I agree with Fons here. We have used essentially this approach when we
needed to play back Ambisonic material on The Morning Line (which is very
irregular and needs preferably to use VBAP) right from when we first set it
up back in 2008 and you definitely won't find it easy to get it work
optimally - manual tweaking is pretty well always needed. However, at this
stage, there probably isn't a better approach for low order Ambisonics,
either automatic or manual.
Dave
On 19 April 2013 23:33, Fons Adriaensen <f...@linuxaudio.org> wrote:
> On Fri, Apr 19, 2013 at 10:12:20AM -0700, Aaron Heller wrote:
>
> > F. Kaiser, “A Hybrid Approach for Three-Dimensional Sound
> Spatialization,”
> > Algorithmen in Akustik und Computermusik 2, SE, May 2011.
> >
> http://iaem.at/kurse/winter-10-11/aac02se/2011_Kaiser_SeminararbeitVersatileAmbisonicDecoder.pdf
> >
> >
> > The toolkit reads speaker locations from CSV files (and other formats,
> > including ambdec presets) and writes out presets files for Fons' Amdec
> > decoder. There's also an initial implementation of a Faust backend, that
> > produces decoders that can be compiled to VST, Supercollider, Pd, MaxDSP,
> > ... (see http://faust.grame.fr/ for more about Faust).
> >
> > AllRAD is a hybrid ambisonic/vbap technique, especially suited to
> irregular
> > arrays. The idea is you design a decoder for a regular array (in this
> > case a 240 virtual speaker spherical design) and then map those signals
> to
> > the real array using Pulkki's VBAP.
>
> This is one of the many methods I tested, and rejected. If applied
> 'blindly' it can easily produce rather suboptimal results (*) - as
> would any method that in the end amounts to linear combinations of
> the outputs of a 'regular layout' decoder. Things improve if the
> VBAP gains are adjusted manually, but that more or less defeats the
> purpose of an 'automatic' method, and even results obtained that way
> are suboptimal.
>
> The problem behind this that adding VBAP (or any other) pannings of
> (partially) correlated signals will not produce the correct result
> - at least not in the frequency range where rE magnitude and direction
> are the metrics. There are solutions to this, but they are not simple.
>
> Ciao,
>
>
> (*) Try for example a 5th order horizontal decode to a regular ring
> of 12 speakers, then remove one of them and it replace by a VBAP
> panning into the two nearest ones.
>
>
> --
> FA
>
> A world of exhaustive, reliable metadata would be an utopia.
> It's also a pipe-dream, founded on self-delusion, nerd hubris
> and hysterically inflated market opportunities. (Cory Doctorow)
>
> _______________________________________________
> Sursound mailing list
> Sursound@music.vt.edu
> https://mail.music.vt.edu/mailman/listinfo/sursound
>
--
As of 1st October 2012, I have retired from the University, so this
disclaimer is redundant....
These are my own views and may or may not be shared by my employer
Dave Malham
Ex-Music Research Centre
Department of Music
The University of York
Heslington
York YO10 5DD
UK
'Ambisonics - Component Imaging for Audio'
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20130420/daeeee26/attachment.html>
_______________________________________________
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound
