Millisecond Forums

seemingly random terminations of the script

https://forums.millisecond.com/Topic27276.aspx

By Motto - 6/11/2019

Hello everyone,

at the moment we are running an oral PSE inquisit script. This script shows a picture for 10 sec and after that, subjects have 3-4 min to tell a story in a headset-microphone. To capture the voice we used the soundcapture feature. After 3 min subjects get a cue telling them, that they have one minute left. If they are ready, they can use the mouse to click on a continue label. They get the last cue after 4 min, that the time is up and they should click continue. The script does not "forceforward" them. It waits for them to click continue. After that the next picture is shown for 10 sec and so on. After 4 pictures and stories we show them an end slide and the script is completed.
Now the problem: The important data in this experiment are the audiofiles, which are only saved, if a next trial starts. We encountered the problem, that on some occasions the script just terminates after clicking on the continue label and so the last audiofile is not saved and all further pictures and stories are not recorded. This happened on different computers, after different picture stimuli and after a different number of pictures shown. There was no error message. It just terminates. But there were also a lot of times, when it just worked perfectly. Everything was shown and we got 4 complete audiofiles and a complete data-file.
We tried different things, to erase the error. We ran the script from a network drive --> random termination. We ran it from local drive --> random termination.
We even tried different things, to replicate this error. We did run it without a microphone --> no termination. We unplugged the microphone during the trial --> no termination.
The script with the soundcapture feature is the only script that produces this problem, so we assume that there must be a problem with it. I will attach a .zip file containing the script and the relevant stimuli. (In the data send command I removed the path on purpose.)
We are at a loss and very grateful for any help you can give us.

Thank you,
Dominik
By Dave - 6/11/2019

Motto - 6/12/2019
Hello everyone,

at the moment we are running an oral PSE inquisit script. This script shows a picture for 10 sec and after that, subjects have 3-4 min to tell a story in a headset-microphone. To capture the voice we used the soundcapture feature. After 3 min subjects get a cue telling them, that they have one minute left. If they are ready, they can use the mouse to click on a continue label. They get the last cue after 4 min, that the time is up and they should click continue. The script does not "forceforward" them. It waits for them to click continue. After that the next picture is shown for 10 sec and so on. After 4 pictures and stories we show them an end slide and the script is completed.
Now the problem: The important data in this experiment are the audiofiles, which are only saved, if a next trial starts. We encountered the problem, that on some occasions the script just terminates after clicking on the continue label and so the last audiofile is not saved and all further pictures and stories are not recorded. This happened on different computers, after different picture stimuli and after a different number of pictures shown. There was no error message. It just terminates. But there were also a lot of times, when it just worked perfectly. Everything was shown and we got 4 complete audiofiles and a complete data-file.
We tried different things, to erase the error. We ran the script from a network drive --> random termination. We ran it from local drive --> random termination.
We even tried different things, to replicate this error. We did run it without a microphone --> no termination. We unplugged the microphone during the trial --> no termination.
The script with the soundcapture feature is the only script that produces this problem, so we assume that there must be a problem with it. I will attach a .zip file containing the script and the relevant stimuli. (In the data send command I removed the path on purpose.)
We are at a loss and very grateful for any help you can give us.

Thank you,
Dominik

I see no obvious issue with the script itself. Given that you're potentially recording for a long time (over 4 minutes, since the script does not force a trial to terminate, but instead waits for a participant to click continue), there audio data accumulating in memory can easily get quite large (several hundred MB if not more). What's likely happening is that the system simply runs out of memory and does not have sufficient resources available to write the data to disk and carry on. You might be better off with having a dedicated audio recording application running in the background to record the entire session, and letting Inquisit handle only the display part.