Learning Synthesis: Sequencers

Part One: A Brief History of Sequencing

Eldar Tagi · 10/06/20

Outside of the world of sound generators and noisemakers, few types of electronic music gear inspire as much excitement as sequencers. They come in many shapes and forms, sometimes as complex stand-alone units, other times as a feature in music-making hardware or software...and on the macro scale, they can be viewed as an electronic substitute for a composer’s old friends: staff paper, and a pencil.

We can define sequencers as devices and software applications capable of recording, storing, and manipulating music performance data. The extent of how much information can be handled by a sequencer varies greatly. Some are simple, only able to temporarily organize a limited set of notes, others are advanced multilayered machines that can manage every single nuance of a musical composition; and there is, of course, an endless plane of everything in between. This is the topic that we will be exploring in depth within this Learning Synthesis article.

Conceptual Origins

Book of Ingenious Device by Banu Musa brothers Book of Ingenious Device by Banu Musa brothers

Although we generally associate sequencers with the advancements of the recent technological era, the ideas of automated musical devices existed long before the industrial revolution was even on the horizon. How long? The earliest evidence can be traced back to the research of the Banu Musa brothers, scholars from the ninth century. In their “Book Of Ingenious Devices,” among many other things, they documented such instruments as an automatic flute, and hydro-powered organ. Around the 15th-century, barrel organs started emerging in Europe. They followed a simple design principle of rotating pinned cylinder with levers, not too different from the much more portable music boxes. At the dawn of the 20th-century, along with the rapid development of mechanization, similar technology could be found in inventions like player pianos, orchestrions, optoacoustic synthesizers, and not-necessarily-musical punched cards which served as the basis of modern computing. The thing that all of these devices shared was an ability to record and store data, which could be played back precisely whenever it was desired.

Electro-Mechanical Sequencers

While at this point it should be evident that the development of a sequencer-like device was long in the making (stemming from multiple sources and locations), it is fair to say that the electronic music research conducted by Raymond Scott in the mid-20th century was quintessential in helping the concept to evolve towards its modern form.

Scott was an American musician, composer, and inventor who pioneered a number of inventions from electronic musical ashtrays to the visionary proto-AI synthesizer, Electronium. Scott made a fortune licensing his music to many animated films produced by Warner Brothers, which gave him the opportunity to dedicate his time and resources to his innermost passion—electronic music.

He established the Manhattan Research project in 1946 and spent the following years creating numerous devices for the production and manipulation of electronic sounds. Sequencing musical pitches was of particular interest to Scott, and he developed several machines that could accomplish such a task. Although the majority of the time Scott’s creations remained in high secrecy, he eventually credited himself as the original inventor of a sequencer.

A young Robert Moog met Raymond Scott in the early '50s and quickly became an admirer of his work, consequently even working for him. When they met, Moog was building theremins in his basement, and it is safe to assume that the exposure to Scott’s ideas and equipment strengthened Moog's confidence to venture further into electronic music instrument design. Moog remained loyal to Scott and halted the addition of a sequencer into his line of electronic instruments until the point when the word spread beyond control, and others started producing their own versions of the concept. Eventually, Moog presented his version of a sequencer which abandoned Scott’s relay-based design in favor of stepping switches. This led to the creation of one of the most important analog sequencers, the Moog 960 Sequential Controller, as a part of Moog Modular Synthesizer.

Analog Sequencers

Essentially, an analog-style sequencer consists of a clock or pulse generator, a counter, and a set of potentiometers for dialing in voltage values per step, and a set of accompanying switches. The pulse steps through each potentiometer one by one, and once the sequence reaches the end it restarts. The output of the sequencer is commonly assigned to control the pitch of an oscillator, but it can be used to animate any other available parameter in the system.

