Spitzer Documentation & Tools
IRS Instrument Handbook


5.1.1             CVTI2R4

CVTI2R4 (“Convert 2-byte Integer To 4-byte Real”) is the first pipeline module that operates on FITS science data (pixels and samples, as opposed to the FITS headers only). The units of data samples are DN. The module converts FITS data from 16-bit signed integer to 32-bit real (float) format. All subsequent pipeline modules operate on 32-bit data.  The files cvti2r4.fits and dmask.fits are the output of CVTI2R4. The cvti2r4.fits file is not available in the Spitzer Heritage Archive, but can be created by running CUPID.  The data conversion process is as follows:

1. The 16-bit integer pixels and samples are converted to 32-bit real format.

2. Saturated samples are identified if they have the input value 32767.0 DN.

3. “Missing” data are samples whose values could not be read during telemetry. Usually missing data consist of whole rows in one or more planes of the data cube. The FITS header contains coded information on the location of these rows. Missing data samples have an input placeholder value of -1. The module sets missing data samples to an output value of NaN.

4. A value of 32768.0 DN is added to every sample of each plane, except for “missing data,” in order to set the valid data range from 0 to 65535.0 DN .

5. A dmask.fits cube of 16-bit integer data type is initialized and has bit # 3 set for those samples identified as saturated in the data cube, and bit # 14 for those samples identified as “missing.”