View Single Post
Old 07-24-2016, 11:59 PM   #4
Human being with feelings
Join Date: Jul 2011
Location: Bergen, Norway
Posts: 92

Originally Posted by ReaDave View Post
I've literally just finished setting up and testing my first 3D cube using mostly Auratone 5C cubes and have been using your ATK (and the Periphonic 3D decoder) for testing and listening. I made another post about that and just replied to it with some details about the cube array.
Cool, I'll check that post out.

Originally Posted by ReaDave View Post
A quick question about the periphonic decoder... Is the elevation angle the angle between the horizontal listening plane (midway between the upper and lower ring in my current setup) and the top and bottom speaker rings?
Yes, it's the elevation angle from horizon to the upper ring, and similar to the lower ring. So at 0 deg both rings would be in the horizon (which doesn't make much sense) and at 90 degrees the rings have collapsed to be just above and below you (which doesn't make sense either). So it's very likely that you'll work with an angle somewhere in between.

There is a small visualisation in the GUI of the elevation angle, in between the two circles and a little to the right.

I should mention that the Periphonic 3D decoder does work correctly yet. When porting the code from the SuperCollider version of ATK I failed to realise that the matrixes used are to be "pseudo-inversed". I have tried and used the Periphonic 3D decoder in it's current state myself, and get quite good results, but according to Joseph Anderson it will sound better (and be more correct with respect to Gerzon's ideas about ambisonics) when I have added the pseudo-inverse. This is tracked in relation to #2, #4 and #5 in the issue tracker:

I hope that it won't take to long before I get around to address this, as it's one of the main tasks left before ATK can be properly released, but I have a pretty packed schedule thought September.
lossius is offline   Reply With Quote