With the Moog Modular Synthesizer becoming commercially available in 1967, the 960 clearly wasn’t the first sequencer around. The RCA Mark II Sound Synthesizer designed by Herbert Belar, Harry Olson, Vladimir Ussachevsky, and Peter Mauzey at the Columbia-Princeton Electronic Music Center in 1957 and its German cousin, The Siemens Synthesizer developed by Helmut Klein and W. Schaaf in 1959, were already equipped with sequencers—however, they still relied on perforated paper technology for information storage and retrieval. Also, Don Buchla completed the first 100-series modular system in 1965, which included 8- and 16-step sequencers (the 123 and 146, respectively). That being said, Moog can safely be credited for popularizing sequencers and establishing an early standard for synthesizer control, including the still widely used volt-per-octave pitch control standard.

Early Buchla sequencers didn’t have an internal clock generator, so a separate module needed to be patched in for the sequence to start running—usually these pulses came from the 140 Timing Pulse Generator, or could be manually produced by one of the system's touch keyboard controllers. Moog’s 960 Sequential Controller was more advanced in that regard: it featured a built-in oscillator as a clock, three rows of potentiometers spread across eight stages, manual triggers and inputs for resetting the sequence to any desired position, and a set of switches that could determine whether the sequence would keep playing, stop, or skip at any particular stage. Although the sequencer was only eight stages long, the length could be extended by chaining two or more modules together. This practice was fairly common until accessible technology allowed for the creation of sequencers with more than sixteen stages.

Another interesting Buchla-inspired, yet very original analog sequencer design came almost a decade later in 1976 in the form of the Touch Activated Keyboard sequencer module for the Serge modular synthesizer. Just as the name suggests, TKB was created as a combo sequencer and touchplate keyboard controller module, meaning that it could be used for both manual triggering of musical events, as well as automating them. Particularly unique feature to the sequencer was the Vertical Clock input designed to address preset voltages on a selected stage, sequentially switching between ABCD rows on every clock pulse. The controller also allowed users to change the direction of a moving sequence either manually or via voltage control, and it also contained an input for random stage selection. With Touch Activated Keyboard, Serge established a new kind of controller interface which lives on to this day in various incarnations, i.e. Make Noise Pressure Points, 0-CTRL, Verbos Electronics Minihorse, and more.

The Dawn (and Dusk) of DIMI

Erkki Kurenniemi Erkki Kurenniemi

While there were many fascinating developments happening in the analog realm of synthesizers, there was a person with an alternative vision of what the future of sound synthesis might be like. In 1961, a young Erkki Kurenniemi was hired to build Finland’s own Electronic Music Studio. As he was only paid with leftover electronic parts, it allowed him to start working on his own ideas for electronic music instruments.

Unique to his approach was an early focus on digital circuitry and computers. As early as 1968, Kurenniemi completed work on his first synthesizer controlled by a digital sequencer, named Integrated Synthesizer. A year later, under a commission from composer Ralph Lundsten, he created The Andromatic—a fully polyphonic digitally-controlled synthesizer. In the following years, Erkki established a company, Digelius Electronics Finland Oy, and developed a few DIMI-series synthesizers (DIgital Music Instrument) exploring the potential of then-new microprocessors and novel control methods. Unfortunately, due to the complex nature of programming the instruments, they didn’t gain much popularity...and by mid-70s Digelius was out of business.

Max Mathews with the GROOVE System Max Mathews with the GROOVE System

GROOVE System

While Kurenniemi was experimenting with emerging digital technology in Finland, Max Mathews was also pioneering computer music research in the United States. GROOVE (Generated Real-time Operations On Voltage-controlled Equipment) was created in 1970 at Bell Labs as a result of a three-year collaboration between Max Mathews and composer Richard Moore. Unlike the entirely digital machines created by Kurenniemi, GROOVE was a hybrid analog-digital system, as Mathews didn’t want to rely on the limited sound-synthesis capabilities of early computers, and instead utilized them strictly for control and memory with a high emphasis on real-time performance.

