So we know coronagraphs are just a circular mask in the center of a telescope to block the sun's bright light and allow our eyes to see into the dark around the sun. But what if this coronagraph could block any bright areas of an image that are too intense for an image sensor to read? The goal would be a bit like an electronic neutral density filter, except it's using a LCD matrix to block bright sections of the incoming image.

I can imagine this possible with a photochromic lens (used in Transition eyeglasses) except divided into a grid of little pixel squares that match with an image sensor.

  • 1
    \$\begingroup\$ What use case are you thinking would benefit from the effort? Why wouldn’t post processing of multiple images solve that problem? Who would pay for the thing if it existed? \$\endgroup\$ Oct 9, 2021 at 23:01
  • \$\begingroup\$ The problem isn't what the device could do to the image when used, but what it will do when not used (loss of light, of resolution, etc....). Or it has to be removable. \$\endgroup\$
    – xenoid
    Oct 10, 2021 at 10:48

2 Answers 2


It doesn't work.

enter image description here

(Source: Wikipedia)

As you can see, the light does not follow a single path from the object to the projected image / camera sensor. Far an in-focus image, we put the object at S1 and the sensor at S2.

We can add another lens and project the image at S2 to a sensor on the other side (e.g. S4). We call the image at S2 an intermediate image.

For a coronagraph, we just add a mask at S2. That only works because the sun and the stars are in focus at the same time.

Usually, we do not have everything in focus, though. Out-of-focus areas will produce circles instead of points at S2, and different circles at S4. So, it is not possible to effectively filter out-of-focus areas at an intermediate image plane.*)

Another problem with such a filter is that the filter's structure is in focus. It will produce nice Moiré patterns when not 100% aligned with the sensor's pixels.

The only place to put this adaptive filter is in the sensor.

*) There is one exception: You can do it for a very small aperture (like a pinhole camera), because there is essentially only one path from any point of the object to the sensor. But the image will be so dark already that you won't need any more darkening. Otherwise you'd need something like holographic technology.

  • 1
    \$\begingroup\$ That's true, at least for a simple lens, but an intermediate image plane would fix this problem. Camera lenses might not use them, but microscopes do \$\endgroup\$
    – Chris H
    Oct 25, 2021 at 15:19
  • \$\begingroup\$ @ChrisH I suspect that a filter at an intermediate image plane will not work for out-of-focus areas of the picture. \$\endgroup\$
    – user24582
    Oct 26, 2021 at 12:03
  • \$\begingroup\$ Now that's an interesting thought. Probably not \$\endgroup\$
    – Chris H
    Oct 26, 2021 at 15:37
  • 1
    \$\begingroup\$ Edited answer to include intermediate image planes. \$\endgroup\$
    – user24582
    Oct 28, 2021 at 8:41
  • \$\begingroup\$ The only place to put this adaptive filter is in the sensor. - maybe that's a possible answer? What if you positioned a piece of photochromic, fibre optic glass directly over the sensor? Requires no power, and it will variably darken to different levels depending on the light that hits it from the lens. Won't be fast but it could be tested fairly simply. \$\endgroup\$
    – AutoBaker
    Jun 27, 2022 at 15:02

This doesn't currently exist on the public market. I can see what you're saying, and it's certainly an interesting proposition.

Currently, when we want to capture an image that uses exposure ranges too great for the camera, the normal solution for this is to use some form of HDR photography:

  • Mobile phones have HDR modes built in.
  • Most DSLR's have an AEB (Auto Bracketing) Mode that takes more then one photo at different exposures.
  • In manual mode we can bracket photos using exposure compensation.
  • In video, we shoot 'flat' to decrease contrast and give us a wider exposure range to edit with in post.

These bracketed photos can be merged using the HDR tool in Photoshop, or you could custom merge them using blend modes and masks to get the result you want.

What I'm trying to say is that we already have ways to achieve the same result as your idea, using tools that are already widely available. I'm not sure if there would be much interest for a custom tool to do what you propose as it has such a specific use case.

As per the comments on the post, it could also be detrimental to photos - if it wasn't used, will it decrease quality, darken the photo or change the refraction of light through this tool?

Another thing to think about, is that for this tool to work correctly, it would have to have a live output from the camera LCD to know exactly what parts to darken and lighten. If your tool is any distance away from the sensor, it will be difficult for it to know exactly where the light is hitting the sensor, so it will have to read this from the camera itself. And then how does it work out how much to modify light?

Might be opening a can of worms. If you've got an idea, go ahead and give it a go :) But realistically I think most photographers would stick with what they have until it becomes mainstream and/or manufacturers develop and offer it.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.