[daip] [Mnj] Calibration still in 31DEC13

[Chris.Phillips at csiro.au]

Thanks Eric

Your point about different phases is a good one - I will get him to look that. 

Maybe the doco needs a warning about that - I suspect that could be a common user failure mode

Regards
Chris

On 22/05/2013, at 9:26 AM, Eric Greisen wrote:

> Chris.Phillips at csiro.au wrote:
>> Hi Eric
>> I suspect this is not related to the current calibration errors, but I a student I have been helping with LBA data has hit what looks to me like a bug in SNCOP.
>> In brief he is trying to copy FRING solutions from a continuum data set with 2 IFs to a spectral line data with 1 IF. When the BOX parameter was blank it was changing the amplitude correction but when he explicitly set BOX is seemed to work OK.
>> If you think  this could be related to the other calibration problems I will get him to try and run MNJ. Otherwise is the best approach to send a more detailed explanation to the aips bug email?
>> Regards
>> Chris
> 
> I have tried SNCOP and it appears to work as it should with BOX=0.
> In my case that averaged up 4 IFs which was not a good thing because the 4 had different phase even though they all had amplitude 1.0.  Averaging up the reals and the imaginaries produced an amplitude near zero. Nonetheless it was the correct average.  I suspect your student hit much the same issue.
> 
> And you are right that this is unrelated - the bug was in fetching values to implement DOCAL=1 not related to tasks most people have never heard of.
> 
> Eric Greisen
>