Manual despiking simply involves examining the data with linplot and display, identifying bad regions by eye and then running change_quality to turn off the bad points. In general this is very time consuming (especially working out the pixel number of a spike so that change_quality can be told the exact location) so two interactive techniques are available:
This script automates the linplot-change_quality cycle - bolometers can be plotted in turn, with spikes identified and removed, all within a few seconds.
The sclean task allows users to simply click on bad points to remove them. This routine has been designed for SCUBA despiking and therefore understands SCUBA quality. sclean provides an integrated despiking environment showing the 2-D image and a 1-D slice, allowing points to be marked bad (or good) in either window. More information can be found in the FIGARO documentation.
Additionally, the rlinplot command, a wrapper for the KAPPA mlinplot command, and the pltbol command, a wrapper for the KAPPA linplot command, can be used to identify spikes and noisy bolometers rapidly without having knowledge of NDF sections or the specifics of each command.
SURF -- SCUBA User Reduction Facility