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

low noise cameras

Expand Messages
  • mark_manner_spot_obsrv
    Ron s comments here and in talks about read noise and dark current (or practical absence thereof) in professional ccd cameras encouraged me to investigate a
    Message 1 of 5 , Nov 1, 2012
      Ron's comments here and in talks about read noise and dark current (or practical absence thereof) in professional ccd cameras encouraged me to investigate a couple of brands. So far I have looked at a couple of Andor and Princeton Instrument models with 2kx2k back-illuminated chips, and multi-stage air cooling. At this point, the read noise for cameras in the $50k price range seem to be 2.5-3.5 e/pix. For many of the 'usual' amateur cameras, the read noise is in the 9-13 e/pix range. Dark current is more or less neglible at the cooling ranges these cameras are capable of. What I haven't seen in this price range is read noise of ~ < 1 e/pix, however, which I believe Ron may have mentioned is the range of one of his cameras. My question for Ron is how significant in practice is a 3x reduction in read noise vs. 10x? There are cameras in the $30k range that have similar back-illuminated ccds, but read noise in the usual 9 e/pix range.
      Mark
    • yahoo@wodaski.com
      It is our EMCCD camera (Princeton Instruments PhotonMax) that has
      Message 2 of 5 , Nov 1, 2012
        It is our EMCCD camera (Princeton Instruments PhotonMax) that has < 1e- read noise.

        Ron Wodaski



        On Nov 1, 2012, at 1:31 PM, mark_manner_spot_obsrv <mark.manner@...> wrote:

        > Ron's comments here and in talks about read noise and dark current (or practical absence thereof) in professional ccd cameras encouraged me to investigate a couple of brands. So far I have looked at a couple of Andor and Princeton Instrument models with 2kx2k back-illuminated chips, and multi-stage air cooling. At this point, the read noise for cameras in the $50k price range seem to be 2.5-3.5 e/pix. For many of the 'usual' amateur cameras, the read noise is in the 9-13 e/pix range. Dark current is more or less neglible at the cooling ranges these cameras are capable of. What I haven't seen in this price range is read noise of ~ < 1 e/pix, however, which I believe Ron may have mentioned is the range of one of his cameras. My question for Ron is how significant in practice is a 3x reduction in read noise vs. 10x? There are cameras in the $30k range that have similar back-illuminated ccds, but read noise in the usual 9 e/pix range.
        > Mark
        >
        >
        >
        > ------------------------------------
        >
        >
        >
        >
      • mark_manner_spot_obsrv
        Thanks for the reference Ron. I looked it up, and see they have a 512x512 and 1024x1024 version. Mark
        Message 3 of 5 , Nov 2, 2012
          Thanks for the reference Ron. I looked it up, and see they have a 512x512 and 1024x1024 version.
          Mark

          --- In ccd-newastro@yahoogroups.com, yahoo@... wrote:
          >
          > It is our EMCCD camera (Princeton Instruments PhotonMax) that has < 1e- read noise.
          >
          > Ron Wodaski
          >
          >
          >
          > On Nov 1, 2012, at 1:31 PM, mark_manner_spot_obsrv <mark.manner@...> wrote:
          >
          > > Ron's comments here and in talks about read noise and dark current (or practical absence thereof) in professional ccd cameras encouraged me to investigate a couple of brands. So far I have looked at a couple of Andor and Princeton Instrument models with 2kx2k back-illuminated chips, and multi-stage air cooling. At this point, the read noise for cameras in the $50k price range seem to be 2.5-3.5 e/pix. For many of the 'usual' amateur cameras, the read noise is in the 9-13 e/pix range. Dark current is more or less neglible at the cooling ranges these cameras are capable of. What I haven't seen in this price range is read noise of ~ < 1 e/pix, however, which I believe Ron may have mentioned is the range of one of his cameras. My question for Ron is how significant in practice is a 3x reduction in read noise vs. 10x? There are cameras in the $30k range that have similar back-illuminated ccds, but read noise in the usual 9 e/pix range.
          > > Mark
          > >
          > >
          > >
          > > ------------------------------------
          > >
          > >
          > >
          > >
          >
        • Stan
          ICCD and EMCCD cameras have extremely low or nonexistent read noise. ICCD uses a photon intensifier coupled to a normal CCD and has essentially zero read
          Message 4 of 5 , Nov 2, 2012
            ICCD and EMCCD cameras have extremely low or nonexistent read noise.

