Processing (Calibrating and Imaging)

SMA Data

Processing Software: MIR, MIRIAD, AIPS, and CASA

The raw data in the SMA data directories, i.e., correlator output, are in MIR format, but can be converted to MIRIAD or FITS (for AIPS) formats, if desired.
  • Processing with MIR
    • Pros: No format translation necessary. Well tested.
    • Cons: No imaging. Memory intensive.

    Details of data processing using MIR can be found in The MIR Cookbook.

    Note: By typing "idl", you will access 32-bit idl on 32-bit machines and 64-bit idl on 64-bit machines.

  • Processing with MIRIAD
    • Pros: Calibration and imaging in same package.
    • Cons: Format translation necessary. Less well tested.

    Details of data processing using MIRIAD can be found in the MIRIAD Users' Manual.

  • Processing with AIPS
    • Pros: Exceedingly well tested (it has been around since the late 70's).
    • Cons: Translation and preliminary processing necessary.

    Details of data processing using AIPS can be found in The AIPS Cookbook.

    Note: To do preliminary processing and convert data to FITS format, consult The MIR Cookbook, above.

  • Processing with CASA

  • Details of data processing using CASA can be found in the SMA CASA Website

Obtaining the Raw Data:

If you do not have an account on the CfA's R&G computers:
  1. Click here for data search and request (or download) instructions.
  2. Install the MIR and MIRIAD software package on your home machine and process from there.

If you do have an account on the CfA's R&G computers:
  1. Retrieve the data files from the appropriate area (see table below). If you are not the PI, please limit your access to data outside the proprietary period (>15 months).
  2. Process the data locally, using either MIR or MIRIAD.

Data Location (on the R&G computers):

Data Year Location
2003-2005 /pre_2006/science/mir_data
2006-2008 /2006_2008/science/mir_data
2009 /2009/science/mir_data
2010 /2010/science/mir_data
2011 /2011/science/mir_data
2012 /2012/science/mir_data

  1. The R&G computers available for SMA data processing are: RTDC2-7.
  2. Prior to May 2007, testing data (baseline, engineering, flux, pointing, and priming) were mixed in with the science data. They are now directed to their own areas, i.e.,


  3. Testing data are also subject to the 15 month proprietary period.

  4. Large data sets (> 4.3 GB) must be processed on a 64-bit machine (RTDC2-4,6-7). Small data sets (< 4.3 GB) can be processed on any machine.
  5. Please do not use your home directory for analyzing large data sets. Ask for a temporary work area, if you need one.