the cd/m² value, it will fill the red sensels just as much as white would.

The reading will be different, but also the advised settings. Whatever surface you shoot, black,brown,yellow,red or whatever will be placed in the middle of the histogram. What the lightmeter does, is calculating a aperture/shutterspeed setting for the received lightintensity so that the resulting lightintensity the sensor reaches is somewhere in the middle of the senstiviness the sensor/film is using.

George

[Manfred M]

The discussions on this thread are not going anywhere and should have been closed some time ago. Time to shut it down.