EIS continues calibration operation (eg. REGCAL071/REGCAL072) to generate hot and warm pixels maps for EIS_PREP routine use: e.g. to mark the hot/warm pixel positions and smooth the values on those positions.
Before EIS launch, the CCD read-out section was assumed in the middle 512 rows of the CCD. That means EIS_PREP only needs hot/warm pixels maps of middle section of CCDs.
However, in post-launch phase, the top and bottom of the CCDs were introduced to cover whole CCDs area (ie. Y-axis: 0-1023 rows, X-axis: 0-4095 columns). Thus hot/warm pixels maps for EIS CCDs have been generated accordingly as top and bottom maps (in detail: REGCAL072 study for top of CCDs and REGCAL071 study for bottom). The EIS_PREP routine has also changed in order to use top and bottom hot/warm pixels maps.
The EIS camera electronics do not readout CCD row 0 in the y-direction (this is due to how the electronics work, row 0 is always discarded), so for the bottom of the CCDs EIS camera readout from row 1 to row 512 (inclusive), and for the top the camera readout from row 512 to row 1023 (inclusive).
So we can see there is an overlap at the 512th row, ie.:
|Bottom half:||row 1 -- row 512|
|Top Half:||row 512 -- row 1023|
"The overlapping row (512) is there because the MDP requires that data to be a multiple of 8 in both the X and Y directions. Rows 1 to 511 is not a multiple of 8, hence we have to read rows 1 to 512 (inclusive). For the top half we must read 512 for 512 (i.e. rows 512 to 1023), hence the overlap in row 512."
-- Jian Sun (2008-10-17)
The middle 512 rows of the CCDs were used at the beginning of the mission until January 2008 when the change was made to read-out the full CCD height (1024 rows). The hot and warm pixel maps are for the middle 512 rows until 18th January 2008. From the 19th January 2008 onwards the hot and warm pixel maps are for the full CCD height (split into the top 512 rows and bottom 512 rows).
-- Louisa Bradley, 13-Aug-2009
When EIS_PREP is running I've noticed that it has used hot and warm pixel maps from different days, why is this?
Following an update to the EIS reformatter software (Feb 2010) the 'recovered' data for REGCAL071 and REGCAL072 can now be used to provide hot and warm pixel maps. Previous to this change the recovered data could not be trusted as it did not appear in the correct position of the CCD. When a file is recovered it means that some of the data is missing, in the case of the REGCAL071 and REGCAL072 files (which are FULL CCD exposures) typically a CCD half will be missing, i.e. 25% of the data cannot be used. Click here to see images from a recovered REGCAL072 file. The remaining 75% of the data can be used and is processed as normal. When EIS_PREP selects the pixel maps that are closest in time to the science observation, it may be that the closest in time are from a recovered file and as one (or more) of the pixel maps are missing it will then have to search for the next closest in time pixel maps. For example if you have a science study (full CCD image, 1024 CCD pixel height) from 10 May 2010 and you run EIS_PREP it will select the hot and warm pixel maps from 11 May 2010 (closest in time). It will find pixel maps for the short wavelength CCD but not a complete set for the long wavelength CCD. EIS_PREP will then search for the missing pixel map from the next closest in time data which would be 07 May 2010.
-- Louisa Bradley, 19-May-2010
Hot and warm pixel maps for 06 June 2010
The hot and warm pixel maps for 06 June 2010 were replaced on 07 July 2010. The original maps were from rescued data and CCDA left was missing. The new maps are a complete set (the reformatter had a little hiccup on 06 June and some files had to be reformatted again). There is also now maps for the top CCD rows where as previously the maps were for the bottom CCD rows only. If you have run eis_prep on data from 04-09 June 2010 it might be worth re-running it so that eis_prep finds the updated maps.
-- Louisa Bradley, 08-July-2010