Welcome to the Bruns-cam 3D live page! Take some aspirin, put on your 3D glasses, and check it out. I'm not always here, so if you don't see the picture changing, check back later.
The local time is:
This page does NOT update automatically. The following links do image updates every two minutes, when active. You may get some idea of whether the image is currently updating by comparing the local time (above) to the time on the image. The REMOTE WINDOW creates a separate, automatically updated window, so that you may keep browsing while keeping an eye (well, two eyes) on me.
*NOTE* : These pictures are in 3D. The anaglyph images require special
glasses (see below) for full effect, and the side by side images require a certain
talent. The interlaced images require LCD shutter glasses.
Changed wall-eye and cross-eye images to be JPS format. Notice the hummingbird feeder.
After a long hiatus, I have restarted the Bruns-cam3D. I now have a computer with two parallel ports running the whole thing, which is much simpler than the network handshaking I had to do between the two computers before. Next, I want to make "JPS" images as another alternative. Are these just jpegs of side-by-side stereo?
A new interlaced format has been added. I recently got a nice pair of VR-Joy shutter glasses which support interlaced modes without need of software. Thus for the first time I have 3D with Linux! My other numerous pairs of shutter glasses require driver software, and are thus greatly limited in what you can easily do.
I finally fixed the flaky cable problem, so in principle the cam can be up for much longer stretches of time. One of the quickcams has a flaky connector to the keyboard power. I finally broke down and actually clipped off the connector, and manually spliced the wire onto one of my quickcam extension cables, from which I had clipped the complementary connector. The cam and the extension cable are now married for life.
Our server has been 'upgraded', so things may be flaky for a while.
I remade the image capturing programs, this time using Patrick Reynolds' cqcam-0.40b programs as a basis. This has fixed a problem I was having with garbage at the beginning of the image for long exposures. Also the exposure and color balance is corrected very nicely in Patrick's programs.
I have finally succeeded in getting everything to work together. The cameras take synchronized images, combine them into an anaglyph, and upload them to the server every two minutes. This won't be on all the time, and there are still several technical issues I want to improve.
These are full-color anaglyph images which may be viewed with red/blue 3D stereo glasses. Such glasses may be obtained from Dan Shelley or from Reel 3-D Enterprises. Be advised that the RED lens should be on your LEFT eye. If you send me a very nice e-mail, with your mailing address, I will mail you a free pair of glasses (limited time offer). The ones I send out are made of paper, but the filters are actually BETTER than those in the fancy looking ones pictured above.
If you would like to see other stereo viewing formats made available on this page, just give me enough information to provide them, and I will.
I know that anaglyph is a somewhat annoying 3D format; but it is accessible to a wide audience. Also, I find that since I began working on this page, I like the format much better. At first I could not see the colors correctly in the images, with the glasses on. But now my brain has learned to directly perceive all of the colors in the scene, even though color information must be combined from both eyes to get this. Therefore some processing of color information must be occuring in the brain not only postretinally, but after stereo convergence as well.
Dozens of 3D enthusiasts around the world have been making digital 3D anaglyph images for many years. Traditionally this would involve importing an pair of images with a digital camera or scanner, and then processing the images by hand with PhotoShop or other graphics program. This technology is well understood and has been very successful. Unfortunately it takes about 20 minutes of focused work to make a single image with this technique.
I realized that in order to make an updating anaglyph web-cam, I would have to automate the entire process. To do this has required that I learn a great deal more about both programming and stereoscopic imaging.
I have two Linux computers, connected by ethernet, each of which is equipped with a color Connectix (now Logitech) QuickCam. I capture the images using the QuickCam drivers that have been developed for LINUX. Thanks to the helpful suggestions of the members of the quickcam-drivers mailing list, I have learned a little network programming and have succeeded in synchronizing the cameras' exposures. I still need to do some more work before I can make nice 3D movies, but I expect that that will be possible as well.
Above is a self-portrait of Bruns-cam 3D, taken in a mirror. The two cameras are mounted side by side with a separation of about 60 mm. The mounting bracket was designed in collaboration with Ulrich Genick, and sits upon a standard photographic tripod.
Above is an image of Bruns-cam 3D looking "inconspicuous".
The images are initially stored as full-color Poskanzer pixmap (PPM) files, captured with a modification of the program cqcam. Jef Poskanzer wrote dozens of utilities (PBMplus) for manipulating such images, so I make use of these. I separate each of the two full color camera images into red, green, and blue channels using ppmtorgb3. I use pnmcut to trim the images to make the alignment between left and right just-so, and to make them exactly the same size. I then combine the red channel from the left eye with the blue and green channels from the right eye using rgb3toppm. Finally, the image is converted to GIF format for general consumption. Naturally, all of these steps are done with a shell script.
My original intention was to make JPEG format images, but it turns out that even a little compression severly degrades the stereo effect. I think that JPEG must make some assumptions about the slop permitted in color perception; assumptions which are violated by this very precise use of the color information. I can make JPEG images which are actually LARGER files than the GIFs, but which look worse. My recommendation if you are making anaglyph images is to AVOID saving them as JPEGs.
I was not entirely satified with the mechanism for making GIF files in the pbmplus package, so I stole some source code from the beautiful shareware program XV, to make a ppm to gif converter.
This web page was last modified on