0925	rabbit rabbit rabbit! Arrive. First task, get corr.g loaded/running
		to show DJ if he stops down. Then, start into the re-gridding
		process!! (or smoothing, or something...)
0945	tweaked the corr.g code - it now only takes from pixel 6 to n-6 from
		both IR and HI data. There was darkening around the edges of
		the IR data (from the smoothing/convolution, no doubt) and
		that made the correlation worse than it should have been.
1015	picked region to attempt re-gridding on...
		calling it region j:
		gal: +101:48:45, -62:05:14  <-- use these!!
		J2000: 00:13:38, -00:42:33
	making j5, j10, j25, j50, for 5x5,10x10, etc pixels
	and in galactic, in case that's easier, j5gal, j10gal, etc
	fits and img formats
	in the /compare2/ folder
	DETAILS:
	going to smooth the SFD maps to match the HI....
1050	So this is really really easy. and is in regrd.g script.
	It worked flawlessly on up to 50x50 images, going to try a 250x250...
	region k: 250x250, centered on (gal) +133:45:26 -69:39:32
1130	k250gal.img/fits for starters... did the 0-moment, and even corr.g on
		this region...
1205	Toney stops down, mentions:
		-Piano - talk to Amy Shelton for a key to the rec center,
			where there is a piano
		-Optical scope - talk to Sue Ann Heather(?) in the visitor
			center, they have some there.
	Off to lunch...
1245	Back from Lunch. DJ said to reproduce everything from the Lockman/
		Condon paper, then we'd start working on GASS data!
	comparing in /compareLC (in /export/home/ninkasi/scratch/sjanowie/)
1355	NOTE: it's important when running regrd() that the reference (image
		#2) be in GALACTIC coords. all Galactic for that regridding.
1505	process/scripts are all written to get everything in place to compare
		HI and IR data. I have the whole datasets from SFD98 as well
		as all the GASS data. The comparisons with the Lockman Condon
		paper all match up, and things look good. Also some summary
		of this match in /compareLC/lockcond.html.
	now, back to reading that Rupen article more...
1530	DJ stopped down. I showed the (relative) agreement between my stuff
		and the Lockman Condon paper, and in general how things were.
	His suggestions:
		-fix html page (done)
		-optimise subvmom.g, by converting to cm^-2 when doin the
			getchunk() command, and saving time. I'm not sure
			if that is possible. If code runs slow, I'll look
			into it later.
		-regrid COMPLETE SFD98 data set to match the GASS data...
			this requires concatting their north and south
			fields! Challenge!
		-"time" things.. how long does this whole process take at
			present? Everything.
1555	He headed back upstairs, and I'm going to start on big regridding!
	Doing a test - if I find the 0-moment of velocities -250 to 0, then
		for 0 to 250, and add those results, will I get the same
		as when I do from -250 to 250?
	NO.
	relatedly, sort of, I'm converting all the GASS sub-cubes 000-014 to
		AIPS files, .img, so that things will be faster in the future,
		much, much, faster... takes a WHILE
	new script: extracti.g, operates on the new img files instead of
		FITS files for obtaining GASS data
	working in /time/ for time test...
	waiting for fits-->img conversations of GASS_WGTMED files...
1700	heading out