[Skip to Content]

AAO

General Links
Professional Links
Australian Human Resources Institute Diversity Awards 2013 Finalist

IRIS2 Linearity

The linearity of IRIS2 data depends on which array was used to acquire the data:

Check the FITS header keyword "DETECTOR" and/or the observation date to work out which correction you should apply.


Engineering-grade (Oct 2001 - Jan 2002; June 2005 - Jan 2006)

Different controller wave-forms were used with the Engineering-grade array in 2005, compared with those available back in 2001. While this improved the functionality of the one "bad" quadrant, the actual linearity behaviour of each quadrant also changed markedly between these two regimes. Please contact Stuart Ryder (sdr -@- aao.gov.au) for advice on linearity corrections to data taken with the Engineering-grade array.


Mk1 Science-grade (March 2002 - March 2005)

IRIS2's HgCdTe HAWAII array suffers, like all such detectors, from non-linearity. However that non-linearity is quite small. Data taken in March 2002 over a range of exposure times to determine non-linearity in DRM mode, is summarised in the figure below (which is also available in postscript and PDF formats).

IRIS2 Linearity curve



The fitted curve is a fourth order polynomial, constrained to asymptotically have unit value as the number of raw ADUs approaches zero. The detector has not reached saturation by 44,000 ADU. However, observers are advised to acquire data staying below 35,000 ADU as there are some pixels on the detector which appear to show lower full well.

The polynomial is of the following form (where r are raw counts in ADU)

Linearised Counts (ADU) = r ( 1 - 8.81352977e-12 r2 + 3.72642828e-16 r3)  with an rms of 0.3%

Over the range 0-25,000 ADU the non-linearity is negligible (<0.1%).


Mk2 Science-grade (May 2006 onwards)

The linearity of IRIS2's replacement HgCdTe HAWAII array was tested in a similar fashion to the Mk1 array above, and found to be somewhat more non-linear, as shown in the figure below (also available in Postscript and PDF format).

IRIS2 Mk2 linearity


The fitted curve is a fourth order polynomial, constrained to asymptotically have unit value as the number of raw ADUs approaches zero. The detector reaches saturation between 50,000 and 60,000 ADU. Thus, observers are advised to acquire data staying below 40,000 ADU.

The polynomial is of the following form (where r are raw counts in ADU)

Linearised Counts (ADU) = r ( 1 + 8.555607e-12 r2 + 1.6591852e-16 r3)  with an rms of 0.3%.

A new version of the ORAC-DR primitive _CORRECT_NONLINEARITY_ is available, which will auto-detect which science-grade detector is in use, and apply the appropriate correction. It needs to be installed in /star/bin/oracdr/primitives/imaging/IRIS2/, or in the directory pointed to by the environment variable $ORAC_PRIMITIVE_DIR.


Return to the main IRIS2 page.

These pages contain information on the functionality of the IRIS2 Infrared Imager and Spectrograph. Pages maintained by Stuart Ryder (sdr -@- aao.gov.au) and Paul Dobbie (pdd -@- aao.gov.au). Page last modified by Stuart Ryder.