sonic surrounds

There seem to be many ambisonic microphones popping up around the place - both high-end and mid-range commercial microphones, as well as home-built DIY microphones.

Here are some photos and links.
(all images’ copyright owned by their respective manufacturers).

The premier ambisonic microphone range by Soundfield is soon to include a (presumably cheaper) A-Format microphone, the SPS200 (pdf link).

AGM Digital are redeveloping their MR1, originally developed with Danish Pro Audio using matched B&K 4011 capsules.

Core Sound are about to release their TetraMic a-format ambisonic microphone.

On the DIY side:
In 2005, I made my own DIY ambisonic microphone using Panasonic electret capsules, which turned out to be great quality for a cost of around $300.

Also in 2005, Etienne Deleflie built his Sound Thief ambisonic microphone using Rode NT5 microphone parts (capsule and phantom preamps).

Earlier, Henry J. Walmsley had already built his own ambisonic microphone also using Panasonic capsules.

In late 2006, I heard of two more new DIY ambisonic mics:
Rafael Kassier’s mic built from Oktava capsules:

and Paul Doornbusch’s beautifully constructed DIY mic built with Transound capsules:

Probably you could find out about more diy ambi mics by joining the Sursound list.

here’s a screenshot of the Pd interface that I developed to perform for Electrofringe 2006.

click on the image for full sized version:

I made heavy use of the (now quite usable) Graph on Parent feature - so that I could instantiate abstraction subpatches that created each of those GUI blocks in the patch.

Blocks aren’t really well labeled in this screenshot, but from left to right, top to bottom, these are the abstractions I developed and used in the patch:

• on the first row:
• a timer (clock), so i know how long i’ve been playing
• a “heartbeat” synth i made which emits a nice low bass pulse
• then three instances of the same ambisonic audio file player (that plays back my 4 channel ambisonic field recordings, and decodes them to 5 channel surround format)
• a single instance of a stereo audio file player
• on the second row:
• two instances of my 5 channel granulator - based on an improved version of the stereo granulator i’ve released elsewhere on my website
• a surround glitch/pop generator based on filtered single sample impulses with controlled stochastic triggering and panning
• and on the final row:
• a CPU load meter
• DSP switch
• mixer for all the abstractions, grouped into foreground and background sounds
• and the 5 channel output and recorder - which was the last thing i implemented, and seemed unreliable, so i didn’t use it to record my performance live. i’ll record a version sometime soon.

The Institute for Computermusic and Sound Technology ICST releases version 1.1 of the ICST Ambisonics Externals for Max/MSP, under the LGPL.

The objects are:

• ambiencode~ - encode n audio-channels in up to 3rd order ambisonic b-format
• ambidecode~ - decode up to 3rd order ambisonic b-format to n audio-channels
• ambimonitor - graphical user interface object for sound-source placement
• ambicontrol - control sound-sources in the ambimonitor in a variety of ways

They can be downloaded here, with sourcecode in the packages.

Get Beehive at CSIRO’s website.

The BeeHive is a Java3D AudioDevice class that renders moving sounds over eight speakers arranged in a cube around the listener. It is a truly 3D spatial sound system with an up/down dimension that is not available in commercial surround sound systems.

You can use the Beehive AudioDevice to augment your Java3D applications with a minimal amount of extra coding. Beehive transparently maps Java3D sounds from the scenegraph to the speaker array using the Vector Base Audio Panning (VBAP) algorithm and the ASIO interface to the soundcard. All you need to do is make sure you have a soundcard that supports eight speaker channels, such as the Hammerfall DIGI9652.

via Stephen Barrass on the ICAD mailing list.

LCS Matrix 3
- Last I looked at this, it had some form of 2D panning to the surface of arbitrary 3D speaker arrays - depth was not simulated, nor was acoustics.

Lake Huron Digital Audio Convolution Workstation
- I think it’s still available, but there seems to be no web presence for it anymore.
update: there’s a webpage here, though it doesn’t seem to be linked to from anywhere.
This hardware is rather old now though, so while it can still do the longest convolutions with the lowest latency in a turnkey hardware product, it is weak in general purpose spatialization even compared to open source software solutions.