The case for PsychoPy

Most researchers in cognitive neuroscience have to program an experiment in a stimulus-delivery environment at some time in their career. Everyone have heard about Presentation and E-prime but much fewer have heard about PsychoPy – a free cross-platform stimulus delivery environment.

PsychoPy logo

Advantages over Presentation and E-Prime:

  1. Free and open source, meaning no license problems!
  2. Cross-platform. You can develop on Windows, Mac or Linux and run your experiment on Windows, Mac and Linux!
  3. Limitless functionality. Psychopy is a Python module, which means that it can be heavily extended with other Python modules for e.g. Matlab-level matrix operations (numpy), image and audio manipulations, advanced statistics etc.
  4. Flexible stimulus-handling: “Online” stimulus-generation, e.g. random-noise masks or moving Gaussian gratings. Also, you can define stimuli-sizes in degrees relative to the subject’s eyes, rather than having to do the trigonometry yourself.

Timing:

It compares to E-Prime and Presentation on timing. I’ve tested it on our stimulus-facilities using a photosensitive diode and comparing the timing of stimuli-appearance on the monitor to the triggers sent on a parallel port. It is reliably below 0,1 ms.

In PsychoPy you can (optionally) define durations of stimuli in monitor-frames rather than milliseconds, which ensures that timing is very well-controlled.

Disadvantages:

  1. PsychoPy’s timing of audio stimuli is still not perfect. I’ve found a ~10ms jitter in audio onset (I’ve seen much worse in some C-compiled experiments…). If you need very precise audio timing, go for something else.
  2. The serious experimenter still needs to script all of his experiment. Although a very nice graphical interface is in development, it is still not thoroughly tested and tried. In my oppinion programming in Python is much cleaner and easier than E-Prime and Presentation.

Getting started:

Good introduction to Python syntax: Learn Python in 10 minutes
Download PsychoPy at PsychoPy.org and read documentation here.