Anticipation error with isynchronous auditory stimuli

Dear Cedrus –

(sorry about the long-winded thread title)

I am running a MacBook Pro Intel Core Duo (OS 10.4.9), and I am searching for a software solution for the following experiment:

Establish an isynchronous .wav file ‘click’ stimulus occurring at precise intervals (i.e. 90 bpm). A metronome, in other words. :slight_smile:

Subjects will attempt to synchronize finger taps with the signal using button presses on an input device similar to the ones presented on your site (for greatest accuracy). Trials will be one minute in length, resulting in 80-95 attempts to synchronize per trial.

I would like to be able to measure the ‘distance’ and direction (before or after) of each tap to each respective click. This is not meant to be a reaction time task, since a subject may tap before a metronome click.

Typical error in a task of this type is less than 100ms, and can be as little as 0ms, so precision and awareness of latency (as much as possible) is crucial. Heh - isn’t that what they all say? :slight_smile:

Will SuperLab be able to do this?

Thank you! Love the site!

I’ve been playing with this a bit on my end to see what I think. First of all, I’m not as accurate as I had hoped, and I guess that’s part of why I’m not a professional musician.

You should be running 4.0.3 or later.

SuperLab will not give you the offset from your 90bpm–you’ll have to calculate this yourself. Your event will be set up with your audio file. It’s important that it’s either WAV or AIFF format and that it is played through the AUDIO event type. You will set the event to end after 60,000 ms, or 60 seconds. You also must reset the RT timer at this event.

Once this is set up, you can run your experiment. Your data file will contain all of the responses with response times relative to the beginning of the audio file. Since you know that at 90bpm, the cycles occur every 666.67ms, you just need to compare the responses with multiples of this value:

667, 1333, 2000, 2667, etc…

Hope this helps!


I just tried to do this with 4.0.3 following your instructions, but I can’t get Superlab to record more than one RT (which is the first one). How do I get it to save more than one button press for each event?



In the input tab in the event editor, set the event to end only after a time limit. You absolutely do not want it to end on a response. If it ends on a response, you’ll only get the first response.