[daip] sum individual channels

[Eric Greisen]

Jose-Maria Torrelles wrote:
> Dear Eric, I am very sorry to ask you directly, but you know within AIPS 
> any form (TASK) to SUM
> a number of individual images of spectral channels, but taking into 
> account the individual rms of each
> channel image?
> 
> I have VLBA H2O maser data (512 channels). Each channel has different 
> rms because
> we are limited by dynamic range. I would like to make
> a linear sum of all the channels, but first estimating the rms of each 
> individual channel and then
> considering the signal of that channel with a S/N ratio higher than a 
> particular value. Tasks like SUMIM, SQASH, or MOMNT cant manage the data 
> in that way (with MOMNT is possible to fix an icut value, but being the 
> same to all the channels).
> 
> Thank you very much in advance,
> 
> Jose
> 
We don't have anything like this directly.  You could split out each 
image as a separate plane.  Then run a procedure which does IMEAN - to 
return the rms as pixstd, followed by BLANK opcode 'SELC' using the 
value of PIXSTD.  Finally MCUBE and SUMIM.  You might have to force the 
blanked pixels to be 0.0 rather than magic blanks (might be available in 
BLANK or with REMAG) so that the sums ignore the blanks.

Eric Greisen