from:	 melvyn wright 
to:	 "Zhao, Jun-Hui" 
date:	 Thu, Sep 11, 2014 at 12:05 PM
subject:	 Re: Rebinning SWARM data with sma2casa.py
mailed-by:	 gmail.com

Hi Jun-Hui,

I'm currently imaging ALMA data using Miriad. I imaged
4 x 3840 channels before breakfast after a week's frustration
with casa. part my unfamiliarity,  part casa orneriness and slowness.
Image each spw (window) separately to keep fractional bandwidth small
enough that common beam is good approx. (1.8/660 GHz in our case)

Carma Mirad maxdim
$MIR/VERSION:
4.3.8 - [automated build Thu Jul 24 11:02:13 EDT 2014]
MIRBIN: /usr/local/miriad/bin/linux
$MIRINC/maxdim.h parameter:
 MIRTEL       = carma
 MAXBUF       =     10000000
 MAXDIM       =        65536
 MAXDIM2      =         8192
 MAXANT       =           64
 MAXANT2      =           28
 MAXBASE      =         1058
 MAXBASE2     =         1058
 MAXCHAN      =         8192
 MAXWIN       =           32
 MAXWIDE      =           32
 MAXPNT       =       100000
static membuf (maxbuf) usage:
 membuf()     =     9999998



Hi Mel,
    Handling interferometer array data with CASA is indeed frustrating when either the UV data becomes large
or one wants to image a large number of large size images.
    Based on my experience with CASA on handing VLA data, only the multi-frequency synthesis and multi-scale clean
algorithm developed by R. Urvashi/Tim Cornwell appears to be useful in imaging wideband data (e.g. 2 GHz at 5.5 GHz)
but the current version is extremely slow. Do you know anyone in the Miriad developer group intends to implement
this algorithm. I know Bob Sault implemented the zeroth order of this algorithm in Miriad quite a while ago but Urvashi
criticized that the Miriad version contains error. Do you know where is Bob S. these days? Does he still work at NRAO
to implement these good Miriad features into CASA?
    The high-order term of Taylor expansion of the clean algorithm in CASA also seems to be problem.
    I tried the new release 4.2.2 that was announced a few days ago. They claimed that the tool plotms is implemented with
a few new features but I could not find them and the new version seems to mess up the GUI layout in plotms. Apparently,
our CASA friends had too much pressure to rush out the new release.
    I just got my performance plan for next 4 months. Hopefully, SMA will renew my appointment....