Page 1 of 3 123 LastLast
Results 1 to 10 of 25

Thread: Inverse Square Law and window-light.

  1. #1
    Downstairs
    Join Date
    Feb 2008
    Location
    Italy
    Posts
    1,449

    Inverse Square Law and window-light.

    The ISL is my enemy and I fight back with:
    Fresnels with scrim flags,
    Neutral density wedges worn sideways,
    Very large window-lights with monstrous jagged-edge foam panel gobbos.
    Window-size beehive grids,
    Backgrounds (and foregrounds) painted dark on the near-side and light on the far-side.
    Black flags, Barn-doors, cardboard and anything available that might gradually feather off the light without screwing the shadows (shadows get screwed whenever the flag or barn-door happens parallel to the far side of the window.

    That's why round sources or sources with central hotspot are so much kinder to noses, boxes or anything else, for that matter).
    In fact, an umbrella with the lamp stuck well inside to avoid spill, and pointed across the subject does a better job with the inverse square law. It is round, has a hotspot (better penumbra and brilliance) and the edges of the brolly feather nicely all by themselves.

    But here is the question.
    Using a real window and a clear sky. There should be no Inverse Square Law dropping a stop or two at between nine and twelve feet from the window. The sky is at infinity. But it does drop. Like when the window is covered with a sheet of frost. Why?

  2. #2
    Between here and there
    Join Date
    Sep 1999
    Location
    Germany
    Posts
    514

    Re: Inverse Square Law and window-light.

    The light coming in through the window from the outside is just a source of light, and the further you move from that source, the less light you get. No escape.

    No, I can't explain it better than that, sorry.

  3. #3
    jp's Avatar
    Join Date
    May 2009
    Location
    Maine
    Posts
    5,628

    Re: Inverse Square Law and window-light.

    You are much better at lighting things than most of us; hopefully I'll be understood as just sharing my understanding rather than teaching.

    Looking out my window I see more trees and branches than sky. So the trees are reflecting and filtering the light coming in. Someone else might be half lawn, half sky. If you have a sunbeam coming in, that is of course coming from practically infinity and will not exhibit ISL. I would pretend the normal diffuse illumination from the window is a big rectangular softbox with a light in it; a natural rembrandt lighting source.

  4. #4

    Join Date
    Aug 2010
    Posts
    84

    Re: Inverse Square Law and window-light.

    Quote Originally Posted by Christopher Broadbent View Post
    The ISL is my enemy and I fight back with:
    Fresnels with scrim flags,
    Neutral density wedges worn sideways,
    Very large window-lights with monstrous jagged-edge foam panel gobbos.
    Window-size beehive grids,
    Backgrounds (and foregrounds) painted dark on the near-side and light on the far-side.
    Black flags, Barn-doors, cardboard and anything available that might gradually feather off the light without screwing the shadows (shadows get screwed whenever the flag or barn-door happens parallel to the far side of the window.

    That's why round sources or sources with central hotspot are so much kinder to noses, boxes or anything else, for that matter).
    In fact, an umbrella with the lamp stuck well inside to avoid spill, and pointed across the subject does a better job with the inverse square law. It is round, has a hotspot (better penumbra and brilliance) and the edges of the brolly feather nicely all by themselves.

    But here is the question.
    Using a real window and a clear sky. There should be no Inverse Square Law dropping a stop or two at between nine and twelve feet from the window. The sky is at infinity. But it does drop. Like when the window is covered with a sheet of frost. Why?
    Hard light coming through a window does not suffer from ISL (well, except to negligible extent that across the room is farther from the sun than less far across the room is). Soft light through windows is 100% bounced light and suffers from ISL the same way bouncing any other source with a soft bounce card (white side, not mirror) would. I'm not sure what the physics are here and it's probably not strict ISL, but this is what's happening.

  5. #5
    ic-racer's Avatar
    Join Date
    Feb 2007
    Posts
    6,737

    Re: Inverse Square Law and window-light.

    The closer the subject is to the window, the larger the angle of view out the window, which allows more light rays to hit the subject. As you move back, the angle of view diminishes. Eventually you will be so far back that you have only collimated rays from the window hitting the subject. These few rays are going to be only a very small subset of all the angled rays hitting the subject when the subject was close to the window, and therefore dimmer.

  6. #6

    Join Date
    Nov 2008
    Location
    Atlanta, GA
    Posts
    669

    Re: Inverse Square Law and window-light.

    Under some circumstances (e.g. w/fresnels) you may calculate the position of a virtual point source placed at a different distance than the actual source in order to see a result that maps distance in f-stop feet to falloff in f-stops.

  7. #7

    Join Date
    Aug 2010
    Posts
    84

    Re: Inverse Square Law and window-light.

    Quote Originally Posted by ic-racer View Post
    The closer the subject is to the window, the larger the angle of view out the window, which allows more light rays to hit the subject. As you move back, the angle of view diminishes. Eventually you will be so far back that you have only collimated rays from the window hitting the subject. These few rays are going to be only a very small subset of all the angled rays hitting the subject when the subject was close to the window, and therefore dimmer.
    I believe this is it; it's an issue with fast fall-off from all soft sources due to the varying angles of incidence, not ISL.

  8. #8
    Downstairs
    Join Date
    Feb 2008
    Location
    Italy
    Posts
    1,449

    Re: Inverse Square Law and window-light.

    Quote Originally Posted by Policar View Post
    I believe this is it; it's an issue with fast fall-off from all soft sources due to the varying angles of incidence, not ISL.
    That is, Ic-Racer has it right, the nearest side gets more sky. So the further the subjects are from the window the less difference there will be. Just like Inverse Square Law unfortunately.

  9. #9

    Join Date
    Jul 2010
    Location
    Chichester, UK
    Posts
    464

    Re: Inverse Square Law and window-light.

    Don't be fooled with all this the sun is 93 million miles from the earth crap. The window is an aperture and as such behaves in the same way as your camera aperture or any point source of light does - the position of the aperture is your light source and all calculations start from that point not not the position of the sun. Diffraction works from that aperture or window in the same way it works in a camera lens, otherwise you'd never get any light fall off from any daylight scene shot at close focus, after all what's an extra couple of inches when the sun is 93 million miles away?

    As an experiment take light meter readings from a window at doubling distances, the inverse square law will apply exactly the same as if the window was a flashgun set at that spot.

  10. #10

    Join Date
    Nov 2008
    Location
    Atlanta, GA
    Posts
    669

    Re: Inverse Square Law and window-light.

    Not so simple as that. There are multiple possible scenarios and one of them is the sun shining directly through the window onto the subject, which is equivalent (approximately, not counting ambient from other sources) to being outside. The window acts more like a scrim; since the rays are parallel there is no effect of limiting the number of converging rays. Not sure what you're referring to WRT diffraction as it relates to the exposure reading of the subject.

    Otherwise, it's as ic-racer describes when the window is transparent, and more like a broad source when the window is translucent or has a curtain etc. For a broad source, since ISL is based on a point source, you have to integrate over the the x and y dimensions to find the total intensity at the point where you're interested, i.e. sum the point sources that make up the broad source.

    I posted my calculations on the virtual point source for sources with a specified beam spread on the cinematography list some time ago, may have that archived somewhere if there's any interest.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •