• News
• Introduction
• Prerequisits
• Installation
• Documentation
• Copyright
• Acknowledgements

Synopsis: Software to reduce Speckle or AO data from the raw data to deconvolved images (in Fourier space), to determine the parameters of a binary or triple, and to find limits for undetected companion stars.

Operating System: Unix, tested on Linux (Redhat and SuSE) and Solaris

Language: C and Tcl/Tk

History: Developed by R. Köhler, see for example Köhler et al., Astronomy and Astrophysics, v.356, p.541-558 (2000)

20. February 2002: I encountered more problems caused by the copyright restrictions of the Numerical Recipes, so I decided to write my own implementation of the amoeba-algorithm. This should make installation much easier.

22. January 2002: Another new version on this website. Again, the changes are mostly bugfixes, this time related to warnings when compiling on non-Intel/non-Alpha machines and the interpretation of new SHARP I-files. There are a few functions which are undocumented, untested or simply do not work in the distributed version. Everything described in the documentation should work, however.

19. June 2001: A new version is avaiable on this webpage. The changes are mostly bugfixes related to the byte-order on Intel- and Alpha-machines. To download it, go to the Installation section.

Introduction

Publications describing the method are of course my PhD thesis (which is unfortunately in German), Köhler & Leinert, Astronomy and Astrophysics 331, 997 (1998), and Köhler et al., Astronomy and Astrophysics 356, 541 (2000). The latter paper contains an especially comprehensive explanation and some examples, so I include it in the tar-file of the package (here is the PDF-file).

Prerequisites

• A C-compiler
  I used the GNU compiler most of the time, but any compiler that understands ANSI-C should work.
• Tcl/Tk

  Version:	8.x (tested with 8.0 and 8.3)
  Available at:	http://www.tcl.tk

  Please note that we also need the tcl/tk-libraries. It's not sufficient to have the "wish" program only! Usually already installed on Linux.
• GNU make

Version:	3.79
Available at:	http://www.gnu.org

This is not only used for compilation, but also during data reduction! Usually already installed on Linux.
• SAOimage DS9 and XPA

Version:	1.9.8 or 2.0 (DS9), 2.0.3 (XPA)
Available at:	http://hea-www.harvard.edu/RD/ds9

In order to use the xpa-protocol on SuSE linux, we had to set the environment-variable XPA_METHOD to "local". In the csh or tcsh shell, this is done with:

      setenv XPA_METHOD local

In the bash shell, use

      export XPA_METHOD=local

• FFTW

Version:	2.1.3
Available at:	http://www.fftw.org

This is used during the deconvolution only. Of course, that is probably the most important part of the program.
• PGPLOT

Version:	5.2.0
Available at:	http://www.astro.caltech.edu/~tjp/pgplot/

The programs tkfmovie, binplot, and maxbright use the /PS and the /XTK drivers. If those drivers are not installed, you have to install them, which might require a complete re-installation of PGPLOT. You need a FORTRAN compiler to install it from the source. Here is the drivers.list I used when I installed PGPLOT.
• netpbm

Version:	probably not important
Available at:	http://sourceforge.net/projects/netpbm/

Used only to create a hardcopy of the results, and only the pnmtops program is needed. The README file in the netpbm package explains how you can compile only one program.

Installation

1. Download the tar-file:

   Here is a gzipped tar-file of the sources

   Here is an example data set (raw and reduced data)   28 MBytes!

2. Unpack it:     gunzip -c Binary-Speckle.tar.gz | tar -xvf -

3. Go to the Binary-Speckle directory:     cd Binary-Speckle

4. Edit the Makefile in the Binary-Speckle directory. The most important change is to enter the locations of Binary-Speckle itself, Tcl/Tk, FFTW, and PGPLOT. There is a section with machine-specific settings for Linux, Solaris, and HP-UX, uncomment the entries appropriate for your machine. The current version has only been tested on Linux and Solaris 8, but older versions ran on other Solaris versions and HP-UX. Please tell us if you have to change something.

5. Type:

   	make
   	make install
         

   This should compile everything and put the executable programs into the toplevel directory.
6. You have to add the top-level directory to your path-variable, so the shell is able to find the programs. This is done best by editing .cshrc (for tcsh-users) or .profile (for bash-users) in your home-directory.

Documentation

1. Reduce the raw data to deconvolved fourier-transformed images
2. Determine the parameters of a binary or triple
3. Find limits for the brightness of an undetected companion

The principle is (hopefully) pretty simple: There is a master program called "mcspeckle" ("mouse controlled speckle", since its predecessor didn't have a graphical user interface). The user tells mcspeckle which data files to use, what is in the files (objects, sky etc.), and what should be done. Mcspeckle than (usually) creates a Makefile, i.e. an input file for the make program that contains all the dependancies between the different data files and the commands needed to do the data reduction. Make then decides what has to be done to create the final results. In the good (?) old days, when one data reduction ran the whole night and quite often ended with a "disk full error", this was very useful. One could simply fix the problem and run make again. Make then checked which files were up to date und ran only the commands for the files that had not yet been reduced. The user shouldn't have to fiddle with the contents of the makefiles.

Of course, someone has to do the work of reading the data, doing something with it, and writing the results to another file. This is the task of most of the other programs in the package (despeckle, findcenter, fits2pnm, fitsmean, maskedit, prespeckle, quickvis, saa, and tkfmovie). Some of them have their own graphical interface to let the user adjust something, but most of them are simply called by make or from the shell, and accept only command line options. The most important option is "-h", which will show all the other options.

Furthermore, there are binplot, amf, and bff, which fit a binary or triple model to the data. Finally, maxbright calculates limits for undetected companions.

Here are tutorials for the three basic tasks:

Reducing the raw data and deconvolution

Finding out if a star is not a binary

Finding the parameters of a binary

Copyright

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA

If you find the program useful, please acknowledge it in any publications that result from its use.

Acknowledgements