selmslie Loc: Fernandina Beach, FL, USA

Ordinarily you can trust the camera's ISO setting and Sunny 16 (1/ISO seconds @ f/16) to arrive at a proper exposure setting for broad daylight that will prevent the highlights from blowing out. The worst case scenario for non-specular highlights in broad daylight can be white sunlit clouds, white feathers, PVC, white paint, etc. This is particularly useful if you plan to develop the result from raw on your computer.

Although this approach is reliable you might want to adjust it for conditions in a particular scene. For example, if there is a lot of white stuff in the scene you might want to use a slightly higher shutter speed. If there are no really bright highlights you can lighten the scene by reducing the shutter speed.

But what happens if you have the Bayer array removed? The sensor receives more light and the effective base ISO increases.

Here is a simple test I used to find the effective ISO for an A7 II monochrome after having the Bayer array removed.

Since I normally use f/11 for landscapes, Sunny 16 translates to 1/(2*ISO) @ f/11.

So all I needed to do is find a shutter speed at f/11 that works for the A7 II monochrome. Even though much of the scene was in shade, the sun was shining directly on the white clouds. At 1/640s and f/11 I got the result I was looking for:



The raw histogram comes close to the limit without blowing out. The camera's ISO was set to 200 so, if the Bayer array had still been there, the exposure would have been 1/400 @ f/11. But the shutter speed of 1/640s shows that the Bayer array was blocking 2/3 stops.

Here is the resulting unfiltered image. The only adjustment in PP was to add a little clarity. I didn't touch the highlights, shadows, contrast or brightness.

BebuLamar

Your result is to be expected. While not equally the r,g,b filter each blocks about 2/3 of the light. I think the g blocks the least. If I am not mistaken the b blocks the most light.

selmslie Loc: Fernandina Beach, FL, USA

It's complicated.

The Bayer ray uses a green filter over half of the cells and a red or blue filter over the remaining half. But the cells each record only the luminance of the light that their respective filters allow to pass. The filters are not sharp cutting filters. The green cells actually record some of the blue and red wavelengths and the same happens with the other two colors. If it were not for some cut filtering at each end of the visible range the UV and IR wavelengths would also be recorded.



But to achieve a panchromatic response the colors are not measured in proportion to the number of cells under the filters (RGB at 25%/50%/25%). They are weighted at 21% red, 72% green and 7% blue with the Bayer array in place to get the overall luminance. So when it comes to measuring luminance, green is by far the predominant color, red is next and blue is almost insignificant.

Once the Bayer array is gone, every cell gets unfiltered light. But the camera doesn't know this. It still applies the same weights to the cells it thinks are covered by the array. But the weights no longer matter. When you average three equal values you get the same average regardless of the weights.

selmslie Loc: Fernandina Beach, FL, USA

Here is another version of the filtered colors.



We can see between about 400 nm and 700 nm. Even the blue and green filters block UV better that the red filter.

Both plots are approximate and depend on the color of the light source, usually daylight.

camerapapi Loc: Miami, Fl.

The only camera that I have without the AA filter is the Olympus EM-1 D. I use "Sunny 16" here in South Florida often and I have not noticed any ill effects from using it. Your scene has a large dynamic range. Confronted with something like this I take the exposure from the bright clouds and correct accordingly known I will have to open the shadows in post. Or I can go with HDR photography that makes it easier.

I have not done any testing with my EM-1 so I am not in any position to discuss your findings.

selmslie Loc: Fernandina Beach, FL, USA

A lot of people dismiss Sunny 16 without understanding the reality behind it.

Having lived most of my life in Miami and nearly all of the rest in either Cuba or north Florida, I can attest that it works for almost all subjects in broad daylight. It even works on sunny days in Scotland, Iceland, Scandinavia and the rest Europe so long as the sun is shining.

So why do cameras need meters? It's for when the sun is blocked or not shining directly on the subject. That's rare in the Sunshine State.

Frontlit:


Backlit:

bwana Loc: Bergen, Alberta, Canada

I just let the camera do the exposure. Works well!

bwa

selmslie Loc: Fernandina Beach, FL, USA

Here is another example of Sunny 16 with a camera that still has its Bayer array. The situation was similar to the first example with direct sunlight on the white clouds while the foreground was not. Since the Bayer array is still present the effective ISO matches the ISO set in the camera.


Nikon D610 1/800s @ f/11 ISO 400

The scene's dynamic range is about 9 stops. The green channel is less than a half stop from its upper limit.

profbowman Loc: Harrisonburg, VA, USA

What is your source for this last graph? Did you get the first one from the list web site? I like to read the surrounding material. Thanks. --Richard

selmslie Loc: Fernandina Beach, FL, USA

The first one came from https://www.monochromeimaging.com/technical/full-spectrum-ir/

The second one came from a Google search for similar plots. https://www.google.com/search?sa=X&source=univ&tbm=isch&q=cmv2000+sensor+spectral+response&hl=en&ved=2ahUKEwjvuYT3hPDxAhXGLc0KHa8YAlkQjJkEegQICRAC&biw=1522&bih=1578