[daip] Re: RA plot labels

[Eric Greisen]

Caroline Simpson writes:
 > Hello,
 > 
 > I'm using 31DEC01 on a linux box (RedHat 7.3), and I've noticed 
 > something strange about the RA scale on my KNTR plots. The printed 
 > coordinates don't seem to be quite right... I think. I would think that 
 > if I'm plotting 64 pixels at 12"/pixel, that should be 51.1 seconds in 
 > RA, but if I use the printed RA coordinate labels to calculate the length 
 > of the plotted axis, I get 61.1 seconds. I noticed this because I was 
 > attempting to convert from milli-RA to proper RA labels on a plot (a 
 > position-velocity plot of RA vs. velocity; we transposed the cube and 
 > used XSUM), and my calculated RA values didn't match my other plots 
 > (but do give me an axis length of 51.1s). The RA scale anomaly appears 
 > on every plot I make using normal RA coordinate labels; even regular 
 > momnt maps etc. 
 > 
 > Interestingly enough, the milli-RA labels appear to be "correct" (they 
 > give me a total length of 51.1seconds, as do the arcsec if I'm plotting 
 > a momnt map. 
 > 
 > Am I just being extraordinarily dense about something? 

The relative arc second scales are in just that relative arc seconds.
But the RA on hours/minutes/seconds is in time.  If the object was at
the North pole the full 24 hours of RA might appear in a single image.
The scale is scaled by cos(declination) compared to a pure arc sec
scale, while the increment in the header is a pure arc sec.

Eric Greisen