Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Camera Mystery

  1. Apr 1, 2005 #1
    This is probably painfully simple, and is not a 'brain teaser' as such but this seemed like the most suitable forum.

    My bro and I were having a discussion tonight and we stumbled across a problem that probably lead to more discussion than it is worth, but since we still can't come up with a proper reasoned answer, I thought I'd post it up here.

    If you have a video camera, giving a live feed on a T.v say, what would happen (what image would you see) if you pointed the Video Camera at the screen broadcasting the feed.

    It is positioned so that it is focused only on the screen part of the TV (you cannot see any of the plastic/ etc around the glass screen). Furthermore, the screen is totally clean (no marks of any kind) and also there is no reflection of the camera or anything else on the screen.

    What do you see?

    I'd test it out, but don't have a working camera, (although I doubt I could emulate this 'perfect' environment anyway).

    As I said, it may sound stupid, but there is enough doubts in our minds to put the question to others.
  2. jcsd
  3. Apr 1, 2005 #2
    hmm i think it would produce some sort of a feedback whereby you would see an infinite amount of television screens in the television screen?
  4. Apr 1, 2005 #3
    yeah I can see why that would happen if you positioned the camera at the WHOLE television ie the frame and the screen.

    But what if you positioned the camera at ONLY the screen....
  5. Apr 1, 2005 #4


    User Avatar
    Science Advisor
    Gold Member

    The original Dr Who title scenes were filmed in this way!
  6. Apr 1, 2005 #5


    User Avatar
    Gold Member

    Why guess? Someone try it!
  7. Apr 1, 2005 #6


    User Avatar
    Gold Member

    My TV screen, even when off, is not pitch black, it's a dark grey-green. You'll never start with black.

    I'm presuming you'd have to do it in a pitch-dark room, otherwise, all you'd get is ambient light from the glass layers, which would be picked up and broadcast.

    But of course, video cameras, will compensate for darkness, which means that, even though your camera will see blackness, it will overexpose it ,and send a signal of grey. That will be picked up by your camera, and eventually, you will have a plain, grey image.

    Ever gotten a snapshot back from the photo lab of a blank picture? They over-process it, so that you don't get a pitch black print, you get an over-exposed grey image.
  8. Apr 2, 2005 #7
    i think that as there has to be some(even if very small) time lag between switching on the camera and getting the picture on the screen. so during that time the camera will capture the blank screen(whichever colour it is, black or grey) when this is shown on the screen you get the picture if the blank screen on it.as the frame is not included, only the screen will appear and it will look as if the screen is switched off. now this is again captured by the camera and again it will send the image of the screen in which it looks as if it is switched off, whereas it is actually not. this goes on in a loop and at all the time the screen will appear as it is switched off.

    did not try it out, but i think even if someone does, the screen will look no different on switching the camera on than it did before that.
  9. Apr 2, 2005 #8


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I pretty much agree with this. I think the final result will be what they call a Moire pattern. Or simply a lot of video snow (noise).
  10. Apr 2, 2005 #9
    In my turn i also pretty much agree with this :smile:
    Some time ago i have taken some snapshots, and even made a small video - there was really some kind of bluish-black Moire pattern.
    However my video was taken in such a way that i saw the boundary of the TV, and while i was rotating the camera here and there the TV boundary on the picture was also rotating however with a certain delay!

    Well, do you want me to post the screenshots and videos in the attachment here?
  11. Apr 11, 2005 #10
    That's what happens, though the number of screens within screens isn't infinite. You reach a certain point where it ends in a blur. If you angle the camera one way or another, the "tunnel" veers off in a curve.
  12. Apr 15, 2005 #11


    User Avatar
    Gold Member

    Actually, I have a question about that. If you have two parallel mirrors facing each other, one mirror would have the image of the mirror it's facing, then itself, and so on. This "curve" that you're talking about occurs in parallel mirrors, not only at an angle, and I'm wondering why.
  13. Apr 16, 2005 #12
    with the mirrors it depends on the direction from which you look into the mirror
  14. Apr 17, 2005 #13
    and if we some how look through the mirror perpendicularly. so we would observe a simmilar pattern.

    and i also suggest that the number of mirror edge we would see if we look in an angle will be able to be predicted with tan.
  15. Apr 17, 2005 #14


    User Avatar
    Science Advisor
    Gold Member

    How the video camera/TV screen scenario works depends on several things. I've done this before. You get an infinite number of TV screens and each one gets brighter and not darker. If you zoom in on a spot on the TV screen you can get some color effects especially around the edges. You can also get an affect that looks like you are going down a tunnel or something.
  16. Apr 18, 2005 #15
    do we have explenation on that
  17. Apr 18, 2005 #16


    User Avatar
    Gold Member

    When you zoom in, each TV screen is going to get larger until the first dissapears, a going down a tunnel effect
  18. Apr 19, 2005 #17


    User Avatar
    Gold Member

    Mirrors and video cameras are not the same thing and do not produce the same effect. A mirror, rotated by 5 degrees (around an axis that passes through both mirrors) does NOT rotate its reflection by 5 degrees, whereas a video image does.

    Yes, you still see an "infinite regression", but it does not rotate.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook