problem with presentation of movies and text in a single trial

Hi,
I have an experiment in which a single trial consists of a text event followed by a movie followed by another text event. The events are not randomized. However, in running the experiment, SuperLab displays the text that is supposed to come before the movie twice, both before AND after. If I replace the movie with another text event (I did this to try to troubleshoot the file, as troubleshooting with the movies being played in the experiment took way too long), the experiment is just fine. It’s only when superlab is playing the movie event that it seems to display the text preceding the movie twice, before and after. The movie is a 90-second .avi movie. Any idea what could be going on?

Joanna

When you say that it displays the text stimulus twice what do you mean? Is the text on twice as long as it should be, or is there a blank screen interval in between two periods with the text? Is the blank screen interval (if any) a longish one, or is it just sort of a flash? If it’s just a flash, then maybe it’s some kind of glitch, possibly as a side effect of the movie starting and stopping.

It might be informative to look at the logfile and see how many times Superlab claims each stimulus was presented and when.

Greg Shenaut

I mean that a trial consists of the following event:

(1) Instructions (text event)
(2) Video (video event)
(3) Question (text event)

When we run the trial Superlab presents the following:

(1) Instructions (text event)
(2) Video (video event)
(3) Instructions (text event)
(4) Question (text event)

It presents the Instructions both before AND after the video.

In order to troubleshoot I changed the video to a text event so we had the following sequence:

(1) Instructions (text event)
(2) Random text (text event)
(3) Question (text event)

The trial was fine. It only presented the instuctions twice when the video was presented.

I don’t know how to look at the log file and I can’t find any instructions on how to do that in Superlab.

Does the Instructions event wait for a response when it is presented the second time, or does it just flicker on the screen?

Internally, movies are played in their own view/window. On Mac OS X, each view is buffered (by Mac OS X itself). Therefore, when the movie goes away, the image that was behind it comes back. Try erasing the screen or stimulus at the end of the first instructions event. This can be done easily in the Presentation Options for the event (click Settings… under the Stimulus tab).

There are two separate logs in SuperLab. There’s the actual data file that is created when the experiment is run. SuperLab is capable of providing extra information through its verbose log. This hidden feature was introduced in SuperLab 4.0.3. Directions for how to enable this are included here for Windows and here for Mac OS X. This spits out a very large amount of information in a window after you’ve run your experiment. It’s not pretty, as it’s basically a big list of everything that happened while the experiment was running with time stamps of .01ms precision. This second log file would be more informative for the given case, but it can be a bit overwhelming. (note that the information/error dialog that comes up when SuperLab is launched in verbose mode is normal).

Hi,

I’m currently working with jmorris on this problem. For your question, the instructions event DOES wait for a response when it is presented the second time. It does not flicker.

The program is being played in windows, but it has been tested on both mac and windows, with the same problem reoccurring.

If/once I’m able to extract the log file, what should I look for in it?

Does it wait for a response the first time? If not, then maybe it’s actually just one presentation of the instructions that is being “covered over” by the video.

I’d start by looking at the regular experimental data logfile, after turning on as much reporting as possible, just to figure out the sequence of events.

The verbose log does have a lot of information in it, but I think it’s more useful for tracking down bugs in Superlab or to determine how long events and pieces of events take to do.

Greg Shenaut

The verbose log is hidden the way that it is because it’s not intended to be the primary way to debug an experiment. It’s an excessive amount of data, and it’s not pretty. If you can make sense of it, then great, but I would much rather spend the time writing a user interface for it than spend the time writing instructions on how to decipher the text version. Greg is right that it’s more useful for tracking down bugs and for determining timing. The only exception to this is with expressions in macros. There’s currently no other method of finding the values calculated in an expression. This is the entire reason the verbose log was released to begin with. It could be helpful with your experiment, but by no means should it be necessary.

Did jmorris send you a copy of the email I sent her last week? Did any of the things I mentioned in the email help? Without listing off specifics: there are two trials that are set to randomize the events in the trial. Also, all of your trigger events explicitly look for input before ending the event, which I suspect is not what you want.

I hope this helps.