Each mic or speaker records a standard test file. It begins with a tone for level calibration, A 440 Hz at -12 dB below full scale digital clipping. Next, white noise and pink noise, 10 seconds apiece. That’s followed by short musical clips. Each sample is intended to bring out different aspects of a device’s character. Hammer dulcimer and mbira emphasize transients and cross modulation. Sax solo for even response, noise and room tone, since this is after all recorded in my not-so-perfect office. Female and male voices, because we record more of that than just about anything. Children’s choir for intermodulation. Synth with heavy subterranean bass and white-noisy percussion.

The “standard speaker” is a pair of heavily modified Radio Shack Linaeum mini-monitors in my office. The mics are set one cubit away from the speaker (fingers touching the speaker, elbow touching the mic, 18.5” or 47cm for the scientists in the audience), far enough away for woofer and tweeter to meld, close enough to be in the near field. For speaker tests, the standard mic is Neumann KM84. They are neutral, cardioid, and stable.

You may want to load a couple of the files into your favorite audio editor such as audacity, Garage Band, Logic, or Pro Tools and A/B them.