[daip] question about multi-freq data

[Hiroko Shinnaga]

Dear Eric,

  Thanks for your quick response.  I'm sorry for my short words.
My data is VLA data.  I used 4 IFs for 2 different frequencies.
So, each freqency uses two IFs.  When I do LISTR 'SCAN' on the
CH0 data which was created by FILLM with cparm(7)=0,
I could see those two frequency IDs as followings;

===
Frequency Table summary
FQID IF#      Freq(GHz)      BW(kHz)   Ch.Sep(kHz)  Sideband
   1   1      22.46088125    1562.5001   1171.8750      1
       2      22.41088125    1562.5001   1171.8750      1
   2   1      23.72198972    1562.5001   1171.8750      1
       2      23.69385238    1562.5001   1171.8750      1
===


> The traditional VLA data set has 2 IFs which are usually at two
> similar but different frequencies.  The traditional VLBA data set has
> several of these IFs.  If you mean FREQIDs, as shown by LISTR 'SCAN'
> mode, then the easiest thing to do is to use UVCOP to write two data
> sets each containing only one of the FRQIDs.  This simplifies the rest
> of the reductions.

  Actually I tried UVCOP to make two different data sets by changing
FREQIDs, because I could see two different FREQIDs by LISTR.
But the FREQID 1 which was created by UVCOP seems to be too small,
although FREQID 2 whic was created by the UVCOP seems to be normal.
Those frequencies were observed at the same time so that each data
should have the same length.  I could tell this length difference when I
did
'VLACALIB' on those two data sets.  The vlacalib process of  the
FREQID1 data had been done so quickly, but the process of the FREQID2
took for a while.  So, I suspected maybe I should do 'FILLM'
with cparm(7)=-1 as written in the AIPS cookbook.   Do you recommend
I should do FILLM with cparm(7)=-1 to reduce the data properly ?
Or do you know any other way to get the reduction done correctly
on the data written by FILLM with cparm(7)=0?



> I worry a bit about your reference to CH0 file.  Is this a
> spectral-line data set and are your separate FREQIDs actually very
> close in frequency, differing only because of Earth rotation and minor
> velocity offsets?

  Those are far enough frequencies, more than Earth rotation.

> If so, then they should have been loaded as only
> one FREQID and FILLM was run incorrectly by whoever ran FILLM for you.
> It is essential for users asking the analysts to run FILLM so that
> they may use ftp (rather than paying for tapes) to provide the
> analysts with all necessary FILLM parameters.  They are not scientists
> and do not know the contents of your experiment.

   You are absolutely right.  I have to do that from next time.

   I'd like to appreciate your response.  Thanks a lot.

 -Hiroko