SAMSON BOX published on 27/04/2012 da Alex di Nunzio
The SAMSON BOX is a digital synthesizer driven by a computer. Has been designed and built by Peter Samson for the Stanford University. In the literature it is also called SYSTEMS CONCEPTS DIGITAL SYNTHESIZER even though commonly named SAMSON BOX, by the name of his inventor.
Brief History – Was developed by Peter Samson of Systems Concepts, San Francisco, on specifications provided by the Center for Computer Research in Musical and Acoustics (CCRMA) of Stanford. Design work began in the early Seventies; for the 1972's International Computer Music Conference at Michigan State University there was a first presentation of its architecture and potentialities. Construction work, on the other hand, started between 1972 and 1973, so that a prototype was installed at CCRMA in 1977 and used for the all Eighties in music compositions and experimental research.1
Past mistakes – The SAMSON BOX project was based, first of all, on some considerations done on other computer music systems developed in the previous years. One was the most significant: Peter Samson has underline how important was realize a new system with a computer dedicated to musical applications, rather that shared among many users as well as for many different purposes, as did during the previous years, without opportunity to satisfy the specific needs of the composers. All that, according to Samson, would be the reason of poor aesthetic quality of the earliest musical works.2 Nevertheless, to the detriment of this, Peter Samson points out the great advantage of the digital synthesis and its application inside the synthesizers that, compared with those analog, allow to go over various limits such as the difficulty to monitoring the many modules, lack of accuracy for the parameters management, and a limited flexibility.3 From a general viewpoint, finally, we can say that this project follow that research line finalized to develop synthesizers, or general hardware, conceived as computer controlled devices. An important previous, in this sense, it's the GROOVE by Max Mathews and Richard Moore.
Research Goals – The aims set by this research concerned, first, the computation speed; to be solved with the design of a dedicated computer. Strictly linked with it, was that of real time: in this way the SAMSON BOX would given to composers the possibility to obtained a instant feedback. The question of the real time was, for a long time, one of the most important topics for computer music community, and for this was considered the first goal of the Peter Samson's project. Not less significant was the openness level of the synthesizer. SAMSON BOX, in fact, that was designed for implementing the common synthesis (such as additive, subtractive and FM), must be able to complement many innovations possible. For this reason Peter Samson, in his paper of 1980, talks about a synthesizer General-Purpose, i.e. specifically dedicated to the music but no designed for a specific use; an open system, so, capable to implement all news developed by researchers. All these features were taken in account for design the SAMSON BOX, although between the prototype of the 1977 and that described by Peter Samson in 1985 there were many differences, so that the last version had less functionality of its predecessor.
Previous Projects – Peter Samson has underline that when began to work for the SAMSON BOX, there were not others projects to refer. GROOVE, indeed, was a hybrid system (i.e. digital control, and analog synthesis), while the SAMSON BOX was designed as a fully digital synthesizer. For this reason took in account all the sound synthesis techniques available, even very different from that digital (traditional instruments, analog synthesis, speech synthesis, and so on).4
Structure – The SAMSON BOX had the same modular structure already adopted with many other systems. Specifically, the hardware consisted of a PDP-6 computer, at least in according with the system configuration presented by Gareth Loy in 1981.5 The PDP-6 was not a computer dedicated to the computer music, but a machine utilized for managing many other device (DAC and ADC converters, and the FRMBOX synthesizer by Richard Moore) and for the care and feeding of the SAMSON BOX.6 For this, in the overall structure, the PDP-6 was dependent of a DEC PDP-10 computer of the Artificial Intelligence Laboratory (AIL) at Stanford University, which it was linked through a particular I/O connection.
Modules – Previously we said that the SAMSON BOX was structured in modules, in accordance with the best analog tradition. The SAMSON BOX was equipped with three different modules, named Processing Elements: generators, modifiers and delay units.7 In detail, the synthesizer was provided of 256 generators, 128 modifiers and 32 delay units, all can run simultaneously. The names not suggested functionality too specific; in fact, for example, generators could be used for modify a sound signal and, vice versa, modifiers could be used as generators. As the case of those modifiers utilized as pseudo-random generators useful for the generation of white noise.8
Generators – Generators, from a functional viewpoint, were divided in two sections, so that could be used as oscillators or envelopes. Generators, in addition, were able to generate many waveforms (sawtooth, sine, square, pulse, etc.), while envelopes could be used for controlling the amplitude.
Modifiers – As the generators, even modifiers could carry out many operations: generate white noise, realize amplitude envelopes, mixing functions, and so on; without association to a specific synthesis technique.
Sampling Rate – The SAMSON BOX was provided with a changeable sampling rate, in according with specific needs. What influenced the sampling rate was the number of Processing Elements run simultaneously: a number that the user could decide arbitrarily.
Input/Output – I/O device implementation is carried out through many years. A technical report of an early prototype shows us a synthesizer equipped with 4 DAC; even Gareth Loy remember that initially were not planned ADC converters.9 In 1985, on the other hand, Peter Samson talks about the final version of his synthesizer and describes two different I/O devices: 16 analog channels (capable to operate as DAC or ADC) and digital I/O devices, for saving and handling data in digital form. About of 16 channels, we must clarify that DAC and ADC converters needed of Low-Pass filters for filtering the input and output sound signal. These could be programmed and even bypassed, if the user wanted to use external filters. Digital I/O, finally, allowed handling data between several devices without loss of information.10
Double Power – One of the most interesting feature of these digital devices, probably, was another. The I/O peripherals, in fact, allowed to doubling the power of the synthesizer through a signal processing to be carried out in two or more steps. Above we said that the SAMSON BOX was equipped with 256 generators and 128 modifiers. If let's assume that a first step of generation concluded with a sound obtained by utilizing all processing elements available, the output could be saved in digital form and then recalled to be added to another sound obtained with the same functionality. The final output, so, was the result of a signal processing done with 512 generators and 256 modifiers; this, clearly, allowed obtaining timbres of high quality.
Utility – During its use over the years, the SAMSON BOX was supported by other software that have integrated and increased its functionalities. This is the case of PLA (even if it has not been used for a long time) and, particularly, of the MUSBOX, a compiler dedicated for the SAMSON BOX.
Musical Works – The most interesting item, however, is the effective use of the SAMSON BOX for the composition of musical works. From this viewpoint is useful mentioning the collection DINOSAUR MUSIC published by Wergo. This Compact Disc collects some works realized with SAMSON BOX, MUSBOX and PLA by Chris Cafe, William Schottstaedt and David Jaffe. Of the latter we remember the notable Silicon Valley Breakdown, awarded at the 1982's Biennale in Venice, and at the 1983's Newcomp Contest.