Recipes also exist for processing engineering data, usually taken with the shutter closed and the telescope stationary. Historically, the real-time sequencer (RTS) was not involved in such measurements which prevented the pipeline from processing these data due to missing FITS headers and other essential metadata.
In September 2010 is was decided that the pipeline be used to process and analyze data taken in engineering mode, and a simulated RTS is used to generate metadata. Support for engineering data is currently quite rudimentary with only a single recipe available.
Note that in order to process data taken in engineering mode, the -eng option must be given to the Orac-dr initialization script.
Measurements of the noise-equivalent power (NEP) are processed with the ENG_NEP recipe. The ‘observation’ consists of a series of NOISE measurements at different pixel heater and detector bias values. The recipe is designed to run offline, after the observation has finished.
The NEP is calculated for each subarray at each setting. These per-bolometer NEP images are combined to form a 4-d hypercube with axes bolometer row, bolometer column, heater setting and bias setting for each subarray. These hypercubes are named sXXYYYYMMDD_MMMMM_nep.sdf where sXX is the subarray label (e.g. s8a).
From these data, the effective and RMS NEP is calculated at each heater and bias setting, stored as a 2-d image with the heater and bias values as the X- and Y axes. These images have suffices of _effnep and _rmsnep respectively.
A number of log files are also written. The first is called log.bolonoise which contains the noise properties for each heater and bias setting and all subarrays. Log files for the ‘mapping speed’ parameter for each subarray are written separately with the name log.mapspeed_SXX where SXX is the subarray (e.g. s4a). The mapping speed is given by and is calculated for the best bolometers between 300 and 1000 in steps of 100 bolometers.
The recipe does not display any data.