  XNoise help page



  GENERAL DESCRIPTION

  The XNoise widget computes the noise standard deviation for
  each pixel of a given image.
  The input image is assumed to be the sum or the mean of a
  set of equally exposed frames. The output is a noise array.
  The overall variance of the noise is the sum of two terms:
  - Gaussian noise, due to the following sources:
    1) read-out
    2) dark current
    3) thermal background
    4) sky
  - Photon noise, associated to the 'astronomical' signal.
  Noise source 1) follows a gaussian distribution of 0 mean.
  Noise sources 2), 3) and 4) strictly follow a Poisson
  distribution, but since they may be often considered roughly
  constant over the frame, the Poisson distribution can be
  approximated by a gaussian one, spatially uniform across the
  frame itself. While noise source 1) introduces no bias in the
  recorded signal, sources 2), 3) and 4) have nonzero mean.
  In general the mean dark current level, the overall thermal
  background and the mean sky are subtracted in the pre-
  processing phase, so their contribution to the noise cannot
  be estimated from the residual signal: this is the reason
  why the corresponding noise levels should be supplied by
  the user.
  Optionally the gaussian noise can be estimated automatically
  by means of a secondary widget, named XNoise_StDev. The
  result is generally over-estimated and represents the overall
  effect of noise sources 1)-4) combined with the (photon) noise
  associated to the diffuse emission originated by nebulosities,
  bright stars haloes and diffuse light from faint unresolved
  sources.
  The photon noise on the 'astronomical' signal is considered
  proportional to the square root of the recorded signal in each
  pixel.
  The XNoise widget can be used in either of the following two
  'meaningful' modes:
  - compute the overall noise per pixel, considering both the
    contribution of sources 1)-4) and of photon noise;
  - estimate the gaussian noise only by means of XNoise_StDev:
    this is the suggested mode for very crowded fields when no
    knowledge on sources 1)-4) is available.
    
 
 
  PARAMETERS

  'Evaluate from data':
      Select 'yes' to estimate the gaussian noise standard
      deviation by means of XNoise_StDev.
      The retrieved value is generally over-estimated, especially
      in very crowded fields.

  'Read-out-noise':
      Read-out-noise for a single exposure.

  'Dark current':
      Dark current level for a single exposure. If the Poisson
      noise is considered and the dark current has not been
      subtracted (a rather unlikely situation!), this field should
      be set to 0, because the dark current noise is accounted
      for as Poisson noise.

  'Thermal background':
      Instrumental background for a single eexposure. If the
      Poisson noise is considered and the thermal background
      has not been removed, this field should be set to 0,
      because the thermal noise is accounted for as Poisson noise.

  'Sky':
      Mean sky level for a single eexposure. If the Poisson noise
      is considered and the sky has not been removed, this field
      should be set to 0: the associated noise is accounted for
      as Poisson noise.

  'Consider photon noise':
      Select 'yes' to consider the effects of both gaussian and
      photon noise.

  'Number of exposures':
      Number of equally-exposed frames which have been combined
      to form the input image.

  'Exposures combined by':
      Select 'mean' or 'sum' to specify how the individual exposures
      have been combined. This will affect noise propagation from
      each single frame to the final image.

  'Electrons/ADU':
      Number of electrons per ADU.



  CONTROLS/BUTTONS

  'Compute':
      Press this button to compute the noise array using the current
      set of parameters.

  'Save':
      Save to a FITS file the noise array.

  'Help':
      Display this help page.

  'Exit':
      Quit XNoise.