Recording verbal responses?

Hi there,

We are considering using Superlab for a series of experiments which will require participants’ verbal responses to be recorded via a microphone. We will also need to measure voice onset times, and each recorded sound file to be named according to the specific stimulus. Can you tell me if Superlab has this capability, as I can’t tell from the manual?

Also, are there any potential issues with accuracy of voice onset timings when the audio is also being recorded?

Many thanks,

Briony

Hello Briony,

SuperLab cannot record the actual response - only its RT. On SuperLab for Mac, you can use a standard microphone, but the timing accuracy is about 20ms. You can get better timing accuracy with our SV-1 device, about 3-5ms.

Thanks for the information Hisham :slight_smile:

Hello Hisham - I wanted to confirm what you mean about “the timing accuracy is about 20 ms”. Does this mean that all voice onset times are delayed by 20 ms, or does it mean that the voice onset measurement has a margin of error of 20 ms?

eahuntley, I believe that the system polls the microphone every 20ms so that the largest error you would get is 20ms but the error will vary for each trial.

Ed, SuperLab polls several times each millisecond. The 20ms delay comes from using USB-based microphones. This is because the operating system buffers the input and passes the data to the application only once every 20ms. This 20ms delay does not happen if you use our SV-1 voice key or our StimTracker Quad, which has a built-in voice key feature. Nor does this delay happen if you have an older PC or Mac with a microphone jack. (please note that having a mic jack is not a guaranty of better timing: some computers implement the microphone interface as an internal USB port).