GROOVE only ran on the Honeywell DDP224 computer and could be programmed using the C language. Since it was equipped with a storage drive, music performances could be recorded, stored, and played back at a later time. The composer could also create a library of code-snippets which could be re-used and re-arranged, effectively simplifying the process of composition. Since all the sound duties were taken care of by an analog modular synthesizer, the computer’s CPU was fast enough to perform all the tasks in real-time. The system was used throughout the '70s but eventually was abandoned because of its high cost and the overall shift towards more compact and affordable personal computing.

EMS Sequencer 256 EMS Sequencer 256

Digital Sequencers

However, more people started seeing the benefits of digital technology, and its breadth of applications started rapidly growing. Digital memory was particularly useful for new electronic instruments, and the first commercial application of such could be witnessed in the stand-alone EMS Synthi Sequencer 256, released in 1971. It allowed composers and musicians to record sequences live, and meticulously edit them later. As the name suggests, the sequencer could store up to 256 “events”, which included note pitch, duration, and velocity. Additionally, Sequencer 256 was equipped with three separate tracks, which could play back their contained sequences simultaneously.

Subsequently, more manufacturers started developing and releasing digital sequencers, including the noteworthy Oberheim DS-2 (1974), and Sequential Circuits Model 800 (1977). Following up, in 1977 an iconic microprocessor-driven sequencer designed by then up-and-coming Japanese manufacturer Roland, the MC-8 Microcomposer, entered the electronic music instrument market, offering a set of never-before-seen features like keypads for note entry, 16KB RAM, allocation of multiple CV sources to a single Gate output for polyphony, and a whopping sequencer length of 5200 steps.

Fairlight CMI II

Around the same time period, the idea of a Music Workstation, an all-in-one device for the production of music, started surfacing—leading up to the development of the pioneering New England Digital Synclavier and Fairlight CMI. While the first versions of the instruments were already groundbreaking in their non-tape-based methods of recording, storing, and playing back sounds, the addition of sequencers in the second series was a significant factor in their popularity. Particularly interesting was the introduction of Page R in the CMI Series II, which eliminated the need for the complicated Music Composition Language (MCL) programming, replacing it with an intuitive graphic interface-based multitrack sequencer. This significantly expanded the potential audience of the Fairlight, and could be considered a prototype of modern Digital Audio Workstations.

Additionally, as the decade drew to a close, Roland started to introduce some of their to-become-iconic drum machines. In 1978 the company introduced integrated circuit technology (IC) with the release of CR-78. This helped to simultaneously reduce the size of the machine and increase the number of features—as CR-78 now featured a dedicated programmer that allowed musicians to create and store up to four rhythmic patterns of their own invention, rather than relying on the preset-based schemes of prior rhythm machines. In 1980 Roland introduced the TR-808 Rhythm Composer and TR-606 Drumatix, which featured much more sophisticated sequencers. Operationally, the sequencers broke down a measure into sixteen steps, with each step corresponding to a 16th note value. The performer would press buttons associated with particular steps to create rhythms. Pattern storage capacity was also expanded, allowing for chaining several patterns together into longer sequences.

The Dawn of MIDI

By the end of the '70s, electronic music instruments were becoming increasingly popular, yet there was no standard for how they were controlled...resulting in incompatibilities between instruments produced by different manufacturers. The initiative to establish a common protocol came from Sequential Circuits president Dave Smith and the head and founder of Roland Corporation, Ikutaro Kakehashi. Thus in 1983, the MIDI protocol (Musical Instrument Digital Interface) was unveiled.

