I'm not totally positive, but I think it depends on the object. You'll need to tune the arrays of speakers such that they generate some kind of 3D interference pattern that perfectly matches whatever you're trying to image.

So as seen in the video, you can get a cool stepped pyramid shape (when the particles are thrown into the air), but you'll have to be pretty smart about what kinds of images you'll be able to display.

Can someone prove me wrong? I'm speaking at a very elementary wave/particle physics standpoint.