[bananas] VLBA frequency label error for LSB data

Tony Beasley tbeasley at aoc.nrao.edu
Wed Jul 8 17:37:46 EDT 1998


VLBA Lower-Sideband Frequency Error
------------------------------------

An error has recently been uncovered in the AIPS program FITLD that leads
to lower-sideband (LSB) data being labelled with incorrect frequencies when
read from VLBA FITS format files. This error was introduced in the 15JAN94
release of AIPS, sometime around December 1993.

Beginning with the 15JAN94 release, FITLD reflects any LSB bands and
negates the phases so that the spectra are stored in ascending order of
frequency. In order that the LSB spectra are labelled with the correct
frequencies after being reflected, the reference frequencies for the LSB
bands must be shifted by an amount that depends on the position of the
frequency reference pixel (defined for the entire dataset). This shift was
implemented incorrectly, and did not take the position of the reference
pixel into account. This leads to an error in frequency labelling of

	(1 - 2 * r) * c

for LSB bands, where r is the position of the reference pixel on the
frequency axis for the entire dataset (shown by IMHEADER) and c is the
bandwidth of a spectral channel in the LSB spectrum. The reference pixel is
normally between 0.5 and 1.0 for VLBA correlator data, which means that the
reported frequencies for LSB bands may be up to one channel bandwidth too
low. This may be significant for some spectral line observations.

We emphasize that this problem only affects **lower-sideband** data that
has been loaded from VLBA FITS format data files. This error will also
effect LSB data from the space VLBI correlators in Mitaka and Penticton
loaded with FITLD.

This error has been fixed in the 15OCT98 release of AIPS and a patch is
currently available for 15APR98 from

http://www.cv.nrao.edu/aips/15APR98/patches.html

The AIPS group apologizes for any problems that this error has caused. We
thank the CfA NGC4258 group for bringing this problem to our attention. If
you have any questions please direct them to tbeasley at nrao.edu.

Tony Beasley 
08July1998



More information about the Bananas mailing list