> Hi, all > > I think there are somebody have worked about beamforming of spherical > array. > Now I want to design the beampattern steering to certain direction and get > the signal from that direction (Sptial filtering). > I get the 32 channels signal p from Eigenmic. Get the Pnm of different > orders according to the spherical fourier transform. > And then weighted and sum all Pnm with the weight, > Wnm=Ynm(omiga)/(4pi*i^n*bn). Ynm is the spherical harmonics and n is the > order of it, omega is the look direction, and bn is the mode strength to > equalize different mode. In my simulation there is a > two voice in the directions 0 degree and 135 degree respectively. When the > order is set n=2, I get the output signal( signal from > the steering direction 0 degree). The voice from 0 degree is enhanced and > the voice from 135 degree is weakened. The > spatial filter takes effect. But when I set the order n=3, the signal is > heard just like the one of order n=2. And n=4, the signal > is even worse than the one n=2. It don��t have any spatial filter effect. > > Can anybody help check what is the main cause of the problem? Thanks > advanced. >
You don't say what happens in first order : is that normal ? Certainly for HOA soundfields that work in first order, but breakdown in higher orders (except for (?) 0 degrees sources (I haven't checked the maths for that, but I think that is the case)), then ... the first thing I'd check is Condon-Shortley*. My apologies if that is stating the obvious and you have already done so. Regardless, do say what happens in first order ... (?). Michael See, for example <http://mchapman.com/amb/reprints/Symmetries.pdf> at pages 5 and 6. _______________________________________________ Sursound mailing list [email protected] https://mail.music.vt.edu/mailman/listinfo/sursound
