If you want to use MediLab for this kind of task and are able to use either the arrow keys or a joystick for participant responses, you could use the On-Line Ratings item type. Take a look at this section of the MediaLab manual for information about it: http://www.empirisoft.com/medialab/h...emtype.htm#olr.
If you want to use DirectRT, you might be able to do this by taking advantage of the having multiple RTs in a single trial of the .csv input file; see this section of the DirectRT manual: http://www.empirisoft.com/directrt/h...ltiple_rts.htm. Essentially, you could list the sound file as the first stim in the trial. Use of the sound file would most likely require that you put the length of the sound clip in ms in the time cell right next to the stim cell with the name of the sound clip file; see the Sound Clip section here: http://www.empirisoft.com/directrt/h...ectrt_stim.htm. You would then follow it with a series of enough rt:any to record whatever number of keypresses the participants should make during the trial. For a ten minute clip, I could see you needing to use a few dozen just to ensure that your participants won't make more keypresses than the trial has defined. Finally, you can use the Automatic Continuation feature of the MinMax cell of the styles file to create a style code that will automatically advance to the next trial once 10 minutes or so of time has passed to move on to the next trial in the .csv input file; see the MinMax section of the manual here: http://www.empirisoft.com/directrt/h..._rt_limits.htm.
Good luck!
Jason Reed
Empirisoft Software Support Specialist