Loading ...
Sorry, an error occurred while loading the content.
 

Re: [ccd-newastro] Digest Number 2455

Expand Messages
  • Roger Hamlett
    ... There are a number of costs to the colour camera. The first is in resolution. The CCD, is covered with a matrix of colour filters, something like: R G R
    Message 1 of 2 , Mar 1, 2005
      > HI Ron,
      >
      > Thanks for that explanation but still I miss the advantage of
      > getting a
      > color camera or a monochrome camera with a filter wheel.
      >
      > What is the advantage of the monochrome ccd camera against the one
      > shot
      > color ccd camera ?
      There are a number of 'costs' to the colour camera. The first is in
      resolution. The CCD, is covered with a matrix of colour filters, something
      like:
      R G R G R G.....
      G B G B G B.....
      R G R G R G...
      Now to get a 'luminance' value, you take the whole group of four pixels
      (first and second from the top row then first and second from the next), to
      generate your luminance. For the next 'virtual pixel', you move forward by
      one, and take the second and third from each row. For the next line, you
      start with the first and second from the second row, and combine this with
      the first and second from the third row. This works fine for signal details
      that are gently changing, but has problems when dealing with bright colour
      changes at high spatial frequencies. If (as a really 'worst case' example),
      you had a grid of white dots, that hit the first, and then the third pixel
      on the top line (happening to line up perfectly with the matrix), the
      system cannot distinguish this from a grid of red dots at the same
      frequency!. The matrix gives spatial resolution, that varies according to
      the nature of the source, from the full resolution of the CCD, down to
      about half the resolution of the CCD, and 'averages' about 0.7* the
      resolution.

      The second problem is loss of sensitivity. Because the filters are there
      all the time, you have the light losses associated with them, always
      present. The amount of loss, varies according to the colour of the light
      being recorded, with green light (selected because this is the most
      sensitive colour for the human eye), getting close to 50% of the unfiltered
      cameras sensitivity, but red or blue light, are effectively only 'seen' by
      one pixel in four, and the sensitivity here is about 25%. Now the human
      visual system is much more 'worried' about changes in intensity than
      changes in colour. This is why systems like TV, allocate the majority of
      their available bandwidth to the 'mono' image, and then a relatively small
      amount to the colours. The colour information, is effectively a low
      accuracy 'wash', applied over the more detailed mono image. With a mono
      camera, you can take advantage of this, and take longer luminance images,
      using the full sensitivity, and resolution of the unfiltered camera, then
      use less detailed (probably binned) colour images to apply the colours to
      this. You are then making 'better use' of the imaging time, in terms of the
      actual acquisition of data for you to see. So for a given imaging 'time',
      you can actually record more of the important data.

      The third problem is also because the filters are always there. You cannot
      'bin' the colour images, and retain the colour information. If your CCD, is
      relatively oversampling, you are stuck with this (and the poorer signal to
      noise ratio this implies). This is one of the most serious 'costs'
      associated with the one shot cameras.

      The fourth problem, is that you are stuck with the filters supplied.
      Generally, these do not give the sharp cutoff associated with the normal
      astronomical imaging filters, giving in many cases, significant overlap,
      and while the filter selection is designed to give a close approximation to
      what you would 'see', it is not necessarily a good choice for some objects
      with unusual colours. If you elect to us a narrowband filter to give better
      signal to noise performance against the light in the sky, you are then
      stuck with imaging through 'double' filters, and probably only detecting
      light using one pixel in four, at half the spatial resolution. :-(

      So what are 'one shot' cameras good for then?.
      The answer here is the simplicity of not having a filter wheel, and the
      lack of cabling involved, combined with recording all the colour
      information at the same time. They are truly 'great' for planetary imaging,
      and for objects like comets, that move significantly against the sky
      background. They also give pleasing general pictures of most objects, and
      given the lack of extra wiring are a lovely solution if wanting to go
      mobile.
      As a classic 'example', I have a SVX-H9, and also the second CCD, to
      convert it to colour use. For 95% of the time, the camera is used in one of
      two combinations, with the mono CCD, and a solar Ha filter, or with the
      mono CCD, and a colour wheel for general imaging. However when I am going
      on holiday, I replace the CCD with the colour one, and combined with a
      short OAG, it makes about the shortest/simplest little colour imager,
      attached to a C5 OTA (which I think is one of the truly great "ability for
      it's size" optical tubes around), I have a imaging setup, that fits in two
      small toolboxes (plus the tripod), and can be tucked into the boot of the
      car to get images from a better site than my normal one...

      Best Wishes

      > OK, I have compared the response curves available at the SBIG site
      > but I
      > honestly do not see an advantage in this case of the monochrome to
      > the one
      > shot color camera.
      >
      > Am I missing something ?
      >
      > I see an advantage in the time saving. If I take 4 images each one
      LRGB or
      > take one single RGB image, in the time I take 1 with the monochrome
      > I take 4
      > with the single shot.
      >
      > regards Rainer
    • Rainer
      Hi, Thanks to everybody now I am more confused as before :-)) No really, thanks a lot for all those explanations as you have given me now more food for thought
      Message 2 of 2 , Mar 1, 2005
        Hi,

        Thanks to everybody now I am more confused as before :-))

        No really, thanks a lot for all those explanations as you have given
        me now more food for thought :-)) and I can begin to save money for
        my SBIG STL-11000 either Monochrome or Color :-)).

