Wrong number of planes in cube image

Got an image problem? Let us help!

Moderator: Mark.Wieringa

Post Reply
saracmbr
Posts: 38
Joined: Fri Jan 20, 2012 3:40 pm

Wrong number of planes in cube image

Post by saracmbr »

Hi there,

I'm trying to make cube images out of pre-CABB data. I've got 2 frequencies, one in 1.3GHz and other one in 1.4GHz. They both have 13 channels. The final cube image using both dataset should have 26 channels but it has 13 channels. I Did the same with CABB data but the number of channels are 256 instead of 2048 (I have split my CABB data into 8 frequency chunks). It seems that invert uses the first input UV file. I'm wondering if there is something I'm doing wrong here?

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

Re: Wrong number of planes in cube image

Post by ste616 »

Hi Sara,

Could you please supply the command you used in invert, and a description of each of the datasets you're feeding into it?

Since this is not the intended behaviour, we may also need to have access to some of the data if we have to debug it.
cheers
Jamie Stevens
ATCA Senior System Scientist
saracmbr
Posts: 38
Joined: Fri Jan 20, 2012 3:40 pm

Re: Wrong number of planes in cube image

Post by saracmbr »

Hi
I have 2 sets of data named data_1 at 1.344GHz and data_2 at 1.432GHz. The Invert command I used is as follow:
Inp Invert
vis=data_1,data_2
map=test.qmap
beam-=test.beam
imsize=1000
cell=2
robust=0
stokes=q
robust=0.1

The data_1 has data in all 13 channels but data_2 has no data in channel=10 so I have to use slop to make invert tolerate this.
I also have some CABB datasets with different frequencies and different number of channels. It displays the same behaviour.
when I use invert with the same input parameters but for a single dataset it gives me 13 planes which is true but when I use the command above(including both dataset) it makes the same number of planes (13 planes) which is not right (it should be 26).

I used "uvglue" to glue both uv-files and I used invert again with the same input parameters but on the output of uvglue. This time it worked and it made 26 planes.

Invert: revision 1.16 2012/09/18
I hope this helps to debug.
Cheers,
Sara
ste616
Site Admin
Posts: 220
Joined: Thu Feb 04, 2010 3:27 pm
Location: Paul Wild Observatory Narrabri NSW

Re: Wrong number of planes in cube image

Post by ste616 »

Hi Sara,

This is in fact the proper behaviour. Invert does not stitch together datasets that have different channel frequencies, and so cannot make a 26 channel cube from two 13 channel sets with different channels.

In this case, your use of uvglue is appropriate.
cheers
Jamie Stevens
ATCA Senior System Scientist
saracmbr
Posts: 38
Joined: Fri Jan 20, 2012 3:40 pm

Re: Wrong number of planes in cube image

Post by saracmbr »

Would you please advise what I should do in case of CABB data. there are many channels with no data left. Invert dose not tolerate this thus I have to use slop.
I used slop=0.1 and it made a 2048 channel cube but Kvis cant display this cube complaining about the bad data and maximum range.

Is there any task to simply remove the channels with no data left form uv-files?

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

Re: Wrong number of planes in cube image

Post by ste616 »

Hi Sara,

There is no problem with using slop; for most cubes I would think slop would need to be used.

However, kvis shouldn't have any problems displaying cubes where some of the channels are bad. Could you elaborate a bit on the kvis problem please?
cheers
Jamie Stevens
ATCA Senior System Scientist
Post Reply