So how does the microphone work? The basic premise is simple enough: the microphone simply counts the number of air molecules flowing through a portal per unit of time, and determines the direction(s) of their flow. From this data, a digital data stream is generated. Pretty straightforward, eh? It's the engineering that is mind-boggling here. The microphone is an extremely small spherical lattice (the diameter is .37 microns) that is "grown" (the actual process is proprietary information, but Dr. Micheals did indicate that it was biological in origin) from an etched semiconductor chip. This lattice has 360 portals around its surface, each just large enough to permit air molecules to flow freely through it. Imbedded in the semiconductor substrate are charge-coupled micro-capacitance lines that run from the outer rim of each portal to its inner rim. As an air molecule flows through the port, its magnetic charge causes an opposite-sign electron flow in the micro-line. Thus, every negative electron flow effectively represents a "positive" molecule and vice versa. Such flows are detected as electrical impulses that are the raw digital data stream. The charge-coupled lines for each of the portals are integrated in the lattice into a semi-conductor "neural network." A global variable-rate clock employing Liburk's Theorem of Inverse Quantity (TIQ) ("Less is More") in alternation with his Theorem of Opposite Quality (TOQ) ("Bad is Good") causes the neural network to transformatively generate, in a quality/quantity matrix, what Micheals calls "Finite Unit Criterion Knowledge" that is presented as multiplexed optical data to the lattice's output, an optical strand that connects the output to its processor/power supply.

The data flow is prodigious: 1,110 Million Instructions per Second (MIPS) in 16 parallel strands, driven by a 150 Megahertz clock in the processor! The processor is a proprietary stand-alone unit connected by SCSI to either PC or Macintosh operating systems, which in turn provide the operating interface for the microphone. The processor receives the data from the microphone and truncates it to a usable 24-bit range determined by the user. This is in place of the normal microphone attenuator. Instead of adjusting the microphone's level, you simply select the portion of the microphone's range you wish to listen to! The notion of Equivalent Input Noise doesn't exist here! A single bit is equal to the sound of one molecule passing by (i.e. approximately -22 dB SPL!). The portals saturate at approximately one tenth of an atmosphere: 170 dB SPL! In practice, using the calibration menu provided, you select the peak level (or the noise floor, if you prefer) you wish to record. The computer simply selects the 24-bit (144 dB) range below (or above) your specified peak (or floor). Sampling rate is selectable at 96, 48 or 44.1 kHz.

Data is received and processed into the appropriate signal set, truncated to the desired bit count (LSBs are deleted if you calibrated to peak level; MSBs are deleted if you calibrated to noise floors - we tried both methods and it didn't seem to matter; there's obviously enough dynamic range for everybody!) Finally, data translators convert the signal(s) to AES/EBU, S/PDIF, and (via conventional, yawn, 20-bit DACs) to stereo analog XLR (15 dB below 0 dBFS equals +4 dBu) and to a DIN plug carrying the discrete Dolby 5.1 signal. Obviously, the days of mic preamps are numbered!

The processor/power supply is a 1 rack-unit box with analog and digital XLRs and RCA (for the S/PDIF) jacks and a SCSI connector for the computer front-end, in addition to the optical cable connector and splitter for future Mikrodigi products utilizing the full capability of 32-bit, 96 kHz. digital audio in Mikrodigi's propriety format, Multiplexed Optical Transmission Holographic Analysis Finite Unit Criterion Knowledge®. The processor is capable of battery operation for field use. When this operation is selected, the processor selects pre-determined defaults for operation, so you don't have to have the computer along in the field (although it worked just fine with my Powerbook 100). You just set the processor up for your desired configuration before you head out. Simple!

The optical cable is 24' long and permanently attached to the lattice's housing. The spherical lattice itself isn't visible to the naked eye (I only saw it under the microscope at Mikrodigi's labs). It is fixed at the end of a tapered wand and is, as you can imagine, extremely delicate! Naturally, you don't tap on it or blow in it to see if it's working, particularly when it is hooked up to the monitors! The unit doesn't need a wind-screen, as response goes to DC and the mic can't really be overloaded - the processor permits high-pass filtering to break up wind artifacts if you so desire.