[alma-config] Re: pointing error in UVCON

[Leonia Kogan]

Dear Koh-Ichiro,

Thank you very much for the explanation of the DR and FI you use.

Here I may suppose you use the definition given by Mark Holdaway in his
paper with colloborators.  Mark gave me the reference to the paper
early but I has been too leasy to read it.

Mark, I am sorry.


Now I figure out that so called MYFIDELITY is almoust coincides with 
your DR:

5. MYFIDELITY = (MAX of the model) / (rms of the image-difference).
3. DR = 
  (Max of the reference) / (rms of Residual on the OFF-SOURCE pixels)

The difference is I calculate the rms using all pixels of the residual image,
and you take only OFF-SOURCE pixels. I do not know what is better?

Anyway Your DR should be bigger of MYFIDELITY. Now I will calculate DR 
following the definition you gave me.

>Do you mean the global error is constant error during observation and
>among antennas?

Yes, this error simmulates the possible error in the source coordinates.
I understand that 5" is too big. This is my mistake.

>This model is quite compact!

Yes, 4" of the source is ~10% of FWHM_PB=30".
My opinion (of course I can be wrong) is that the case of bigger ratio of 
source/FWHM_PB is the subject of mosaic, not of the single pointing 
observation.


Regards
Leonia


----- Begin Included Message -----

>From owner-alma-config at kochab.cv.nrao.edu Sun Jun 25 22:09 MDT 2000
Date: Sun, 25 Jun 2000 22:08:42 -0600 (MDT)
From: "[Leonia Kogan]" <lkogan at aoc.nrao.edu>
To: morita at nro.nao.ac.jp
Subject: [alma-config] Re: pointing error in UVCON
Cc: alma-config at zia.aoc.NRAO.EDU
X-Sun-Charset: US-ASCII
Sender: owner-alma-config at kochab.cv.nrao.edu
Precedence: bulk
Content-Type: text
Content-Length: 3521
X-Lines: 109
Status: RO

Dear Koh-Ichiro,

Now I am rather free after my week of the AIPS designating.

You wrote:

>Please let me know DR and FI of images without the pointing error.

I can not do it because I estimate the quality of the image using my own 
criterion. It is just because I do not know the generally used definition 
of the DR and FI.

I carried out the following sequence of the AIPS tasks to estimate the 
fidelity of the restored image:

1. Task UVCON to simmulate the UV data corresponding the given model
2. Task IMAGR to restore the image using CLEAN.
3. Task COMB to create the image-difference of the image-model and 
   the restored image.
4. The verb IMSTA to estimate the rms of the image-difference.

5. MYFIDELITY = (MAX of the model) / (rms of the image-difference).

I used the M51 image as my model. This image was given to me by John Conway.

The result of my simulation with pointing error is given at the table:

__________________________________________________
Pointing error |     0      |  5+2=7" |  5+10=15"|
               |            |         |          |
MYFIDELITY     |   342      |   236   |   103    |
__________________________________________________

The pointing error's first item is the global error simmulating the
possible error of the source position. The pointing error's second item
is the random component of the error with the given rms.

I used the wavelength=1.8mm that gives (for D=12m) FWHM_PB=30".
The model size is ~4".

I may suppose that your right definition of FI will give a lower value
for my model, because the M51 is much more complicate than your
gaussian sum.

I simmulate VLA snapshot with M51 model and got a very ugly image.
Having included 4 hours tracks I got a resonable good image.

The new version of the UVCOn is in the AIPS system now and is available
at the AIPS version of DEC31.

If you want and can use the AIPS you can play with UVCON. 
Anyway I'd like to play with it using your model.

I'd appreciate if you teach me the right (which you use) definition of the FI.

Thank you very much


Leonia




----- Begin Included Message -----

>From owner-alma-config at kochab.cv.nrao.edu Thu Jun 22 10:12 MDT 2000
Date: Fri, 23 Jun 2000 01:15:25 +0900
From: morita <morita at nro.nao.ac.jp>
X-Mailer: Mozilla 4.7 [ja] (Win98; I)
X-Accept-Language: ja
MIME-Version: 1.0
To: "[Leonia Kogan]" <lkogan at zia.aoc.NRAO.EDU>
CC: alma-config at zia.aoc.NRAO.EDU
Subject: [alma-config] Re: pointing error in UVCON
References: <200006200457.WAA05334 at bonito.aoc.nrao.edu>
Content-Transfer-Encoding: 7bit
Sender: owner-alma-config at kochab.cv.nrao.edu
Precedence: bulk
X-Lines: 24
Status: RO

Dear Leonia,

sorry for my slow reply!

> I used the new option of UVCON for one of the doughnut configuration using
> model of M51 given to me by John Conway.
> 
> In this test I could not reproduce the result given at your memo (internet).
> I mean the decreasing of DR and FI in > 6times (for spiral)
> when pointing error is becoming 2.4% of the FWHM_PB.
> 
> In my test the effect of the pointing error is not so dramatic.

Please let me know DR and FI of images without the pointing error.

RMS imaging error due to the pointing error is very small in 
my simulation but the deconvolution error due to limited (u,v)
coverage is also very small.  DR without the pointing error
is larger than 10,000.  In such case, the imaging error due to
the pointing error is comparable or larger than that due to 
the deconvolution error.

Regards,
Koh-Ichiro


----- End Included Message -----



----- End Included Message -----