Slashdot Mirror


"Dream Team" to Create Gigapixel Photo System

neutron_p writes "An eclectic group of artists and scientists that organizers have dubbed the "dream team" of imaging and visualization are gathered at New York University this week to begin to create a photographic system capable of capturing and displaying a gigapixel of visual information in a single image. The first Big Picture Summit, Dec. 8 and 9, is organized by artist-photographer Clifford Ross. Ross says his goal is to bring closer to reality his desire to create a "you are there" photographic experience for those who have not personally witnessed the sublime beauty of natural scenes such as Mt. Sopris in Colorado."

6 of 215 comments (clear)

  1. Large Format film cameras by mrm677 · · Score: 5, Informative

    Why? 8x10 cameras have existed for 100 years. Using modern film and a drum scanner will create a digital image with more than 1Gb of pixel data.

    Even my 4x5 camera yields over 100 megapixels when scanning film with a $300 Epson flatbed.

    1. Re:Large Format film cameras by vought · · Score: 4, Informative

      I've got mod points, but I'm going to contribute to this thread instead of modding. There's a lot of assumption about image quality based on numbers alone, and fortunately I worked and taught in a terrific digital studio for the past two years, so I want to set striaght a couple of things.

      4X5 inch film comprises an imageable area of about 19 square inches with a lens that covers this film adquately.

      19 square inches of high-resolution color film - Fuji Provia or Velvia (limiting myself to color for the sake of this discussion) will capture about 2400 dpi worth of useful image data at 8 bpp. The film can easily be scanned to higher resolutions and higher bit depths, but the main reason to do this on a high-end drum scanner is to avoid having to use software to interpolate the image for extremely large prints that exceed the resolution of the original scan.

      19 square inches * 2400dpi * 8bpp = just under 300MB.

      That's about 100 megapixels, give or take a few percentage points. Cost of film and developing per shot is about $3.50 for color, about $.80 for black and white.

      Scanning a 4X5 inch sheet of film costs about $80.00 at my favorite lab. (I use westcoastimaging.com, even though I'm down the street from Calypso Imaging in Santa Clara. WCI does an incredible job, and nearly everyone on staff is actually a photographer, or married to one.)

      Add $20.00 for FedEx back and forth from the lab, and you've got a 100 megapixel image with some slight imperfections (dust spots, chromatic aberration from some older lenses) for about $105.00 per exposure. At this point, the fun begins; the photographer can use Photoshop or the GIMP to make tonal and contrast changes, attempt to match the chrome, or get really fanciful.

      You could make a 16-bit scan of a photograph that contains super-subtle tonal gradation to ensure against banding in the final print, but since most digital photographic printers like the LightJet and Chromira only print 8-bit files, it's usually a moot point.

      Normally, lower-end scanners have to scan in 16-bit to eliminate noise and increase quality to a point where they can stand close to an 8-bit drum scan from a Tango.

      Without explaining the vagaries of scanning backs, it is possible to directly capture a 100MP image from a conventional 4X5 inch camera - but only if the subject isn't moving. Even the 40-year old shutter in my Schneider 90mm lens can work at 1/500 sec, given sufficient light for the film I have loaded. No "gagapixel" camera can do this yet - not even remotely.

      This whole "gigapixel" push is a scam. After making some 30X40 test prints from the Canon EOS1-D mkII the other day, I can say without question that digital cameras are pushing the boundries of medium format film while remaining under the 30 megapixel benchmark.

      The proof is in the print. In a world where most digital images are posted and viewed on web pages, no one will easily tell the difference between a 30k JPEG that started life as a high-resolution scan and one that started life in a .06MP Apple QuickTake from 1996. A print on paper at a equivalent resolution is the best wayo to test real image quality.

      In this case, megapixel comparisons are moot. Because the characteristics of film and digital are different, you can't accurately compare a scan from film and a digital file of the same size on screen alone.

      The most accurate way to determine the quality of an image is to look at a print. When they reach 30-40 megapixels, with forgettable battery life and no crashes, I may be tempted to give up my view camera for a DSLR, but some features still won't be there (full tilt/swing/shift movements, for one).

      For me, 20 pounds of view camera equipment (using exactly one battery, for my spot meter) is still (and may reamin for several years) the easiest way to capture high-resolution photographs in the field. That's what I like to do with my camera - if my goal was to get quick turnaround studio shots, then I'

  2. Something like these? by TroZ · · Score: 5, Informative
  3. Another Gigapixel camera by saccade.com · · Score: 3, Informative

    Check out Gigapxl.org. The guy creating the cameras for this project is a serious optical genious.

  4. zerg by Lord+Omlette · · Score: 2, Informative

    Maybe a few more details over @ the boingboing coverage of story...

    --
    [o]_O
  5. Re:Why artists? by BWJones · · Score: 2, Informative

    This is very much like a conversation I just had in another discussion group. The issue is that you have certain optical properties of your eye. Namely starting with the density of photoreceptors which are about 10^5 per square mm. You then have to deal with an imaging surface (the retina) at 2.5 cm from the lens revealing an object of 1mm at 25 cm which gets projected to a size of about 0.1mm onto the back of the retina. If one assumes approximately 320 photoreceptors/mm (averaged over the eye), given a perfect lens the ideal resolution would be about 30 pixels/mm at 25cm away from your eye! which gives you an approximate optimal resolution distinguishability of 10 microns (important for tying flies). Given that most folks do not have perfect lenses, we are really looking at about a 4 micron resolution that can be distinguished monochromatically. So, if one backs away from the object in question resolution becomes much less important for overall perception and the huge Apple Cinema display three feet in front of me right now does a pretty good job at rendering a close approximation of reality at 100dpi. In fact, an 8MP image from my Canon camera on the Cinema Display is almost indistinguishable from a picture of the valley below taken from my office window.

    --
    Visit Jonesblog and say hello.