        regards Rainer


        --- In ccd-newastro@yahoogroups.com, "Roger Hamlett" <roger@t...>
        wrote:
        > > HI Ron,
        > >
        > > Thanks for that explanation but still I miss the advantage of
        > > getting a
        > > color camera or a monochrome camera with a filter wheel.
        > >
        > > What is the advantage of the monochrome ccd camera against the one
        > > shot
        > > color ccd camera ?
        > There are a number of 'costs' to the colour camera. The first is in
        > resolution. The CCD, is covered with a matrix of colour filters,
        something
        > like:
        > R G R G R G.....
        > G B G B G B.....
        > R G R G R G...
        > Now to get a 'luminance' value, you take the whole group of four
        pixels
        > (first and second from the top row then first and second from the
        next), to
        > generate your luminance. For the next 'virtual pixel', you move
        forward by
        > one, and take the second and third from each row. For the next
        line, you
        > start with the first and second from the second row, and combine
        this with
        > the first and second from the third row. This works fine for signal
        details
        > that are gently changing, but has problems when dealing with bright
        colour
        > changes at high spatial frequencies. If (as a really 'worst case'
        example),
        > you had a grid of white dots, that hit the first, and then the
        third pixel
        > on the top line (happening to line up perfectly with the matrix),
        the
        > system cannot distinguish this from a grid of red dots at the same
        > frequency!. The matrix gives spatial resolution, that varies
        according to
        > the nature of the source, from the full resolution of the CCD, down
        to
        > about half the resolution of the CCD, and 'averages' about 0.7* the
        > resolution.
        >
        > The second problem is loss of sensitivity. Because the filters are
        there
        > all the time, you have the light losses associated with them,
        always
        > present. The amount of loss, varies according to the colour of the
        light
        > being recorded, with green light (selected because this is the most
        > sensitive colour for the human eye), getting close to 50% of the
        unfiltered
        > cameras sensitivity, but red or blue light, are effectively
        only 'seen' by
        > one pixel in four, and the sensitivity here is about 25%. Now the
        human
        > visual system is much more 'worried' about changes in intensity
        than
        > changes in colour. This is why systems like TV, allocate the
        majority of
        > their available bandwidth to the 'mono' image, and then a
        relatively small
        > amount to the colours. The colour information, is effectively a low
        > accuracy 'wash', applied over the more detailed mono image. With a
        mono
        > camera, you can take advantage of this, and take longer luminance
        images,
        > using the full sensitivity, and resolution of the unfiltered
        camera, then
        > use less detailed (probably binned) colour images to apply the
        colours to
        > this. You are then making 'better use' of the imaging time, in
        terms of the
        > actual acquisition of data for you to see. So for a given
        imaging 'time',
        > you can actually record more of the important data.
        >
        > The third problem is also because the filters are always there. You
        cannot
        > 'bin' the colour images, and retain the colour information. If your
        CCD, is
        > relatively oversampling, you are stuck with this (and the poorer
        signal to
        > noise ratio this implies). This is one of the most serious 'costs'
        > associated with the one shot cameras.
        >
        > The fourth problem, is that you are stuck with the filters
        supplied.
        > Generally, these do not give the sharp cutoff associated with the
        normal
        > astronomical imaging filters, giving in many cases, significant
        overlap,
        > and while the filter selection is designed to give a close
        approximation to
        > what you would 'see', it is not necessarily a good choice for some
        objects
        > with unusual colours. If you elect to us a narrowband filter to
        give better
        > signal to noise performance against the light in the sky, you are
        then
        > stuck with imaging through 'double' filters, and probably only
        detecting
        > light using one pixel in four, at half the spatial resolution. :-(
        >
        > So what are 'one shot' cameras good for then?.
        > The answer here is the simplicity of not having a filter wheel, and
        the
        > lack of cabling involved, combined with recording all the colour
        > information at the same time. They are truly 'great' for planetary
        imaging,
        > and for objects like comets, that move significantly against the
        sky
        > background. They also give pleasing general pictures of most
        objects, and
        > given the lack of extra wiring are a lovely solution if wanting to
        go
        > mobile.
        > As a classic 'example', I have a SVX-H9, and also the second CCD,
        to
        > convert it to colour use. For 95% of the time, the camera is used
        in one of
        > two combinations, with the mono CCD, and a solar Ha filter, or with
        the
        > mono CCD, and a colour wheel for general imaging. However when I am
        going
        > on holiday, I replace the CCD with the colour one, and combined
        with a
        > short OAG, it makes about the shortest/simplest little colour
        imager,
        > attached to a C5 OTA (which I think is one of the truly
        great "ability for
        > it's size" optical tubes around), I have a imaging setup, that fits
        in two
        > small toolboxes (plus the tripod), and can be tucked into the boot
        of the
        > car to get images from a better site than my normal one...
        >
        > Best Wishes
        >
        > > OK, I have compared the response curves available at the SBIG
        site
        > > but I
        > > honestly do not see an advantage in this case of the monochrome to
        > > the one
        > > shot color camera.
        > >
        > > Am I missing something ?
        > >
        > > I see an advantage in the time saving. If I take 4 images each one
        > LRGB or
        > > take one single RGB image, in the time I take 1 with the
        monochrome
        > > I take 4
        > > with the single shot.
        > >
        > > regards Rainer
      Your message has been successfully submitted and would be delivered to recipients shortly.