voice rt sound files longer than requested

Hello, I'm working on an experiment where particpants are required to give vocal responses to auditory stimuli, and we're measuring RT's for the vocal response. I also want the experimenter to be able to press a button to code the vocal response. I have it set up so that there are three stimuli on each trial: sound file, text(just to have something on the screen to indicate that it moved from the sound stimulus to the next stimulus), text (to allow the experimenter key press). The audio RT is taken to the second stimulus. (We are actually interested in rt's to the sound file, but we can't use headphones and if we take an rt to the sound file, the sound file itself activates the rt.)

So this is all well and good, and I have it so the microphone records for 300 ms after voice rt begins: rt:aud,rec,-300 (just enough time for the participant to say "stop" or "go"). However, when I check the recorded sound files, a few of them are way longer than 300 ms, maybe a couple of seconds. They start immediately with the voice response, but then just keep going past 300 ms. There are also some that seem to start recording before the voice response. That may just be an issue of microphone sensitivity, but I don't think so. On both of those types of weird recordings, the reaction times in the log file seem to be longer than they should be.

Any insight as to why some of the recordings of the voice rt's are longer than the requested 300 ms, why some recordings start before the voice response, and if this means that reaction times aren't being recorded accurately?

Thanks.

Let me be more concise; hopefully someone knows the answer. Please let me know if you have any ideas.

There are really two separate questions:

1) I am recording voice responses with rt's, and I have the time set as rt:aud,rec,-300, but some of the sound files that are recorded are up to five seconds long, even though I only requested 300 milliseconds of recording. Why might that be happening?

2) Some of the voice response sound files seem to have started recording before the actual voice response occurred, which probably means the reaction times are inaccurate. Is there a reason this might be happening other than microphone sensitivity issues?

As a matter of fact, none of the recordings are 300 ms. The shortest seems to be about 440; is that normal? Some are around 800 or 900, and then there are also the ones I mentioned already, that are several seconds long.

Well let's try to tackle the duration issue first. Remember that [I]rt:aud,rec,-300[/I] will cause DirectRT to record the vocal response for an additional 300ms after the RT occurs. So if the RT is 200ms, you should get a voice clip that's around 500ms long. Alternatively, you can specify [I]rt:aud,rec,300[/I] to record for exactly (give or take) 300ms.

See also: [URL="http://www.empirisoft.com/directrt/help/directrt_samples_of12_voice_rts_.htm"]www.empirisoft.com/directrt/help/directrt_samples_of12_voice_rts_.htm[/URL]

Okay I think I see now. I had thought that it waited to start recording until the beginning of the voice response, but from what you said I guess it starts recording right away, then once the voice response occurs, it records for 300 more milliseconds.

In other words, if it took 30 seconds for the participant to give a voice response then the sound file would be 30.3 seconds long; is that correct?

Thanks very much!

Yes, that is correct.