Without seeing the code, it's like trying to diagnose what's in the blackbox. ;-) I'd expect to see something like this happen if the list of texture area names was predefined in the script rather then read from the figure at run time. If the figure being used had texture area names different from the pre-defined list, there could be problems. That's just a shot in the dark. You might want to add a 'try' program structure in the loop that's changing the parameters to see if you could trap an error that way. A well placed 'print' statement to list what's happening in the loop would also be useful. It would at least let you know how may iterations occured before the loop broke. You could set it up to print the current values for the variables being used.