Hello everyone,

We’d like draw your attention to this Python implementation that we’ve been 
working on. It’s an implementation of real-time binaural rendering of signals 
from rigid-sphere microphone arrays. So far, it only supports head tracking 
along the azimuth, but there is more to come. 

If you follow the quickstart instructions, you will hear a binaural rendering 
of a (casual) Eigenmike recording.

---- Sorry, we’ve only tested things on macOS so far. ---

ReTiSAR is the real time version of 
https://github.com/AppliedAcousticsChalmers/sound_field_analysis-py, which we 
used, for example, in

        J. Ahrens, C. Andersson, “Perceptual evaluation of headphone 
auralization of rooms captured with spherical microphone arrays with respect to 
spaciousness and timbre,” in J. Acoust. Soc. Am. 145 (4), April 2019

Hannes Helmholz, the author of ReTiSAR, will have a paper at the upcoming IEEE 
ICASSP and Forum Acusticum in which he used it. 

There is certainly some overlap with the functionality of other existing 
implementations. ReTiSAR was our first attempt to write a heavy-processing 
application in Python. We were hoping to avoid some of the programming overhead 
that we have been facing in C++ with SoundScape Render 
(http://spatialaudio.net/ssr/). We are happy to conclude that Python together 
with the JACK Client for Python 
(https://github.com/spatialaudio/jackclient-python) is indeed a viable 
framework. We can’t clearly conclude on if or by how much Python makes it 
easier. 

Currently, ReTiSAR can render 12th order on an iMac Pro. In this case, we 
emulate the microphone signals by convolving a single-channel signal with the 
measured impulse responses of an according array from the Cologne dataset 
http://audiogroup.web.th-koeln.de/wdr_irc.html (we are not aware that a 
12th-order array exists; we don’t have one in any case). Note that 12th order 
corresponds to significantly more than 100 channels. Admittedly, ReTiSAR needs 
a long block length of 4096 samples or so for this, which makes the head 
tracking laggy. But it still does it without interrupts. Eigenmike recordings 
(4th order) can be rendered with a block length of a few hundred samples on a 
Macbook.

So, here is the first public version: 
https://github.com/AppliedAcousticsChalmers/ReTiSAR 

We would like to highlight (and thank for the fact!) that we received funding 
and other support from Facebook Reality Labs, particularly from Ravish Mehra, 
Philip Robinson, David Alon, and Sebastià Amengual Garí. 

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
[email protected]
https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe here, edit 
account or options, view archives and so on.

Reply via email to