Direct RT stimuli into to Media Lab

Hello!
I found information in the help menus about using ML responses in DRT, but I am wondering about using information from DRT and putting it into media lab.

Specifically, in DRT we play a little conversation "movie" in which we have pictures randomly assigned to different clips of narrative (e.g., so sometimes female1.bmp is matched with clip1, other times with clip 2, and so forth). We have almost all of this stimuli on one line in DRT to make the conversation less choppy (found that tip on the threads-thanks!) and if the same person talks more than once, we can easily refer back to the randomly picked picture (so if female1.bmp was paired with clip 1 and should also be paired with clip 4 because clips 1 and 4 are the same person talking, we used ?b1t2s1 to recall that image-again, thanks, this is an awesome feature).

Later, we ask questions about the clips they heard, but we wrote these questions in ML. Is there any way we can have ML call up the correct image that was randomly selected, like the ?b1t2s1 command in DRT? Or could I maybe somehow have DRT write the stimuli used to a file to be called up later? We already employ a pretty complicated response.xls file, so adding a little more won't kill us :)

It's not a dire need, but since we ask about a few different people we thought it would be nice to have their picture come up when we ask a question about what they said. I know the easiest solution would be to write all of the questions about the audio in DRT, but I'm curious if there is another way.

Thanks for your help,
Susan

First of all, thanks for all the nice comments about features you have found useful :)

Wish I had an easy answer for you on this one but I'm a frayed knot. Hmm.

Only thing I can suggest off hand (without some excruciating programming on my part) is to write a short program that can read the csv log data from DirectRT. The log files contain the data you need I think and can be read pretty easily by any program using an input function. Once these values are found in the input file the same program could save them to your responses.xls file (if you want to get real fancy) or to an HTML file which could be presented within a specified region on any MediaLab screen. If you have access to a programmer, the code required for this would be pretty doable--probably less than an hour to read the log and output the values you want to an html file (which is really just a text file with an .htm file extension).

Then, you can just run the exe from within MediaLab after the DirectRT session but before the ML questions. An exe can be run silently in the background (no graphical interface or pause in ML):

[QUOTE][I]From the user's guide re Executables in MediaLab:[/I]
Note that for all [..] item types, an executable file may be launched at the start of the item by adding a "c" parameter followed by the file name, e.g., (p7,cmyprog.exe) would cause [I]myprog.exe[/I] to launch at the start of a 7pt scale response. This is especially useful for programs that should run silently in the background as MediaLab will [I]not[/I] attempt to keep the program visually in the foreground [I]unless [/I]it is launched via the Executable item type. [/QUOTE]

Now with all that said, what would be super crazy cool is a [I]responses.xls[/I] file that could be accessed by BOTH DirectRT AND MediaLab during the same session. You know I have never even thought of that before (esp. given that DirectRT does not currently use that functionality). Hmm. Looks like we have another feature on our to-do list!

Let me know if you come up with something that works--I'm guessing many other users would love to know how you did it! And of course, please feel free to ask any additional questions or for help with it if you run into any trouble.