            ICCD uses a photon intensifier coupled to a normal CCD and has essentially zero read noise. But ICCD has some issues with QE (most are below 20% but there are a few very expensive ones in the 50% range) and MTF is limited by intensifier micro channels. I assembled my own ICCD (ZeroCam):

            http://www.stanmooreastro.com/ZeroCam.htm

            EMCCD achieves a similar effect via pixel level electronic intensification (not photonic). EMCCD has generally high QE that is tempered somewhat by stochastic noise (intensification noise). And CIC is a sort-of low level effective read noise but is less than 1e-. Because the intensification is pixel based the MTF is cleaner than ICCD but EMCCDs suffer from trapped charge lag effects that result in faint vertical smears (actually every CCD suffers from this but it is completely hidden by common read noises). I recently obtained a Photometrics EMCCD but have not posted web comments yet.

            When operating in zero noise mode, ICCD and EMCCD have extremely shallow photon well depths of only a few dozen or so, which means that in most cases the exposures must be extremely short, which in turn means that the readout/download times must be very fast to avoid losing photons during the download cycle. To avoid lost photons, the camera should operate in frame transfer mode so that the next exposure is taking place while the prior exp is downloading.

            All of this needs to happen very quickly because exp saturation can occur within a fraction of a second. Thus most ICCD and EMCCD have low pixel counts (by today's standards) to accomodate bandwidth limitations, typically 1x1k (1 megapix). HD+ video technology will increase that size but if you like a huge image then this technology is not for you.

            Such a camera will produce enormous amounts of data requiring software that is not commercially available. I write my own software for these cameras and have created stacks made from nearly a million frames. Try doing that in CCDStack or PixInsight or Registax! <g>

            High speed intensified astro-imaging is a whole new ball game and is on an exciting edge of amateur endeavors. I know of only a few pioneering practitioners. If you are interested then contract me for more info. BTW, I will probably sell my ZeroCam because I simply do not have the time to operate both cameras.

            Stan
          • mark_manner_spot_obsrv
            Thanks for the information Stan. I have followed your zero-cam experiments. My reluctance to experiment with the zero-cam is both not enough time and my
            Message 5 of 5 , Nov 3, 2012
              Thanks for the information Stan. I have followed your zero-cam experiments. My reluctance to experiment with the zero-cam is both not enough time and my inability to write software to deal with the data.
              Mark

              --- In ccd-newastro@yahoogroups.com, "Stan" <stan_ccd@...> wrote:
              >
              > ICCD and EMCCD cameras have extremely low or nonexistent read noise.
              >
              > ICCD uses a photon intensifier coupled to a normal CCD and has essentially zero read noise. But ICCD has some issues with QE (most are below 20% but there are a few very expensive ones in the 50% range) and MTF is limited by intensifier micro channels. I assembled my own ICCD (ZeroCam):
              >
              > http://www.stanmooreastro.com/ZeroCam.htm
              >
              > EMCCD achieves a similar effect via pixel level electronic intensification (not photonic). EMCCD has generally high QE that is tempered somewhat by stochastic noise (intensification noise). And CIC is a sort-of low level effective read noise but is less than 1e-. Because the intensification is pixel based the MTF is cleaner than ICCD but EMCCDs suffer from trapped charge lag effects that result in faint vertical smears (actually every CCD suffers from this but it is completely hidden by common read noises). I recently obtained a Photometrics EMCCD but have not posted web comments yet.
              >
              > When operating in zero noise mode, ICCD and EMCCD have extremely shallow photon well depths of only a few dozen or so, which means that in most cases the exposures must be extremely short, which in turn means that the readout/download times must be very fast to avoid losing photons during the download cycle. To avoid lost photons, the camera should operate in frame transfer mode so that the next exposure is taking place while the prior exp is downloading.
              >
              > All of this needs to happen very quickly because exp saturation can occur within a fraction of a second. Thus most ICCD and EMCCD have low pixel counts (by today's standards) to accomodate bandwidth limitations, typically 1x1k (1 megapix). HD+ video technology will increase that size but if you like a huge image then this technology is not for you.
              >
              > Such a camera will produce enormous amounts of data requiring software that is not commercially available. I write my own software for these cameras and have created stacks made from nearly a million frames. Try doing that in CCDStack or PixInsight or Registax! <g>
              >
              > High speed intensified astro-imaging is a whole new ball game and is on an exciting edge of amateur endeavors. I know of only a few pioneering practitioners. If you are interested then contract me for more info. BTW, I will probably sell my ZeroCam because I simply do not have the time to operate both cameras.
              >
              > Stan
              >
            Your message has been successfully submitted and would be delivered to recipients shortly.