Engineering the Case

As stated earlier, signals at very high frequencies behave strangely. Our new skin was not only loss-y; at very high frequencies it was behaving simultaneously as a conductor, a reflector, an inducer, and a transmitting antenna, if not more. One of the first steps we took to reduce the unpredictability of the behavior of high frequencies in the Firewall was to work on the material for the case. This proved highly audible and was not overly difficult to solve.

Wood was eventually selected: natural wood, due to the fact that all metal cases create a resonant chamber which will haphazardly redistribute the high frequencies within the enclosure, creating nodes that are arbitrary in both nature and location. These nodes not only intermodulate amongst other nodes, they also reflect off the metal case itself and are induced back into the filter. In a word: noise.

The use of a wood enclosure is remarkably advantageous to audio performance. Wood enclosures do not reflect; instead, wood tends to naturally absorb a spectrum of this noisiness which the ear easily appreciates. Contrary to common belief, it's not an acoustical resonance issue as much as it is a high frequency absorption spectrum issue.

When we speak of a "natural" sounding filter, we are in fact experiencing a lack of high-frequency noise reflecting haphazardly into and within the filter. (Incidentally, there's nothing altogether "natural" about 50 or 60 Hz. All audiophile problems and solutions are man-made. Let's not forget that.)

Having established a good-sounding solution for the enclosure, we were able to finally move on to more exciting and much more challanging territory.