Emulation of film speed in DSLRs

Discussion in '35mm Cameras' started by Blipvert, Mar 6, 2004.

  1. Blipvert

    Blipvert Guest

    What is the technical process of emulating film speeds in DSLRs... is
    the CMOS physically charged more (hence more noise) At higher speeds?
    Testing both the Nikon D100 and Canon D60, I've found that it's fairly
    true to film, i.e. the less versatile "100" emulation is cleaner than
    the 800/1000 speeds.
     
    Blipvert, Mar 6, 2004
    #1
    1. Advertisements

  2. Blipvert

    S. Cargo Guest

    Yes.. Increasing ISO is done by increasing the amplification of the
    sensor output.. Noise is introduced in the same manner as you'd
    experience with an audio amplifier..

    If you turn an audio amplifier up full tilt, you'll hear some hiss in the
    speakers if there's no sound. And.. unless you have a very expensive amp,
    the sound will be somewhat distorted at high volume as well.

    You obviously can't hear a video signal, but the 'hiss' and distortion still
    appears at high amplification. These products appear in the video signal.
    As a result, at high amplification, this hiss and distortion yields an image
    that appears grainy.

    The appearance is much like the grain you get with high ISO film.

    The early digital cameras were horrible at more than ISO 200.

    The latest sensors and technology have actually succeeded in letting
    some digital cameras produce better quality output at ISO 800+ than their
    film counterparts. This is seen in some of the current DSLR's
     
    S. Cargo, Mar 6, 2004
    #2
    1. Advertisements

  3. Blipvert

    columbotrek Guest

    They just control the gain in the amplifier before the Analog to Digital
    converter. That is how DSLRs do color balance as well. Another way to
    think of noise is... If you crank up the volume of your stereo without a
    signal source the hiss or hum you hear is the noise. That noise is
    always there but when the input signal is strong you need less amp so
    the sound is cleaner.
     
    columbotrek, Mar 6, 2004
    #3
  4. Blipvert

    Tom Thackrey Guest

    Really? How would that work? Do they adjust the gain for each color
    separatly?

    It seems to me that it would be much easier to do it during or after the
    conversion to RGB.
     
    Tom Thackrey, Mar 6, 2004
    #4
  5. Blipvert

    columbotrek Guest

    There are more than one way to do things. Post capture processing is how
    Photoshop works. By the time it gets to Photoshop though you no longer
    have the option of adjusting things anywhere else. My understanding of
    how color balance works is that the sensors only do shades of gray. They
    approximate color by placing a filter over certain pixels. If you want
    less red then all you have to do is reduce the gain in the red channel.
    If you adjust the color balance in the amplifier stage all you need to
    do then is record what you get. If you did it post capture, you would
    have to apply a mask to each pixel during the capture. More work for
    the CPU.
     
    columbotrek, Mar 6, 2004
    #5
  6. Blipvert

    Tom Thackrey Guest

    I understand how it might work. My question is do you know or are you just
    guessing that it works by using a different gain on each 'color' of pixel?
    Not much work, the post capture computation is trivial.
     
    Tom Thackrey, Mar 6, 2004
    #6
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.