The 2020 project is the most ambitious ever undertaken at Aphex Systems. Virtually every detail of the hardware and software was designed in-house by our small but seasoned engineering team.

	Marvin Caesar, President

I want to give credit to the people of my engineering team who brought this project to life. Each individual shared in the responsibilities of electronic and mechanical design, autocad drafting, schematic capture, prototyping, testing, software development and documentation. It is this flexibility which allowed our small staff to complete the entire project in-house.

Many thanks are owed to Marvin Caesar, the President of Aphex Systems, pictured at right, for sharing a vision of the importance of this product and persevering through the costly and lengthy development process. We literally shut down all other design projects at Aphex in order to allocate the maximum resources to the FM Pro project.

The Model 2020 FM Pro from Aphex Systems Ltd. is a complete FM broadcast audio processor self-contained in a single chassis. All functions are integrated within one interactive yet stable and optimized processing framework. Over 12 Aphex patents in audio signal processing technology are incorporated in the FM Pro making it the most innovative broadcast audio processor of all time.

The unit is housed in a two rack space chassis, much smaller and more convenient than competing equipment. It's self contained power supply automatically accepts any mains voltage from 80 to 268VAC 50/60Hz, making it suitable for use anywhere in the world. To further internationalize the 2020 MkIII FM Pro, we designed it to operate with either 75 or 50 microseconds pre-emphasis, user selectable through panel controls.

The model 2020 is NOT just a combination of previous Aphex processing units. Every stage of processing is a completely new design. We transplanted some of our best circuits into the model 2020, such as our high performance servo balanced input and output stages and our world famous VCA1001 voltage controlled amplifier, but the processing algorithms are substantially new. Every stage of processing is crafted to work seamlessly with the next so there is no sense of competition or "fighting" for control between the leveler, compressor, and limiter elements. At every stage, the sonic effect of the process was carefully tuned for the minimum objectionable artifacts and maximum performance. The design process started with architecture based on years of broadcast engineering experience, and all the processing techniques were reinvented to excel at their function. No standard "cookbook" processing circuits were even considered in the design. Systematic research and design experiments were conducted in a lab equipped with studio quality monitor speakers and the best lab instruments known to man: human ears.

Extensive research was conducted also in the field on real radio stations to further improve the design in ways substantial and important to real-world broadcasting. Importantly, no multiband clipping is used in the model 2020. The multiband clipping incorporated in several competing audio processors causes grunge and listener fatigue. It also stimulates and exacerbates the effects of multipath distortion and can actually degrade your coverage area. We developed techniques to gain comparable and even greater loudness with far less garbage placed in the sound. Equally important is that the model 2020 without composite clipping can achieve dominating loudness. Many broadcasters have had to add the crunch of an overdriven composite clipper to get the last ounce of loudness they needed with other audio processors. Unfortunately, such a use of composite clipping heavily trashes up the stereo subcarrier and can greatly increase the effects of multipath distortion, actually decreasing the effective coverage area of the radio station. Without composite clipping, the model 2020 helps you attain a wider, louder, and cleaner coverage area.

Through digital control, the unit can be operated from the front panel and from remote control software running on a PC Windows platform. The remote control can be operated from long distances using dial-up telephone lines. In this way the broadcaster can adjust processing from an idealized listening environment such as a studio, home or car while the processor is permanently installed at the transmitter site or central rack frame. Password protection can be set to allow only approved operators access to the controls.

One may ask why we designed this product around analog circuits instead of the modern trend toward digital signal processing (DSP). The fact is that, within the present state of the art of DSP, analog signal processing still sounds better. The chief advantages of digital signal processing remain its consistency, relative ease of manufacture, and natural interface to computer control... not its sound quality. Additionally, DSP algorithms presently available are relatively crude compared to the highly developed and mature art of analog processing technology, particularly the advanced processing used in the model 2020. To marry the best of both worlds, we developed the model 2020 around our advanced analog circuits and exhaustively researched ways to cleanly tune and adjust them by digital control. This ties all the convenience and feature options of computer control together with superiority of analog audio processing. A carefully designed high definition (20-bit) digital I/O interface is provided for users who want to place the model 2020 within a digital audio framework. The sonic results of converting digital to analog, processing in the model 2020's analog domain, and converting back to digital are demonstrably superior to comparable all digital audio processors.

Optional professional standard AES/EBU I/O is available for digital audio systems while analog I/O is standard. Both methods of I/O can be selected for use at any time, and a defeatable autoswitch feature is available to automatically switch to the analog input should the digital audio data become corrupted or fail. The digital audio input accepts standard sample rates from 32KS/s to 48KS/s while the output sample rate may be synchronously slaved to the input or set up independently at any chosen sample rate. Both input and output are 20-bits wide. Because of the very high-grade converters and the extensive reconstruction afforded by the intermediate analog audio processing, the model 2020 serves as an outstanding sample rate converter. An example of how this applies is interfacing between a digital audio STL and a digital FM exciter. Digital audio can be supplied to the model 2020 from a typical 32KS/s 16-bit digital STL while the digital output from the model 2020 to the exciter can be set to 48KS/s (with 20 bits) for noticeably better on-air audio quality.