For cameras that photograph visible light...

I understand that much of the light entering the outer lens element never reaches the camera sensor. So considering this, I am wondering further...

Of all the light that passes through the lens's entrance pupil, what proportion of that is useful for image creation?

In comparison to the amount of focused light that strikes the sensor, is there much that is out of focus and that does not increase the information content of the final image?

My interest is in how camera lenses "filter" information (not in the lens add-on filter sense but by way of overall information reduction).

  • $\begingroup$ Anything that hits the sensor should contain image information, since the sensor has no way of distinguishing between photons that do and photons that don't $\endgroup$
    – user107153
    Commented Jan 3, 2018 at 8:15
  • $\begingroup$ The image is everything that the sensor detects. All light through the lens from the objects registered on the sensor creates the image, except for the light lost due to (1) glass imperfections, such as surface reflections and others, (2) sensor efficiency, and (3) the aperture limitation. The lens does not filter anything from the image. Perhaps you meant something else, like how the part of the image in focus relates to the part of the image out of focus, but then you should state you question more clearly. $\endgroup$
    – safesphere
    Commented Jan 3, 2018 at 18:02
  • $\begingroup$ So to sum the major sources of intentionally designed light reduction (this is still in question form :) (1) The lens barrel/hood (2) the aperture (increasingly as it is stopped down). (3) lens magnification (reduces field of view) For unintentional sources of light reduction (1) glass imperfections (2) sensory inefficiency. $\endgroup$
    – Nick Mirro
    Commented Jan 3, 2018 at 19:24
  • $\begingroup$ I would like to emphasize that I meant 'useful' image information. That is light that helps form the meaningful image. I have one further clarification question. For a lens focused at infinity and without visible vignetting on the sensor, does a roughly known portion of the light passing through the entrance pupil not focus on the sensor? Less than 10%, 1/3, 2/3? Thank you for helping. $\endgroup$
    – Nick Mirro
    Commented Jan 3, 2018 at 19:30
  • $\begingroup$ You should use the @ sign followed by the name (with no spaces) of the person you are asking. Otherwise no one is notified about your comment. The sensor efficiency includes the angle. Light that comes to a pixel straight passes to the sensor in a bigger proportion that light coming to the same pixel on an angle from the opposite side of the lens. $\endgroup$
    – safesphere
    Commented Jan 5, 2018 at 6:17

1 Answer 1


The camera does not filter the light per se.

The actual truth is the all the light rays that go through the lens contains information about SOME images - the thing you need to understand is that only rays emitting from a certain object will come into focus on the sensor.

Generally speaking, a lens will always produce an image of any given object (as long as the object is in the field of view of the lens), but the location of the image is directly dependent on the distance between the object and the lens (and also the effective focal length of the lens).

Since the sensor is always a fixed distance from the lens (ignoring for a moment focusing mechanisms), then only objects with the right distance from the lens can come into focus on the sensor, BUT objects that are a different distance from the lens also come into focus - just on other planes that do not coincide with the sensor.

So the fraction of light 'with information' (aka, focusing on the sensor) is the fraction of the total incoming light that has originated from the 'right' distance such that it would come into focus. This region is called the focal plane of the camera (in the object space), and its depth is called the 'depth of field' (you may have heard of it:).

  • $\begingroup$ And depth of field increases when the lens aperture is decreased. $\endgroup$
    – user137289
    Commented Jan 3, 2018 at 12:19
  • $\begingroup$ Correct. I don't understand what you're getting at? $\endgroup$ Commented Jan 3, 2018 at 12:59
  • $\begingroup$ Just an additional note, mainly directed at the OP @Nick Mirrow - one can get an image with higher information content by admitting less light. $\endgroup$
    – user137289
    Commented Jan 3, 2018 at 13:12
  • $\begingroup$ So stopping the aperture down is collimation that shrinks the area of the entrance pupil, further reducing light that will reach the sensor. $\endgroup$
    – Nick Mirro
    Commented Jan 3, 2018 at 18:51
  • $\begingroup$ I wasn't aware of the 5 minute edit limit. In the above comment, I am assuming that the aperture reduces light even in the wide open position. Light is then further reduced by stopping down. Less light striking the sensor increases the information in the image by "filtering" metaphorically. $\endgroup$
    – Nick Mirro
    Commented Jan 3, 2018 at 19:05

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.