The introduction of MIDI changed everything. Now, composers and musicians could use one piece of equipment to control the other, even between manufacturers. Constructed of a series of commands and messages that control parameters of a synthesizer, MIDI offered a non-destructive method of composition, as the data could be freely edited and manipulated without permanently affecting the sound. Sequencers proved to be an excellent platform for executing MIDI commands, and started being ever-present in synthesizers and rose in popularity along with personal computers and standalone samplers. In 1983, Roland released the first standalone MIDI sequencer, the MSQ 700 Multi-Track Digital Keyboard Recorder. MSQ 700 could host and simultaneously play back up to eight sequenced tracks. It also introduced an original workflow, requiring the user to record musical phrases into the sequencer as they played them on a keyboard, and then play them back with options for overdubbing and alterations to the original tempo. This style of sequencing later was translated into the legendary SH-101 synthesizer, and even survives to this day—it recently has also been adopted in a series of Moog machines, including the Grandmother and Matriarch.

In the same year, Roland also introduced a new drum machine, the TR-909 Rhythm Composer, which was equipped with MIDI, a step sequencer in the style of the previous iterations of the TR-series, and a mixed synthesis/sample-based sound engine. The improved sequencer could store up to 96 patterns and featured updated control over Accents, as well as a new Shuffle and Flam control. The machine had a very short production life span, as the public was drawn to entirely sample-based competitor products like Linn Drum. It was succeeded a year later by the TR-707.

The TR series of instruments gained a cult-like following throughout the '80s and beyond. The machines undoubtedly left a giant impact on several music scenes and the style of sequencing implemented in them even acquired a unique nickname—x0x-style sequencing.

Tracker Sequencers

As you may have noticed, music has always been incredibly receptive to new technology, occasionally acting as a catalyst for its evolution. As soon as computing started to emerge on the horizon, it began to be integrated into the world of sound creation and music. In as early as 1957 Max Mathews (responsible for the GROOVE system mentioned above) working at Bell Labs wrote MUSIC, the first computer program geared towards music creation. So as soon as computers were cheap enough to be personalized in the 1980s, various software applications designed to streamline the process of making music started being developed.

Among such was the Ultimate Soundtracker developed by Karsten Obarski for the Commodore Amiga in 1987. This was the first tracker sequencer program, originally intended as a tool for the creation of game music on Amiga. It featured four tracks for sequencing sampled sounds, allocated respectively to melody, accompaniment, bass, and percussion parts.

The workflow involved entering the data that described basic elements of music composition. Pitches, note durations, and velocity could be entered as numbers or letters on a vertical timeline, and the playback would go from top to bottom. At the moment of release, Ultimate Soundtracker did not resonate with the public, as it was considered too complicated and unintuitive, resulting in it being mostly used for game sound on Amiga. However, the format did appeal to some users and over the years was updated and improved starting with freeware clones like NoiseTracker, and then ProTracker.

To this day, trackers occupy a special niche in the music-making community, and even in recent years, examples of the format being reenvisioned and modernized appeared such as with Renoise and Sunvox software, as well as XOR Electronics NerdSeq and Polyend Tracker hardware units. The use of music trackers is highly associated with the demoscene—a global subculture of computer generated art and music enthusiasts, where artists create small self-sufficient audio-visual demo programs. Additionally, some of the well-established electronic music producers and composers are known to use tracker software, notably Aphex Twin, and Venetian Snares.

The Rise of The Groovebox

Roland Corporation introduced the term “Groovebox” in 1996 as they used it to describe their MC-303, a self-contained instrument capable of both synthesis and sequencing that, above all, facilitated uninterrupted live production of electronic music. Since then the term is generally applied to any portable hardware instrument and in some instances software (i.e. a few Native Instruments Reaktor ensembles) capable of producing full multi-part songs. Even the classic machines from the TR-series are occasionally referred to as such ex post facto.

This format became increasingly popular and was refined with the release of each new machine, resulting in several designs that significantly impacted the way music was being produced, e.g. SP-303, SP-404. x0x-style sequencing was often implemented in grooveboxes, as it offered a fast and intuitive way of creating rhythmic patterns. Yet it was limited, and only a handful of parameters could be automated.

