acoustic separation for stereo field recording

Acquiring a stereo image requires the acoustic separation of two microphones. The separation arises from a combination of materials, geometry, physical space, and the directional nature of the capsules. People who record natural environments are often highly vested in acquiring accurate stereo images, and are a great source of information and innovation regarding separation strategies. There is lots of material on the internet regarding this; links to some of my favorite sites are listed at the end of this post.

Microphones with independent spatial buffers can accommodate a wide variety of physical geometries and proximity.  For this reason, I particularly liked the PIBO concept.  This becomes highly relevant in a human environment, where a soundfield arises as much from the field (buildings, streets, planters, open spaces, humans, etc) as from the things creating the sound (humans, machines, wildlife, etc). The shape, size, and materials of the buffers can also vary, thus further affecting recording of the soundfield and its character.

In the interest of time and convenience, I used a flush-cut (aka “pull”) saw to hack an American-style foam football into 4 pieces. I cut 2 disc-shaped sections from the center, and in the process obtained 2 cone-shaped sections from the ends. The foam appears to have an open-cell structure, is highly resilient (springy), and fairly dense. I drilled holes in the center of each section with a spade bit for the microphones, and then hot-glued some cylinders of plastic mesh (gutter guard material) to each section. The gutter guard material is really inexpensive, readily available, and relatively rigid. The cylinder shape provides a spatial buffer to the microphone while supporting various wind-mitigating covers. A piece of PVC pipe or conduit can be used a handle (or for mounting to a tripod), and to help secure the microphone to the baffle. The disc shape seems to provide a more pronounced stereo separation, and has the added benefit of easily accommodating discs other materials (plastic, wood, metal, foam, etc) at the microphone boundary.

Like most microphones, the EM 172 capsules need to be protected from the wind. Screens of faux fur and other fabrics are common, though my internet research failed to reveal any decisive or conclusive strategies or materials. After some quick experimentation with a few fabrics (faux fur, cotton dishcloth, old t-shirt), I found that they all offered some protection from wind when combined with a buffering space (the space between fabric and microphone). Larger spaces provided better wind protection, at least up to a point. Fabrics that performed well with my plastic-mesh cylinders often performed quite poorly when pressed against the face of the capsule. The faux fur tended to muffle the sound a bit more than the others. I’ll likely include a more rigorous study of this in a separate post.

These microphone assemblies are quite conspicuous; they look like somebody is recording something, or at the very least is up to something unusual. Faux fur always attracts attention, particularly when its on the end of a stick. So, they’ll be great in an environment with subjects that don’t include humans or that depends on their lack of awareness. For that, a more inconspicuous adaptation is required, which will also be the subject of another blog post.

The following audio files provide a qualitative perspective for how the cone and disc shapes compare with regard to the stereo field. In each recording, a series of sound sources (box of matches; pig wind chime; large metal bowl) are moved counter-clockwise around the mics, starting and ending in the “front” of the pair. The sources are then moved up, between, and behind to give a sense of the how the array captures the vertical field. There are also 3 different wind-mitigation materials used in the recordings. The recordings were made outside, so there are a variety of background noises to listen to as well, which are sometimes fantastic for comparing the windscreen materials. The Scarlett had the same gain for each, and none of the recordings have been normalized. They were converted to the .mp3 format with Audacity.
Foam disc:
with dishcloth windscreen

with tshirt windscreen

with faux fur windscreen

Foam cone:
with dishcloth windscreen

with tshirt windscreen

with faux fur windscreen

Internet Resources:
general and overview:
Rob Danielson has a series of blog posts that summarize different design strategies for stereo separation, including his own.

SASS array:
Vicki Powys has a fantastic blog post about a DIY SASS array, with lots of pictures and recordings.

I came across this rather interesting site devoted to binaural recording and related topics; sometimes having a thick head is a good thing.

Olson plates:
Curtis Olson has incarnated several configurations of stereo arrays utilizing different materials and geometries.

According to author Andrew Skeoch, PIBO is an acronym for “Pair of Independently Baffled Omnis”.


Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Leave a Reply

Your email address will not be published. Required fields are marked *