Multi-camera Rig Makes Trees Say Cheese


[Nick Nichols] set out to take a photograph of a full redwood tree. Here’s the catch, these redwoods are over 300 feet tall and they’re not just standing in the middle of a vast desert. If the photo is taken from a distance, you will only capture the top part of these majestic beauties. How can you take the shot from close up? Build a custom rig to take multiple shots and stitch them together for a composite photo.

[Nichols] built the rig to hold three cameras focused to the left, middle, and right of the tree. The frame includes a gyroscope to keep the cameras steady. By lowering the cameras from the top to the bottom of the redwood they were able to capture 84 pictures to assemble the final shot. The result is featured in the October edition of National Geographic magazine. We’ve included the final picture and embedded a video of the rig in action after the break.



[digg=][via NPR]

[Thanks Kent]

42 thoughts on “Multi-camera Rig Makes Trees Say Cheese

  1. @Caleb Kraft: Hang on a second–they wouldn’t be printing the picture on paper if people weren’t buying the magazines.

    You can’t buy a magazine and then complain about how much paper is being used for actual content!

  2. a lot of paper is made from trees in tree farms. The more paper you use the more trees they have to plant. it not like there is a grizzled ax wielding guy running around an old growth forest singing “eff you trees, I need a memo pad” lumber on the other hand may be a different story.

  3. the final image was stitched together using Microsoft ICE.

    As the site suggests, its from microsoft’s research dept.. tech used in programs such as photosynth.. but photosynth isn’t really about creating panoramic/stitched images, you’d be looking more along the lines of windows live photogallery for something intergrated for that use.

    Microsoft’s ICE is dedicated technology to stitch images together.

  4. Am I the only one that was bothered by the stitching of the foreground trees? I think a better solution than a 3 camera rig and stitching would be to use the sensor from a scanner and put a lense in front of it. Then fly that rig up while “snapping the photo”. I think I’ve seen a project like that here before.

  5. @GeneralFualt & lolerskates: The ‘crappy stitching’ is an artifact that arises from the different perspective of each camera and the relative distance to the tree when the images were shot. To do this correctly, you’d need one camera that rotates around the center of the *first* lens element. This minimizes stitch error. (I am by no means defending Micro$oft either – photo stitching software is nothing new)

  6. @bloke_from_ohio yes too true. i grew up near wausau paper. in some areas many square miles of replanted trees. very dull, no life but linear lines of trees for as far as you can see. truly renewable though…

  7. @medix
    What I was proposing was using a setup that only captures a single row of pixels. Then “stitch” those rows together into one long photograph. That would remove the perspective problem.
    Using the element from a bed scanner and a camera lense should do the trick. There was a project here some time ago that used this technique to take some very strange and interesting photos.

  8. @lolerskates

    Agreed on the crappy M$ technology. They should have used Apple’s image stitching QTVR I have on my Mac I find it much better. They actually probably did, but I bet M$ paid them to say this.

  9. Besides the parallax issues, the camera platform descended from the canopy on a single rope, which was subject to wind and tree sway. Thus the tree to background perspective shifted slightly through the shoot, causing insurmountable problems for the numerous auto stitching programs that were tested. The final redwood tree was stitched together in Photoshop—it took over 120 hours.

  10. Geez, some of you twats are so zealous over the use of a MS tool (god forbid!) that you can’t just enjoy and comment on what is a pretty cool image? Here’s a clue … the perspective shift would have affected ANY stitching tool, even your beloved Apple tools. Get over yourselves. Maybe get out of the basement and go look at a real tree.

  11. ricz – you don’t find the QT tool much better. You just like it because it is Apple. Admit it, you have never even used the MS tool and don’t have a clue what you are talking about. What a twat.

  12. Very nice way to get a high rez shot. Would love to see this via Silverlight’s Deepzoom so we could dive into the detail that should be in this shot.
    And ICE is a very-very good stitching technology notably better than anything Adobe has though I can’t speak to Apple’s tech as I haven’t used it. If you are using Windows Live Photo Gallery, make sure to go to add-ons and download the ICE stitching tech. What Live Gallery uses by defualt is very good (same tech as Adobe)but no where near as good and flexible as ICE.

  13. Great photo, but it has been done before by James Balog, during his pioneering “Trees” project from 1998-2005. He shot the same tree, same perspective, and even used the same tree rigging team. The only difference is that instead of using a mechanical device, James climbed the rope, over 300 feet into the canopy, and shot the images himself as he lowered down the tree.

    The National Geographic team here was just remaking an original and unique photograph of James’. See and to see the true story, and all the other original work that will likely one day be ripped off by the big yellow box.

  14. Adam wrote that Balog already did this, and with this tree, more or less.

    Would sure like to see a link to this tree by Balog if you happen to know of his photo online.

    I’ve seen Balog’s big aggregate image of the Stratosphere Giant, and what looks like a photostitch of Del Norte titan, but don’t recall seeing this tree. Although I’m pretty certain I read about Balog being at this tree that Nichols did.


Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.