A revolutionary update came from a still-young Swedish manufacturer Elektron with the release of their highly-acclaimed Machinedrum UW in 2005. The sequencer offered a new feature called Parameter Locking, which allowed users to change any parameter of sound per step, essentially a form of live, on-the-fly automation recording. To counterpart the repetitiveness, step probability was added, allowing users to set the likelihood for any particular sequencer stage to trigger. This opened up a lot of potential for creating intricately detailed and evolving parts intuitively without long and tedious programming. Presently, all of the Elektron groove machines from Digitakt to Octatrack are equipped with the same immensely powerful sequencer that can, besides sequencing internal voices, be used to control up to eight different devices in your studio via MIDI.

Generative Sequencers

So far we’ve mostly looked at what could be generally described as programmable step sequencers. This indicates that we are manually programming rhythms and notes into sequencer and it plays it back as specified. We’ve touched upon the notion of probability, but mostly using it as a definable feature applied towards the overall pre-programmed pattern. There are a few designs that take a rather different and less deterministic approach towards melody and rhythm generation.

Euclidean Rhythms

Euclidean sequences are created on the premise of distributing an arbitrary number of beats as evenly as possible across a measure of predefined length. The concept emerged from the research paper “The Euclidean Algorithm Generates Traditional Musical Rhythms,” published in 2005 by Godfried Toussain.

To generate a Euclidean sequence we only need two numbers, where the smaller number represents the number of beats and the larger one determines the total number of steps in the sequence. According to the Euclidean algorithm, the equal distribution of pulses is determined by finding the highest common denominator between the two numbers. This way, if we try to spread two beats across a four-step long measure equally, the pulses will naturally fall on steps one and three because the common denominator between the numbers is two. Using the same logic applied to numbers three and eight, we get a syncopated triplet feel. From the latter example it is also evident that whenever there is a remainder, it also gets placed as evenly as possible. It’s important to note that any two numbers can be used. Stacking multiple euclidean sequences on top of each other, and offsetting some of the sequences can produce complex evolving rhythms very fast.

Interestingly, it was discovered that rhythms in many folk music traditions follow the Euclidean rules. For that reason, Euclidean sequences often feel natural despite their seeming complexity. These days, there are a plethora of software and hardware Euclidean sequencers. In the Eurorack format alone there are Mutable Instruments Yarns, vpme.de Euclidean Circles, Rebel Technology Stoicheia, Hikari Instruments Eucrhythm, and more. You can also play around with euclidean sequences on this website.

Analog and Digital Shift Registers (ASR, DSR)

Shift registers can be very powerful musical sequence generators, and as they are mildly influenced rather than systematically programmed by the user, their results can be full of happy accidents and surprises.

Analog Shift Registers, like the one found in Verbos Electronics' Random Sampling module, are much like an advanced sample-and-hold circuit, and the idea comes from none other than Serge Tcherepnin. ASRs accept a clock and a stream of continuous voltage at the input. The voltage is sampled at the rate of the clock and then passed to a series of cascading outputs one by one. This means that on first clock pulse the sample voltage is sampled, on the next pulse it is passed to the first output and the new value is being sampled at the input, yet on the next clock the voltage from first output shifts to the second output, and the previously sampled value is sent to the first output, while the device is sampling new value, etc.

Digital Shift Registers work on the same principle as an ASR, yet instead of sampling continuous voltage streams they sample digital bits, making them suitable for generation of non-linear, ever-evolving beats and rhythms. An exemplary DSR module in the Eurorack format is Dual Digital Shift Register from omiindustriies.

To Be Continued...

As you may see, music sequencers are a very dense subject...and while we’ve touched upon key historical precedents that influenced the development of this technology, we are still far from being done talking about it. In the last decade alone, there has been a huge amount of new takes on this centuries-old concept, and in the second part of this article we will explore them, as well as discuss some common and not-so-common methods of interacting with sequencers. Stay tuned!