Hi,
i’m wondering what the speaker setup in the spat.tut.17 is…
It looks like it permits an accurate comparison between hoa and vpap, is that correct ?
Does it mean that this configuration follows a specific geometrical form, which would make sense in a real world of 24 speakers ?
I’ll have the opportunity to test such a configuration so i’m curious about it !
Thanks
Damien
Hi Damien,
The speaker setup certainly makes sense in a real-world configuration:
this is what we have in one of our studio. It is rather well suited for HOA.
The primary goal of this tutorial was to introduce the “virtual speakers” approach
(which btw can also be achieved with the spat.virtualspeakers~ object)
rather than to offer an “accurate comparison between hoa and vbap”.
(“accurate” in which sense ?)
T.
Hi T.,
thanks for your reply, in fact I wondered wether this configuration was a platonic solid or not.
But it’s usefull to know that it’s well suited for HOA !
About the comparison with vbap, forgive my lack of experience in this field, I just want to be able to test differents spatialization methods for a given configuration, including HOA an vbap. Does that make sense ?
Hi,
This configuration is not a platonic solid, but it is rather regular (in a spherical sense), and therefore provides a HOA decoding matrix that is well balanced.
Comparing VBAP and HOA makes sense, and you should do it if you have the opportunity! Just be aware that the techniques are quite fundamentally different. (both having some pros and cons)
T.
Hi,
thanks for your help !
yes I’m aware of that, I also intend to compare with dbap in 3d (as spat.pan~ permits I believe) but I wonder if that make sense in a sphere.
The point is: in a spherical configuration, would dbap allow movements of a source that vbap wouldn’t ?
I’m thinking about the distance of a source to the listener, and it’s not yet clear to me if vbap panning remains on a spherical plan or allows for changes in distance.
D
Hi,
You might want to have a look at the ‘knn’ panning type. It is quite similar to dbap and works in 3D as well.
(i’m not 100% confident about dbap in 3D, however)
These (dbap or knn) would certainly generate panning laws that are different from vbap.
Whether this is really noticeable on movements of a source, hard to say. Try it out.
About distance: besides ‘knn’ and ‘dbap’ which are fundamentally based on distance (between virtual sources and loudspeakers), all other panning algorithms do not account for the distance of the source (only azimuth and elevation are used).
If you need convincing distance rendering, then you need to also include reverberation effects (see spat.spat~ and spat.oper). Panning alone is not enough.
T.
Thanks T. !