I believe I understand. The problem is that you are already using the BGR function to keep your ratings together with your stimuli. So then you can't use BGR to randomize the order of image/audio.

There are a number of ways you could approach this. I think the easiest would be to create two input files--one for images and one for audio and then have MediaLab take care of the administration of the files. Or you could run the two sessions back to back manually using a pre-determined random ordering by subject ID. In the next version, we hope to have the capability to use multiple levels of BGR and WGR randomization for just this very kind of thing.