[daip] BLANK - excluding single-channel peaks

[Eric Greisen]

Linda Watson wrote:
> Dear AIPS Help Desk,
> 
> I'm using the AIPS task BLANK to blank an HI data cube. I'm using the 
> method where you blank a convolved version of your cube to exclude flux < 
> 2-3*noise and then use that convolved cube as a filter to blank your 
> original image (using OPCODE 'SELC' and then 'IN2C').  I would also like 
> to exclude single-channel flux peaks.  Is there a way to only keep pixels 
> with signal that is above a set level in a specified number of consecutive 
> channels?
> 
> It seems like using one of the other OPCODE settings should be the way to 
> go, but my attempts have not been successful.  I've mostly been trying to 
> work with OPCODE 'FLTW' but it has become very clear that I don't 
> understand what AIPS actually blanks when I give it a window size.

The FLTW is complicated.  What I did with my recent analysis was to 
smooth the spatially smoothed cube in frequency as well with XSMTH.
This then reduces the single spiky channels a lot.

Eric Greisen