Bandpass table size incorrect - Error in mfboot

Is MIRIAD being a pain? Let us know your experience.

Moderator: Mark.Wieringa

danieljohansson
Posts: 3
Joined: Mon Jan 24, 2011 11:07 pm

Bandpass table size incorrect - Error in mfboot

Post by danieljohansson »

Hi Miriad experts!

I am reducing 7 mm ATCA data taken in November 2010. I have a problem with the flux density bootstrapping in mfboot. When I want to scale the gain of the phase calibrator using the observations of Uranus, I get the following error:

### Fatal error: Bandpass table size incorrect. This happens when the inputs for mfcal and gpcal are inconsistent.

I use the following sequence for calibration after flagging

-- mfcal on bandpass calibrator (which interval to take?)
-- copy bandpass solution to gain calibrator and flux calibrator (with options = nocal)
-- mfcal on gain calibrator (with options = nopassol)
-- mfboot on uranus and gain calibrator (this is where I get the above error). I get a solution for the scaling and spectral slope of uranus.

I would be happy for any input on this. In some reductions with certain flagging the mfboot step worked, but I cannot see a major difference between them...

Thanks,
Daniel
ste616
Site Admin
Posts: 220
Joined: Thu Feb 04, 2010 3:27 pm
Location: Paul Wild Observatory Narrabri NSW

Re: Bandpass table size incorrect - Error in mfboot

Post by ste616 »

Hi Daniel,

The problem here is not that the bandpass table size is incorrect (technically), but rather that there is no gain table on your flux calibrator, since you've used 'options=nocal' when copying from your bandpass calibrator to your flux calibrator.

In the case you've described above, you should do the following:
  • mfcal on bandpass calibrator (you can leave the interval option unset if you like, it shouldn't make a difference in this procedure)
  • copy bandpass solution to the gain calibrator with gpcopy, and no options
  • gpcal on gain calibrator (with options=xyvary,qusolve), or if you prefer, mfcal on gain calibrator (with options=nopassol, and a short interval=0.1)
  • copy calibration solution from your gain calibrator to the flux calibrator with gpcopy, and no options
  • mfboot on Uranus and gain calibrator, which should now work as expected
The most important thing to remember when running mfboot is that all the datasets that you give it must have exactly the same bandpass table, and gain tables that have the same scaling factor.
cheers
Jamie Stevens
ATCA Senior System Scientist
danieljohansson
Posts: 3
Joined: Mon Jan 24, 2011 11:07 pm

Re: Bandpass table size incorrect - Error in mfboot

Post by danieljohansson »

Thanks Jamie, that worked very well. Also explains why this problem seemed to come and go; I was probably not using the options = nocal with gpcopy every time... Thanks a lot for your quick answer.
ehone
Posts: 5
Joined: Sat Jun 07, 2014 1:18 am

Re: Bandpass table size incorrect - Error in mfboot

Post by ehone »

I just ran into this problem and couldn't solve it using the tips listed above. My calibration steps are as follows:
  • mfcal vis=bandpass refant=2 interval=0.2 select="time(good time)"

    gpcopy vis=bandpass out=gain
    gpcopy vis=bandpass out=flux

    gpcal vis=gain options=xyvary,qusolve
    gpcal vis=flux options=xyvary

    mfboot vis=neptune.88910(flux),j1326-5256.88910(gain) select="source(neptune)"
mfBoot: version 1.0 07-Jul-06
Applying bandpass corrections to neptune.88910
Applying gain corrections to neptune.88910
Found data for source neptune
Applying bandpass corrections to j1326-5256.88910
Applying gain corrections to j1326-5256.88910
Scaling the data by 0.701
Adjusting spectral slope by -0.675
### Fatal Error [mfboot]: Bandpass table size is incorrect. This happens when the inputs for mfcal and gpcal are inconsistent​


The flux of the phase calibrator seems to have been scaled by the scaling factor shown but it doesn't appear to be correct. I have also tried the method suggested above where the bandpass and gain tables are copied from the gain calibrator to the flux calibrator but the result was the same.
Can anyone offer any help?
ste616
Site Admin
Posts: 220
Joined: Thu Feb 04, 2010 3:27 pm
Location: Paul Wild Observatory Narrabri NSW

Re: Bandpass table size incorrect - Error in mfboot

Post by ste616 »

You definitely shouldn't be doing gpcal on the flux calibrator when the flux calibrator is a planet. Does this make any difference?
cheers
Jamie Stevens
ATCA Senior System Scientist
ehone
Posts: 5
Joined: Sat Jun 07, 2014 1:18 am

Re: Bandpass table size incorrect - Error in mfboot

Post by ehone »

Thanks for the quick reply.
Went through the steps again without using gpcal on Neptune and got the same error message, again the flux scale has been applied so is it possible that the step has worked and just produced the error message for no reason?
ste616
Site Admin
Posts: 220
Joined: Thu Feb 04, 2010 3:27 pm
Location: Paul Wild Observatory Narrabri NSW

Re: Bandpass table size incorrect - Error in mfboot

Post by ste616 »

Can you try just the following please and report what happens.

rm */bandpass */gains */leakage
mfcal vis=bandpass refant=2 interval=0.2 select="time(good time)"
gpcopy vis=bandpass out=flux
mfboot vis=flux,bandpass select="source(neptune)"
cheers
Jamie Stevens
ATCA Senior System Scientist
ehone
Posts: 5
Joined: Sat Jun 07, 2014 1:18 am

Re: Bandpass table size incorrect - Error in mfboot

Post by ehone »

I just input those commands and the result was the same error message as before.
Really not sure what to make of this.
ste616
Site Admin
Posts: 220
Joined: Thu Feb 04, 2010 3:27 pm
Location: Paul Wild Observatory Narrabri NSW

Re: Bandpass table size incorrect - Error in mfboot

Post by ste616 »

I think we'll need to see your data to debug this further. Could you make it available to us please?
cheers
Jamie Stevens
ATCA Senior System Scientist
and48t
Posts: 2
Joined: Wed Jun 25, 2014 8:20 am

Re: Bandpass table size incorrect - Error in mfboot

Post by and48t »

did this ever get sorted out? I'm having the same issue as Daniel was having and I followed your directions carefully and it's still producing the error. I did everything you suggested:

--mfcal on bandpass calibrator (you can leave the interval option unset if you like, it shouldn't make a difference in this procedure)
--copy bandpass solution to the gain calibrator with gpcopy, and no options
-- mfcal on gain calibrator (with options=nopassol, and a short interval=0.1)
--copy calibration solution from your gain calibrator to the flux calibrator with gpcopy, and no options
--mfboot on Uranus and gain calibrator, which should now work as expected

Is there something I'm missing since I"m working with 3mm data?

Cheers,
Crystal
Post Reply