Software for the SR Eyelink II and Eyelink 1000
David Stracuzzi and Jeff Kinsey have written a suite of software for conducting experiments on the SR Eyelink II and Eyelink 1000/2000 eyetracker. It is conceptually based on software written at Saarbruken and provided to us by Christoph Scheepers, but all the code is new and subject to OpenSource license. If you use this software in published experiments, we would appreciate a reference to "EyeTrack" on this web page.
There are multiple versions of the experiment-running program EyeTrack. You can download the one you want from the right or from the "museum." Changes through 0.7.9 are documented in the EyeTrack change-log file under "downloads" (changes in 0.7.10 mostly involve improvements to display change routines). The latest version is 0.7.10m, but 0.7.10k is also available on this page (both do multiple display changes on multiple lines by insisting on a legal fixation on a line before any changes occur on that line, so that return sweeps don't accidentally trigger display changes; 10m has some useability enhancements, e.g. popup help windows). We recommend using the latest version, but the others are there for history buffs. All versions are beta (bug reports gratefully accepted by Chuck Clifton). Each can be unzipped into its own directory, maintaining "code" and "documentation" subdirectories. You may also want to download "Scripter," a Perl script written by Adrian Staub, that permits development of scripts in tab-delimited files prepared in Excel.
A slightly-updated collection of html and Word documentation files (including Scripter instructions and side notes) now exists as EyeTrack Documentation.
There are also multiple versions of the data-cleaning program EyeDoctor (it replaces the DOS program EyeWash). Again, download the latest from the menu to the right, unless you have old data files that require an older version. Check the EyeDoctor change-log.html for some history. EyeDoctor 0.6.5 deals well with data that contain multiple display changes on a single trial, correcting a few minor bugs that were in 0.6.4b and saving display change information in the temporary .edd files to permit a user to take a break while eyedoctoring a file with display changes. (Note: the zip file now includes a critical .dll file that was missing from some earlier distributions.)
HTML help files for EyeDoctor are available for download. They are not fully up to date as of 11/27/09, but should be useful. Unzip the package and open index.html with your web browser to read them.
Note, these programs require some .dll files provided by SR Research. Install their Display Computer software to get these files, even if your computer is not connected to an EyeLink. You'll be able to run miniEyeTrack to check and debug your programs.
You can use the program edprep60.exe, found in the full dataanal package (below), to prepare .cnt files that allow you to use eyedry with the output of eyedoctor. You can use the program nurse, also available as part of dataanal, to prepare .sen files from DPI .stm files so that you can use EyeDoctor on DPI .dat files. Read the comments at the start of the .c versions of these programs for some brief instructions in their use. Question-answering accuracy can be scored using ascacc1.exe, also found in the dataanal package (and again, minimal instructions are available in ascacc1.c).
The zip file dataanal (available in the downloads at the right) is a collection of DOS data analysis programs for eyetracking. ReadMe.txt gives a brief synopsis of what is in dataanal, and notes that you can get any of the old data analysis programs that used to come along with these more useful programs directly from Chuck Clifton.
Download eyedry.zip to get a current version of eyedry compiled on a Windows port of GCC (modeled on a Linux compilation that Patrick Sturt did a few years ago). It should eliminate the old memory limitation problems, and maybe run faster.
WINDOWS EYEDRY IS NOT CURRENTLY MAINTAINED: DOS-phobics, you can now use a Windows version of EyeDry (available at the right). The program was done by Jeff Kinsey. It has been quite thoroughly checked out, but use it with caution. All the actual computing code is taken from the DOS version and it seems to produce identical outputs. It has some stylistic quirks in the interface (e.g., it crashes without any error message if you have a missing or illegal parameter value) but will be improved. Let Chuck Clifton hear suggestions for improvement. He'll pass them on to Jeff. The latest version, WinEyeDry_0.4.5, has an .ixs output formatted for easy input into R.
If you still run a DPI tracker, you can get the suite of experiment-running programs, called pcexpt, from "downloads" as well. You may want to get the data analysis programs for the DPI directly from Chuck Clifton as well.