[daip] uv coordinates calculation in AIPS

[Eric Greisen]

On 09/16/2017 10:58 AM, Feng Gao wrote:
> Dear Eric,
> 
> Hello, this is Feng Gao, I used to work with Jim Braatz on the MCP 
> project in Charlottesville and now I'm working within GRAVITY, the 
> IR-interferometry at MPE.
> 
> Right now we're trying to implement the GRAVITY data (in OIFITS format) 
> into AIPS, to benefit from the imaging capability in AIPS. And I have 
> two questions about how the u and v value are calculated in AIPS:
> 
> 1. I thought, for example, the u value is calculated by multiplying the 
> uu parameter value listed in each random group with the frequency of the 
> visibility data directly. However, I've tried to confirm this on some 
> existing MCP processed data but it doesn't match with what I saw in 
> AIPS. For example, the uu parameter value listed in the 1st row of data 
> in the attached .uvfits is 0.012523099178076431, the frequency of the 
> 1st channel is 21904.00 MHz (derived from the header), which gives the 
> calculated u of 274305964.4, while the u value of that data point from 
> AIPS is around 274305802. So how should I get the u v value exactly as 
> in AIPS? I've attached that data and a screenshot from UVPLT for your 
> reference.
> 
> 2. Will the PSCALn value listed in the header been used in some way, 
> other than in FITTP, as mentioned in the AIPS memo 117 by you? I'm using 
> FITLD to load the data.
> 
> Thank you so much!
> 
> With best regards,

I would not look at the U and v Values in the FITS format file but look 
at them after they have been loaded into aips.  There they are in 
wavelengths at the frequency in the header - note that the header may 
say that  this frequency applies at some channel N not = 1.  Not knowing 
the header information or the UVPLT adverbs I cannot comment further. 
This looks like the difference between u and v at the reference and u 
and v at say channel 1.

Eric Greisen