Hello all, For most of the day and evening I've been trying to find the error in some of the code I use to compute AMB decoders and which has been updated and extended recently. It fails on one of the test cases.
If - I compute a systematic 3rd order decoder for a regular icosahedron, using the standard pseudo-inverse method, - apply the per-order rE gain factors which AFAIK are 1.0, 0.862, 0.612, 0.305, - then it should produce a uniform rE of 0.862. But it doesn't. I get a uniform rE of around 0.8 only if the 3rd order gain is decreased to something like 0.1 relative to the zero order one. The same test for 2nd order on a regular dodecahedron works perfectly, giving a uniform rE of 0.775. As does 3rd order horizontal on an octagon with rE = 0.924. I've gone through the code at least 10 times and can't find any error. Is there something wrong with my assumption that this should work ? TIA, Ciao, -- FA There are three of them, and Alleline. _______________________________________________ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound