Hello everyone, This post aims at anyone who has data on sound source directivity available.
Stefan Bilbao of the University of Edinburgh and I published a paper at this past ICASSP on a method that allows for computing a complete spherical harmonic (SH) representation of a sound source directivity independent of how much (or how little) information on the directivity is available, i.e., independent of how many data points are available. In a nutshell, we impose the magnitude of the directivity that is known on a spherical surface around the source (which we term the “finite-distance signature”) onto a spherical carrier wave. There are no limits in terms of how much interpolation we involve as we’re using only the magnitudes. (We add a minimum phase in the end to design the time-domain structure of the directivity). There is quite a bit of directivity data available as polar diagrams or other magnitude-only data or data that are available only in the horizontal plane and the like that do not allow for a complete conventional SH representation to be computed. We can now revive these old data and make them suitable for applications that require such complete SH representation (like FDTD simulations). --- Here’s the thing: We would like to avoid browsing papers and read magnitude directivities from polar plots with a ruler. We would rather like to ask all of you who have such data to submit them to us electronically. --- We would then process the data and make them available online. Here’s a first outline of what this could look like: https://github.com/AppliedAcousticsChalmers/sound-source-directivities/ So far, the repository contains only those data that we used in the papers that we wrote about the method (so far, only the ICASSP one has been published, so please stay tuned!). There’s a loudspeaker, a singing voice, and a handful of musical instruments. The long-term goal is that this repository can serve as a tool to help finding out what kind of data representation is most useful for what application. A complete SH representation can be a good starting point as it allows for any other representation to be computed. But we haven’t really understood yet what people do with directivities. We might end up concluding that some other representation is more useful. But we don’t know yet, and we’d like to contribute to resolving this question. (Just to clarify, we are aware that formats like SOFA exist. We’d primarily want to know what we need to put into the SOFA file…) Thanks for your support! Best regards, Jens -- Jens Ahrens Associate Professor Division of Applied Acoustics Chalmers University of Technology 41296 Gothenburg Sweden +46 (0)31 772 2210 http://www.ta.chalmers.se/people/jens-ahrens/ _______________________________________________ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe here, edit account or options, view archives and so on.
