 News |
FAQ
|
| Search
|
Home
|
| Getting Started | Documentation | Glish | Learn More | Programming | Contact Us |
|
| Version 1.9 Build 1556 |
|
We think that ideally data loaders for specific observatories and correlators should not be required; data output from various places should use a common format and data content to allow transparent access by the users. The goal of the FITS standard is to establish this. However, with the complexity of VLBI correlators and specific requirements for calibration of certain observations, we fear that such a wish is not realistic, and data loaders for the various correlators (VLBAMkIIIJIVESHEVEK4) will be required. Note that the Canadian S2 and Japanese VSOP correlators will use the VLBA FITS binary table format. Furthermore, the JIVE correlator also intends to use the FITS binary table format. For the ATNF LBA this would be RPFITS (as in ATLOD), MkII correlators produce UVFITS. However, FITS readers for VLBI are not only a matter of data format, but in particular a matter of information content, providing an instrument specific path to convert the data (content) to a data set with extension tables suited for the calibration tasks. At least all the required data loaders should be separate individual tasks, not part of any generalized FITS readers (see section 3.2).
It is part of the AIPS++ consortium requirements that the code can be used for on-line data reduction, e.g. at correlation time; this requires that it will be possible to develop data loaders for internal archive formats at various correlators.
AIPS++ must be aware of any additional data carried in MkIV, VLBA, MkIII, S2 and VSOP format before correlation. Support, in this context, may be no more than maintenance of an awareness of the contents and structure of this data. In principle, these different formats should not have any impact on post-correlation data formats; in practice, we should be aware of additional data which are carried in these or any new formats, since these may become relevant to some aspects of off-line data-processing (e.g. the fixed phase offset between upper and lower sidebands in MkIII and MkIV).
Some correlators (notably MkIII) are known to produce redundant data; to obtain all baselines in a specific experiment multiple correlator passes are necessary and sometimes data on certain baselines is produced more than once. Data readers for such data should allow the user to select a certain correlator pass or store redundant data (see also section 4.7).
Individual correlators will introduce different digital signal processing effects which will need to be corrected; one example of this is the FFT artifact in VLBA data. State count corrections (if not performed by the correlator) are another example. Ideally we think it is the responsibility of the correlators to deliver data in which such specific calibration has been applied. However, the AIPS approach for VLBA data has been to apply this correction on input, by means of the data loader (FITLD). Separate tasks for each correlator, each of which potentially requires correction templates or additional information, will be needed. (AIPS equiv: ACCOR, sections of FITLD).
Similar to CE interferometers, VLBI uses a large number of associated data sets for calibration and data flagging. VLBI is different in the sense that this information is not always availableapplied during correlation and needs to be appended to the data product. Below we list a number of such data sets that possibly need to be read from external sources. Because it can not be anticipated beforehand which calibration method will be used for a certain project (e.g. which ionospheric model), this requires in some sense that the Measurement Set can be flexibly extended.
Examples are: