I've imaged an extragalactic supernova with ATCA, and although I don't expect to resolve it the image shows a source that looks extended in one axis at least, as confirmed by imfit:
Object sn1978k
RMS residual is 5.41E-05 (theoretical image noise is 2.11E-04)
Using the following beam parameters when
deconvolving and converting to integrated flux
Beam Major, minor axes (arcsec): 1.92 1.63
Beam Position angle (degrees): 66.3
Scaling error estimates by 4.7 to account for
noise correlation between pixels
Source 1, Object type: gaussian
Peak value: 9.3272E-04 +/- 7.2344E-05
Total integrated flux: 2.3485E-03 +/- 2.8603E-04
Offset Position (arcsec): -1.278 -0.867
Positional errors (arcsec): 0.158 0.055
Pos error ellipse (arcsec): 0.159 0.055 87.75
Right Ascension: 03:17:38.786
Declination: -66:33:03.867
Major axis (arcsec): 4.719 +/- 0.422
Minor axis (arcsec): 1.673 +/- 0.135
Position angle (degrees): 88.09 +/- 2.82
Deconvolution appears to produce a point source
-------------------------------------------------
How come the fit results above confirm that the source is extended along the major axis cf. the beam, and yet still states "Deconvolution appears to produce a point source"?
Cheers,
Stuart
When is a point source not a point source?
Moderator: Mark.Wieringa
-
Mark.Wieringa
- ATCA Expert
- Posts: 297
- Joined: Mon Feb 08, 2010 1:37 pm
Re: When is a point source not a point source?
Hi Stuart,
I had a look at the code, the message means deconvolving the gaussian source fit with the beam fit failed (the minor axis was <0), so the source maybe narrower than the beam in one direction and wider in the other. There are two failure modes like this, your source was close enough to the beam size in the minor axis that it printed the 'point source' message. If the source is even narrower it will just say 'deconvolution failed'.
This could be due to insufficient signal to noise ratio or residual calibration errors affecting the clean.
Some things to try: change the robust parameter in invert to change the resolution a bit, if you think you have enough signal and the image supports it: fit the source with 2 point components.
Cheers,
Mark
I had a look at the code, the message means deconvolving the gaussian source fit with the beam fit failed (the minor axis was <0), so the source maybe narrower than the beam in one direction and wider in the other. There are two failure modes like this, your source was close enough to the beam size in the minor axis that it printed the 'point source' message. If the source is even narrower it will just say 'deconvolution failed'.
This could be due to insufficient signal to noise ratio or residual calibration errors affecting the clean.
Some things to try: change the robust parameter in invert to change the resolution a bit, if you think you have enough signal and the image supports it: fit the source with 2 point components.
Cheers,
Mark