I’ve run into a mpeg bug I think. I’m trying to run two movies back to back and repeat them about 15 times. They are about 15 sec. each and the file size is around 20 KB. It seems when run the experiement, multiple issues arise. 1) Sometimes the screen doesn’t clear the screen all the way when the 2nd movie starts and 2) there sometimes is a delay of 1-2 seconds before the next movie starts. Also, when I try to exit out of the experiement (esc), I can’t. Superlab just keeps running (sort-of) and the movies try to keep running. I have to force quit, which takes a long time as well. Any suggestions?
You’re running this on 4.0.2 pre 5, right? build 2260? I’ve made a few minor changes since then, but they shouldn’t be related to running an experiment.
There is a known issue with the movie plugin on the Mac that has been improved (though not completely removed) in your build. When a movie is removed, it marks the area of the full screen window behind it as dirty. The next time the operating system is given a chance to do its thing, it draws the default pinstriping. Try as I may, I haven’t been able to get rid of this, though I have been able to significantly reduce it.
Do you mind sending me your experiment? I may be able to find a workaround for you.
I believe we’ve mostly resolved this over the phone, but it’s worth saying on the forums:
When you set a time limit on an event, the timer does not start until after the event is considered to be done. If you are waiting for a movie to finish before moving on, this timer does not start until after the movie has finished playing. If you are not waiting for the movie to finish (in the event settings in the event editor), then the timer starts immediately after the movie is started. As a result, the time limit on a synchronous event (with no participant input) winds up increasing the length of the event, not decreasing it.
The escape key is not handled perfectly everywhere when running an experiment. For example: it currently is not handled correctly during the delay that occurs to synchronize trials, and it may not be handled correctly if no events look for a response. The combination of these two appears to be why it seems a little unresponsive. For now, just keep hitting the escape key and it will eventually cancel the experiment.
All text and picture events draw to the background of a fullscreen window. Movies (and string input) count as a child to this window. On the Mac, the actual background is not erased until the OS is given a chance to erase it (which I try my hardest NOT to do). As a result, when the movie goes away on the Mac, the preexisting background shows back up. Insert an event to erase the background in the trial that shows the crosshair, and the crosshair will not show up again unless redrawn. On MS Windows, the background of the window is not stored by the OS, so it immediately redraws the background color when a movie or string input field is erased. This is set to be your experiment’s background color. String input is a little different from movies in this area, as string input has to give the OS time to do its thing–otherwise you wouldn’t be able to navigate the edit field.