Photoshop & Foveon

Printable View

  • 02-09-2004, 05:10 PM
    kes
    Photoshop & Foveon
    I see that the Foveon sensor will be in a consumer camera by Polaroid. Does Photoshop make an assumption that the image comes from the type of sensor used in 99.9% of the digicams. Or, can it handle images from the Foveon whose pixel can measure all three colors?
  • 02-10-2004, 01:04 PM
    Trevor Ash
    That's an interesting question. I suppose only the Adobe engineers can answer that one.

    At a guess, I'd suppose that none of Adobe's products curently do anything to better support images from the fov sensor.

    I'd also guess that the only software likely to be modified to better support fov sensors would be RAW convertors.

    Finally, I doubt they'll do much because once an image is in RGB it's RGB but I suppose they could take advantage of certain characteristics that fov sensors might exhibit, even after converted to RGB viewing and display devices.

    That's my guess :) What are yuor thoughts?
  • 02-10-2004, 03:05 PM
    kes
    Photoshop & Foveon
    I need to learn what a RAW convertor is. I should find more sources describing the engineering differences between Foveon and all the rest. I'll try to send a note to Foveon.

    I wonder whether the photo wizard I stumbled into on Microsoft Windows can handle an image from Foveon. It handled photos from a borrowed Nikon CP2500 AOK.

    In April, I shall attend a workshop in digital imaging at the local art museum. I shall ask the presenter.
  • 02-10-2004, 04:12 PM
    Sebastian
    Trevor is right, once the image is RGB it's RGB. It wouldn't even make a difference in RAW, except that there is no need to use Bayer interpolation or anything similar.

    RGB is RGB, no matter how it was made, it's all just a pixel in the end.
  • 02-10-2004, 08:50 PM
    Trevor Ash
    Hi Kes,

    Be sure to let us know what you've learned on this subject when you get back from the class. Have fun!
  • 02-10-2004, 09:10 PM
    Sebastian
    Quote:

    Originally Posted by kes
    I need to learn what a RAW convertor is. I should find more sources describing the engineering differences between Foveon and all the rest. I'll try to send a note to Foveon.

    I wonder whether the photo wizard I stumbled into on Microsoft Windows can handle an image from Foveon. It handled photos from a borrowed Nikon CP2500 AOK.

    In April, I shall attend a workshop in digital imaging at the local art museum. I shall ask the presenter.

    Kes, just noticed this...

    The Windows wizard WILL NOT handle the Foveon output from ANY camera that does RAW, fFoveon or otherwise. This has to be done with software specifically designed to handle all the different RAW fiormats out there.

    Once again, it has nothing to do with the sensor, but with the way the file is stored. Once converted, any graphic editing package will handle the image like any other image. The new crop of Foveon cameras can convert into JPEG iinside the camera, which is a format that just about any program can open. The initial crop of Sigma Foveon cam,eras could only save RAW files, conversion to other formats had to be done through their software.

    Again, the Foveon is jsut a chip, how it captures data has nothing to do with whether or not the data can eb edited further on down the line by sofware such as Photoshop.
  • 02-11-2004, 01:30 PM
    kes
    Quote:

    Originally Posted by Sebastian
    Kes, just noticed this...

    The Windows wizard WILL NOT handle the Foveon output from ANY camera that does RAW, fFoveon or otherwise. This has to be done with software specifically designed to handle all the different RAW fiormats out there.

    .

    Sebastian:

    If Foveon's RAW must be converted to a format which Photoshop and other image processing software can manipulate, does the conversion destroy the advantages of the Foveon engineering? Presumably the Bayesian interpolation, which uses neighboring pixels to achieve color, must be an algorithm which is very different from the one used to achieve color from a Foveon pixel, which has three color numbers attached.

    Kes
  • 02-11-2004, 01:44 PM
    Sebastian
    Quote:

    Originally Posted by kes
    Presumably the Bayesian interpolation, which uses neighboring pixels to achieve color, must be an algorithm which is very different from the one used to achieve color from a Foveon pixel, which has three color numbers attached.

    Kes


    Your presumtion is wrong. Every single pixel of every single color has three values, one for red, on green, one blue. Typical CCDs are filtered, so a pixel will read 128 green, but will have to guess the other two values based on the data from the surroudnign pixels, that's the interpolation. The Foveon just directly gives the file all three values. No magic, no engineering trickery, no special software required. The Foveon simply has no algortihm for guessing like regular sensors do. No matter which sensor, they all have to provide three values for each pixel.