I've got this array of speakers: A ring of 8 on the 'main floor' - stage- and audience-level - and a ring of 8 speakers on the balcony level. 16 total. I'm using ICMS's ambisonics objects to do some z-plane panning from the main floor ring to the balcony ring. The sources start at a random x-y coordinate in the main floor ring, and then pan up (rise? tilt?) to the balcony one at a time over a duration of 3 seconds or so.

The problem I'm having is the z-plane behavior in the ambimonitor object doesn't seem to be getting encoded in the output, though the x-y positions seem correct. I'd expect to see that the live.gain objects I have on channels 8 - 16 (to show me what will go to the balcony-level speakers) coming out of the ambidecode object light up a little later than channels 1 - 8 (the ring of 8 on the main floor) as the sources pan upward. This isn't the case though, all 16 channels have about the same amount of audio going to them.

I'm not really using the ICMS ambisonic objects for 'traditional' ambisonics, but rather to just get a visual representation of the speaker array and the motion of the sources up and running quickly, and bake some motion into some multichannel recordings. Anyone familiar enough to show me where I'm going wrong? The patch is attached. I've tried messing with 'distance coding mode' and a few of the other settings in ambiencode's inspector window, but nothing seems to make much of a difference. Thanks.