Most of the sound structure and elements have stayed the same from the first semester:
During the first part of the design process, the sound team decided that environmental ambience would be coupled with relatively static environmental music. The ambience and music would be rendered as 30-second loops, which would effectively play from one to the other as smoothly as they would loop. There would be six ambient loops and six music loops per region, and they would be played back via the sound engine by randomly being picked as the last one stopped playing. This provided us with a good amount of sonic material without audible repetition.
Footsteps were going to be designed as loops, fit to a creatures animated walk-cycle. These were going to be randomized and played back in the same manner as the music and ambient environmental sounds. However, switching to the Torque engine has made it possible to trigger individual footfalls to specific points within the walk-cycle, making the footstep sounds much more dynamic.
Limitations
Calls and vocalizations, as well as attack and eat/drink sounds, will be treated the same in Torque as they would have been in Director. They will simply be triggered every time the specified action is triggered for the creature in the engine. However, Torque has made it much easier to position all sounds in the 3d space relative to their host creature, saving our team and the programming team a significant amount of coding.
Last but certainly not least: at Hopeful Monster's inception, the sound team had the goal of integrating audio emitters into the game engine. An audio emitter is a sound object, positioned in a 3d space, which has specified acoustic attributes (ex: which direction its main "throw" is facing, reference distance to max. distance, etc...). We were planning on using these to enhance the environmental ambience, for example: positioning birds in trees, or patches of grass which rustle, the sounds of which would stay in their relative position as the player moved past/around them. This appeared to be a huge challenge when we were working with Director, but Torque has emitter capabilities built in, so it turns out that our "pie-in-the-sky" will most likely be much easier than most of the other audio commands we will have to create.