Author Topic: PixInsight 1.6.1 - ImageCalibration: New Pedestal Functionality  (Read 8105 times)

Offline Juan Conejero

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 7111
    • http://pixinsight.com/
ImageCalibration: New Pedestal Functionality

A pedestal is a small quantity added to the result of an image calibration process to ensure positivity of pixel values. Negative pixel values may result from subtraction of overscan regions and bias frames. Normally, negative values are statistically irrelevant and are truncated to zero by default without further consequences. However, there are situations where the amount of zero-truncated pixels is too large in the calibrated result, especially in master calibration frames, and this can be a problem for subsequent image calibration tasks. In these special cases, adding a small pedestal to the result --usually in the range from 50 to 200 ADU-- solves the problem.

The ImageCalibration and ImageIntegration tools that ship with PixInsight 1.6.1 fully support pedestal values. There are two pedestal sections in ImageCalibration: input pedestal and output pedestal.


Output Pedestal

This is a small quantity expressed in the 16-bit unsigned integer range (from 0 to 65535). It is added at the end of the calibration process and its purpose, as noted above, is to prevent negative values that can occur sometimes as a result of overscan and bias subtraction. Under normal conditions, you should need no pedestal to calibrate science frames because mean sky background levels should be sufficiently high as to avoid negative values. A small pedestal can be necessary sometimes to calibrate individual bias and dark frames. The default value is zero.

Input Pedestal

The input pedestal is subtracted from input images as the very first calibration step. This quantity must be expressed in the 16-bit unsigned integer range [0,65535]. If present,  an input pedestal has been added by a previous calibration process to ensure positivity of the result. For example, you might have specified an output pedestal to enforce positivity of previously calibrated master bias, dark and flat frames. The input pedestal is controlled with the following parameters:

Pedestal mode

This parameter specifies how to retrieve input pedestal values. It can be one of the following:

- Literal value lets you specify a pedestal value that will be subtracted from all target and master calibration frames. You usually don't need this option, except in very special cases.

- Default FITS keyword. This is the default mode. When this mode is selected, ImageCalibration will retrieve pedestal values from 'PEDESTAL' FITS keywords, when they are present in the input images. Note that ImageCalibration writes PEDESTAL keywords to calibrated images when a nonzero output pedestal is specified.

- Custom FITS keyword allows you to specify the name of a custom FITS keyword to retrieve pedestal values, when the specified keyword is present in the input images.

Pedestal value

This is a literal pedestal value in the 16-bit unsigned integer range (from 0 to 65535),  when the Literal value option has been selected as pedestal mode.

Pedestal keyword

Custom FITS keyword to retrieve pedestal values in 16-bit DN. This is the name of a FITS keyword that will be used to retrieve input pedestals, if the Custom FITS keyword option has been selected as pedestal mode.


ImageIntegration Pedestal Functionality

The ImageIntegration tool automatically detects pedestal values when they are present in its input images. Pedestals are specified by (actually nonstandard, but de-facto standard) 'PEDESTAL' FITS keywords, which are always written to the headers of calibrated images by ImageCalibration when a nonzero output pedestal is specified.

The only thing that ImageIntegration does when there are PEDESTAL keywords in its input images is propagating the same PEDESTAL keyword to its output integrated image. In this way you can integrate a set of calibrated flat frames with varying pedestals, for example, to build a master flat frame.

Note that for the automatic pedestal detection feature to work, the set of input integration images must specify coherent (i.e. identical) pedestal values; otherwise existing pedestal keywords are ignored. Appropriate informative and warning messages are written to the console so you can always know what's happening.
« Last Edit: 2010 August 16 03:02:48 by Juan Conejero »
Juan Conejero
PixInsight Development Team
http://pixinsight.com/