Millisecond Forums

Hi All,
    I have a(nother) timing question.  The Inquisit help files make it quite clear that visual sitmulus presentation times are rounded to the nearest integer frame/retrace (which is system dependent), or presented in the specified frame if the /frames attibute is used.  In fact I have done some informal testing with a photometer and a CRO measuring interstimulus intervals and the like for a masked priming script and found the timing to be absolutely spot on when using /frames.  What I have not been able to determine is what happens with
a) soundfiles, and
b) intertrial pauses.

Given that system resolution is considerably finer than screen resolution, how does Inquisit handle the timing of soundfiles?  Are their start times locked to the screen refreshes or can they start playing at arbitrary times with millisecond accuracy?

Presumably, since trials are executed in terms of their frame contents, the start time of a trial is rounded to the nearest frame, so writing /stimulustime = [0 = Something]   will always yield the exact same timing as /frames = [0 = Something]?
Is this the case or are intertrial times, pauses etc asynchronous with respect to the screen refresh and the frame/stimulustimes shifted or rounded to fit?


Any clarification is greatly appreciated!

Cheers,
     Ben.

Good question. Currently, all stimulus onset (including sounds and port signals) are synchronized with the vertical refresh of the monitor. (Note that with sound, there will likely be a bit of a delay before the physical sound actually eminates from the speakers depending on the length of the sound files and the quality of the sound card, but the play command is nevertheless issued in sync with the monitor's vertical scan.)

I've often thought that nonvisual stimuli should be decoupled from the refresh rate, but to my surprise, no one has yet requested this feature.

Finally, intertrial pauses are not directly linked to refresh rate, although they may be indirectly affected by it given that stimulus presentation is. As an example, let's say you specify a pretrialpause of 46ms. Inquisit will wait 46 milliseconds (actually, it will use this time to prepare the trial and any stimuli to be presented), then it will wait until the next vertical blank to begin stimulus presentation. So, the interval may be extended a few milliseconds as the screen repaints.

Regards,

Sean