Eyelink recording question

Hi all,

I'm trying to implement an eye-tracking experiment in PsychToolbox3, and I'm running into a recording problem. I'd like to start recording at the beginning of a trial, continue recording through two or more trial phases, then stop recording to do a drift correction/central fixation, and finally start recording again at the beginning of the next trial.
The challenging part of my problem is that it's inconsistent, but the short version is that sometimes in the stop/drift-correct/start interval something goes wrong. It seems like the Host PC fails to catch the stop-recording command, but even putting Matlab into a while loop that issues the stop command, waits, and checks the recording status each time through fails to stop the recording. Longer trials might be more of a problem (~30s), but I think that I've observed the same issue when I keep the trials short as well. Has anyone else run into this kind of problem? Thanks,

Dave

--
Post-doctoral Fellow
Neurology Department
University of Iowa Hospital
davideugenewarren@...