Documentation for Project #2 - Ambisonic Soundscapes

Recording with an ambisonic microphone is different from recording 4 or 8 channels with individual mics. Two major differences: 1) the 4 microphones and each respective preamplifier must be calibrated for equal output for a known sound pressure, and 2) the gains of each channel of the line amplifier/analog-to-digital (a/d) converter (interface) must also be calibrated.

The second point also means that there is no "gain riding" during recording. One must set the gain in anticipation of the expected soundfield volume. For example, when I recorded the soccor game at the UT Intramural Field and was in the parking lot about 40 meters from the players, I set the interface gain for a zero decibal (db) gain. However, in the church sanctuary and within 5 meters of the horn and piano, I set the gain at -10db. I chose these values from looking at the zero levels on my computer and leaving a 6db margin of safety. Running out of headroom on an a/d converter is not a pretty sound and cannot be "fixed in the mix." Although the preamplifier gains are quite stable, I calibrate the interface before each recording session because of the knob adjustments.

I am recording in what is known as B-format (introduced in Project 1). Level adjustments in any of the 4 outputs would skew the directionality of the surround pattern. In fact, one can, with the right system or software, "drive" and rotate the soundfield. One ambisonic component not explored in my projects thus far is the vertical axis.

I edit surround recordings by listening in stereo and looking at the waveforms. Through this project, I have not had a 5.1 system for edit monitoring.