= Changes/mfcal =

 ||Template                  ||            ||          ||           ||      ||
 ||Item                      || Complexity || Research || Milestone || Time ||
 ||Add time variability      || hard       || substantial    ||  10       ||  20d    ||

== Changes ==
 * Potentially add time variability to the bandpass solution. Similar to gpcal we may want to change the current solution vector of scalars to a 2d spline fit in frequency and optionally time.[[BR]]

== Research ==
 * Based on test observations of strong calibrators at a large range of elevations and time, determine if we need to deal with time variability in the bandpass shape over the 2GHz band. Engineers are suspecting changes may be significant near the edges of the receiver bands.

== Comments ==
 * Suggestion from Bjorn Emonts: It may be more useful to focus on interpolating between multiple scans on a bandpass calibrator. I.e. leave the solution process the same, but allow multiple solutions separated in time and then
 interpolate between them.  He notes that for some HI absorption experiments a bandpass cal is needed every hour to get sufficient spectral DR.

== Resolution ==

The latter option, interpolation between multiple bandpass solutions, has now been implemented. You can specify a bandpass solution interval in mfcal, the programs that read uvdata and apply bandpass solutions will interpolate between them.

[wiki:Changes Changes Home][[BR]]
[/trac/miriad4cabb Wiki Home]