[Top] [All Lists]

Re: Operating VRI (Radio Interferometry)

Subject: Re: Operating VRI Radio Interferometry
From: Martin Brown <[email protected]>
Date: Fri, 17 Mar 2006 09:23:47 +0000
Newsgroups: sci.astro
W. Watson wrote:
On another thread, we drifted off on to the topic of the Java VRI program that's on the web at <>. That thread is getting a bit long, so I thought I'd pick up on VRI here. I'm interested in completing a single example. The other thread gave me some clues how to proceed, but I thought I'd try this out here (* see note at bottom). This is pretty much an arbitrary and ficticious example based on a limited knowledge of how this works. I learned in the other post that the sequence, Source, FFT, Apply and FFT-1 is basic to the operation. The rest was obtained by playing with this.

There are four panels arranged as:

Image Panel (IP)     |  Antenna Panel (AP)
Transform Panel (TP) |  Coverage Panel (CP)

AP has a green background with a single dish icon
CP has a background with a red point. If the nearby
In (Zoom) button is pressed, it's dashed line.

I created the acronyms for convenience here.

There are various buttons around the panel. Initially, there are three buttons to the side of each panel that are common: In (Zoom), Out and Reset. As I proceed with the example as below, a FFT and FFT-1 (inverse) will appear. Above the four panels are fill-in boxes for the source and antenna. Below the panels are fill-in boxes for the source, HA interval, frequency and bandwidth of the source, and source declination.

1. Select source as Wide Double, two point objects appear in the IP
2. Set the Lat to 45.0. Press Enter to make sure it's set.
   Use the slider bars at the bottom to set
   the Dec to 0.0 and the HAs to -5.1 (top slider) and +5.1. You may
   not get +/- 5.1, so do what you can to be close.
3. Drag the antenna icon four times to form a square with
   one icon in the center. Make each side about 1".
     A     A
     A     A

OK but this won't give particularly good u-v coverage. As a nice simple example to get started use the classic E-W equally spaced dish configuration. It is much easier to understand.

A  A  A  A  A

Then when the Earth rotates for 12 hours you get ellipses depending on the the declination of the object (perfect circles for object at the pole).

Another cute configuration for small numbers of antennae is a Golumb ruler:

A   A        A     A
  1      3       2

This gives a set of unique baseline lengths 1,2,3,4,5,6

   Press the Plot button and some red horizontal lines will appear
   in the CP. The FFT button appears on the left. When this button
   appears seems to vary. It may appear in an earlier step.
4. Press the FFT button and vertical grey lines appear in the FP
   with a white background, and the FFT-1 button appears on the left.
5. Press the Apply button below the CP. if The Ampl choice is set
   below the TP, then a group of dashed horizontal lines appear in the FP.
6. Press FFT-1. Two vertical rows appear of dots (circles?) in the IP.

So what have I got in the IP? Can someone offer a more meaningful example?

The transform of on instantaneous set of data. To get a meaningful image you need to set the start and end hour angles of the observation to be -6h to +6h (any 12 hour period will do provided the source doesn't set).

After changing this press "Reset" and you should see a nice set of red ellipses in the U-V coverage domain.

* I find VRT sometimes rather erratic (inconsistent) in its behavior. Sometimes buttons like FFT and FFT-1 just appear at almost a whim.

They should appear when there is something relevant for them to act upon. I find the naming of the "Reset" button to make observations less than intuitive. In a way it is a shame that the Image Panel is reused for the output synthesised image since it prevents a direct comparison.

I'm forced to start again because of some missed item, I reload the page. Proceeding again often does not get quite the same responses. The behavior of this program is somewhat inconsistent. If one sets up the antennas as above, but uses the sliders before setting them, then the extreme values (5.1 vs 4.9 hrs) will be slightly different than if done afterwards. The initial default value for lat ( changes in the number of decimal points once you select Custom. I sometimes wonder about the effectiveness of the Reset button. Sometimes it doesn't seem to help recover from a previous bad entry.

It worked as I expect when I tried it before, but then I probably do what it is expecting (quite possible that it doesn't recover gracefully from some unexpected or inconsistent initial states I suppose).

Just checked again. Hmmm... something has gone wrong with it!

The Hour Angle settings offered here currently range from 0h to 0h for both start and end of observations. Unless you can get it to observe for a few hours to generate tracks in the U-V plane it will always give unsatisfactory results in the synthesised image.

The mirror at Jodrell Bank that I use is still working OK.
Try that instead :

Martin Brown

<Prev in Thread] Current Thread [Next in Thread>