Page 1 of 2 12 LastLast
Results 1 to 10 of 16

Thread: dove prisms, can they do this?

  1. #1
    Join Date
    Mar 2018
    Posts
    96

    Default dove prisms, can they do this?

    I need an optical element that when rotated will rotate a video projection.
    Someone on reddit suggested dove prisms. I have no experience with them so although I found few illustration videos from thorlabs and few suppliers, I would still like to have an idea which one I should try out and if my expectations can actually be met.
    https://www.youtube.com/watch?v=lMyuHy9WZGM

    I only found one video that maybe does something similar: https://www.youtube.com/watch?v=Jvv22wI3qY0

    It is my understanding right now that the faces where the beam goes in and out of have to be big enough to fit the projection beam diameter.

    Not sure if this is all I need to consider.

    I would like to know for example, if the beam can reflect inside the prism and still look fine as long as it isn't clipped coming out from the other face.

    BTW, it doesn't have to be a dove prism, if there are other less known prism types that can do this or other not expensive optical components those also may work so please mention them. Dove prisms have one disadvantage for example for my use case, which is one full rotation causes double beam full rotation so I'm open to other ideas.

  2. #2
    Join Date
    Jun 2009
    Location
    St. Louis, MO
    Posts
    1,197

    Default

    Yes a dove prism would work, but it would a whole lot easier to play the video back using a computer and rotate the image using a tiny fraction of the power of a modern video board...
    "There are painters who transform the sun into a yellow spot, but there are others who, with the help of their art and their intelligence, transform a yellow spot into the sun." Pablo Picasso

  3. #3
    Join Date
    Mar 2018
    Posts
    96

    Default

    Sure, but you would lose resolution in one axis because the aspect ratio isn't 1:1.

  4. #4
    Join Date
    Jun 2009
    Location
    St. Louis, MO
    Posts
    1,197

    Default

    I grew up in a planetarium - think in terms of inscribed fish-eye lenses or multi projector setups mapped and blended to display content on the dome. If you want to maintain a rotating rectangular image at the resolution of one projector stack two projectors and blend them.

    If you must a dove prism you'll need collimated light passing through the dove - then focus it again for the projection lens. You'll want to design your optical system so that all of the light from the projector makes it into the projection lens. i.e. Select the focal lengths and f numbers of the lenses in your collimator and focus lenses to match the entrance pupil of your projection lens. Then you'll know how big a dove will be needed.
    "There are painters who transform the sun into a yellow spot, but there are others who, with the help of their art and their intelligence, transform a yellow spot into the sun." Pablo Picasso

  5. #5
    Join Date
    Mar 2018
    Posts
    96

    Default

    Any ideas what happens with uncollimated projection beam through a dove prism? Edmund says it recommends collimated light for "optimal results" and Thorlabs says it may produce astygmatism. Just how bad it is isn't clear though.

    Adding lenses to collimate and diverge the projection beam again sounds difficult.

  6. #6
    Join Date
    Jun 2009
    Location
    St. Louis, MO
    Posts
    1,197

    Default

    Very little of the image will pass through the dove prism if it's uncollimated...
    "There are painters who transform the sun into a yellow spot, but there are others who, with the help of their art and their intelligence, transform a yellow spot into the sun." Pablo Picasso

  7. #7
    Join Date
    Nov 2008
    Location
    Cleveland Ohio
    Posts
    2,599

    Default

    Why does it not also break up the colors? What if render the video out in a box resolution but leave the unused part black. Now rotation is just a box nothing lost. I know nothing about this so. Just saying I could sound stupid.

  8. #8
    Join Date
    Jun 2009
    Location
    St. Louis, MO
    Posts
    1,197

    Default

    The first face must disperse the colors somewhat, the refection off the base makes the different colors emerge from the second face somewhat offset, but parallel to the input and thus focus to a common point...
    "There are painters who transform the sun into a yellow spot, but there are others who, with the help of their art and their intelligence, transform a yellow spot into the sun." Pablo Picasso

  9. #9
    Join Date
    Mar 2018
    Posts
    96

    Default

    Quote Originally Posted by laserist View Post
    Very little of the image will pass through the dove prism if it's uncollimated...
    What do you mean? Transmittance, cropping, or something else?
    I don't much care for changing the throw ratio of the projection to make it collimated, but not sure what the correct approach for collimating a video projection would be.

    Quote Originally Posted by kecked View Post
    Why does it not also break up the colors?
    I'd imagine because it's symmetrical, but I may be wrong. I think this is also what laserist means.

    Quote Originally Posted by kecked View Post
    What if render the video out in a box resolution but leave the unused part black.
    You can definitely do that but for a 16:9 aspect ratio projection you will lose quite a number of pixels. Same with cropping a 16:9 projection into a 9:16. You can do in software but you lose pixels.
    You can also use two vertically stacked projectors like laserist suggested but then you need two projectors which adds to the cost and processing. This works by having a intersection area between the projections and adding a gradient fade mask to both video sources so they will blend together. It's called "edge blending".
    Last edited by shoujin; 03-12-2018 at 22:45.

  10. #10
    Join Date
    Sep 2006
    Location
    Netherlands
    Posts
    1,435

    Default

    Dove-prism are fairly common because they're used in microscopes and binoculars, but I've never seen them in sizes that could accommodate the lightpad of a beamer. Doing so will require extra telescoping optics, aberrations and losses. Check surplusshed.com for good deals. Make sure you pick dove, porro is useless.

    You can make a contraption similar to the dove-prism using 3 (first surface) mirrors.
    Mirrors will have the advantage of no chromatic distortion.
    Downside will be the increased physical size.

    If you make a box or tube that folds the lightpad hitting 3 (or any odd nummer) of mirrors, the image will rotate 2 times the angle of the box.
    (Given that the lightpad of the output is coaxial to the input.)

Posting Permissions

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