[Chapter 2] FM Tone Generators and the Dawn of Home Music Production

The standard bearer for digital synthesizers, the DX7 (May 1983, shown at front), and its successor, the DX7IIFD (October 1986 second from front). At rear is the DX1 (1983), and immediately in front of it, the DX5 (1985). With an FM tone generator, 32-voice onboard memory (64 voices with an external ROM attached), and a price tag of 248,000 yen, the DX7 and other instruments in the DX series were hit products that sold around 420,000 units worldwide. The body of the DX7 was given a brown color to match the body color of the YIS home computer that Yamaha was selling at the time.

[Chapter 2] FM Tone Generators and the Dawn of Home Music Production

Great advances in semiconductor technologies

photo:Testing FM tone generation at Standard University in the USA

Testing FM tone generation at Standard University in the USA

photo:Dr. Chowning, of Stanford University, using a GS1 voice programmer to create sounds

Dr. Chowning, of Stanford University, using a GS1 voice programmer to create sounds

The start of the 1980s saw an explosion in the popularity of semiconductor-based electronic components, and devices that simply had not been possible with earlier technologies started appearing on the market in rapid succession. Terms such as "integrated circuit" and "large scale integration" began to show up on university entrance exams, and companies started producing electronic games based on this type of circuitry. The breakthroughs made in the field of semiconductors during those years were truly remarkable.

One of the more notable technologies made commercially viable by these rapid advances was the digital frequency-modulation (FM) tone generator. This sound creation method was originally developed at Stanford University in the United States, and Yamaha—the first company to recognize its true potential—signed an exclusive licensing contract with the university in 1973.

Our research team started working with FM tone generators as part of a scheme to switch over the Electone┬« to digital technologies, and by 1974—when the analog SY-1 Yamaha synth was released—we had already successfully completed a prototype instrument with a digital FM tone generator at its core. Unfortunately, it was not yet possible to bring this instrument to market due to the huge number of integrated circuits required by the semiconductor technologies of the time, and also because of the difficulty experienced in balancing size and function satisfactorily. As further advances were made in the field of semiconductors, we finally achieved an instrument with specifications we considered acceptable. And in April 1981—seven years after the start of development—Yamaha released its first FM tone generator product in the F-70, a classic Electone model. This was followed one month later by the GS-1 keyboard, an instrument intended for stage use.

photo:A voice programmer for the GS1

A voice programmer for the GS1

FM synthesis is notable for its ability to recreate with remarkable levels of realism those sounds that are full of variety and rich in harmonics—such as the electric piano, brass instruments, and glockenspiel. Sampling has now become the mainstay of tone generation, and because this technique makes use of actual recordings, we take it for granted that our synths can effortlessly reproduce the sounds of a vast array of different musical instruments. However, the analog synthesizers of the early eighties were simply unable to produce certain kinds of voice—bell-type, metallic sounds being a notable example—and this made the FM sounds of the GS-1 truly sensational.

The GS-1 was not actually marketed as a synthesizer, possibly because sounds could not be edited on the instrument itself. Voice cards could be used to change the bank of 16 voices that the GS-1 was able to produce, but a special, programming device for use by developers (see the photograph below) was needed to create or modify these sounds. Truth be told, this ability to make voices editable proved to be quite a hurdle in realizing a commercially viable synthesizer product.

photo:Released in 1981, the GS1 digital keyboard featured FM tone generation and touch-sensitive control. Notable for its grand piano-like body, the GS1 was an expensive instrument that sold for 2.6 million yen.

Released in 1981, the GS1 digital keyboard featured FM tone generation and touch-sensitive control. Notable for its grand piano-like body, the GS1 was an expensive instrument that sold for 2.6 million yen.

The user interface concept

The sounds produced by an analog synthesizer can be changed by adjusting the values of resistors and other electronic components that make up its tone generator circuit; therefore, knobs and faders containing variable resistors can be added to provide sound editing functionality. The way in which these controllers are arranged depends on the design and size of the synth itself, and instruments such as the CS-80 introduced in Chapter 1 already needed a huge array of knobs. Digital synthesizers have many more sound-related parameters than their analog predecessors, so assigning a physical controller to each and every one would have been totally impractical.

It is also important to remember that digital synthesizers operate based on programs in much the same way as computer software. To produce a new sound, one simply needs to add the required program. However, if the parameters of the sounds themselves are to be editable, then the synthesizer also needs an editing program. Needless to say, the editing program would require its own buttons and knobs for entering parameter values—in modern parlance, its own user interface (UI).

photo:A catalog showcasing the GS1 and GS2 (for Japan). Chick Corea appeared in the opening pages.

A catalog showcasing the GS1 and GS2 (for Japan). Chick Corea appeared in the opening pages.

One of the UIs most familiar to us is the computer's screen, keyboard, and mouse. Neither Windows nor Mac yet existed back in 1980, and entering commands and text using a keyboard was the principal way that users interfaced with their computers. The methods we have grown used to today—for example, working with a graphical interface and mouse or a touch-sensitive screen—were not available back then. In developing the digital synthesizer, creating a clear and simple UI for the musician who wanted to interface with sound in a more intuitive fashion and also for the user with no experience of computer programming was perhaps the greatest challenge that had to be overcome.

One of the UIs most familiar to us is the computer's screen, keyboard, and mouse. Neither Windows nor Mac yet existed back in 1980, and entering commands and text using a keyboard was the principal way that users interfaced with their computers. The methods we have grown used to today—for example, working with a graphical interface and mouse or a touch-sensitive screen—were not available back then. In developing the digital synthesizer, creating a clear and simple UI for the musician who wanted to interface with sound in a more intuitive fashion and also for the user with no experience of computer programming was perhaps the greatest challenge that had to be overcome.

As a solution, our developers devised the new type of programmer shown below. This instead utilized a combination of lamps and buttons that the sound designer could use to confirm the parameters' previous settings when making edits.

The UIs of modern synths provide full and free access to all internal parameters—we do not realize how lucky we are, because this was not always the case. Back when semiconductor and program technologies were developing at breakneck speed, countless rounds of trial and error were needed to perfect a UI conducive to creative sound design. Nevertheless, this was a crucial step in the development of the synthesizers of that era.

photo:The TRX-100 prototype, the basis for the GS1

The TRX-100 prototype, the basis for the GS1

photo:The programmer for the TRX100

The programmer for the TRX100

photo:One of the programmers built as a test, consisting of a combination of lights and buttons.

One of the programmers built as a test, consisting of a combination of lights and buttons.

Arrival of the DX7 to transform the music scene

Two years after prevailing against all odds to develop an FM tone generator, create a UI for programming and editing sounds, and successfully release the GS-1, Yamaha introduced the world to its DX7 frequency modulation synthesizer. At the heart of the FM tone generator lies the operator—a fundamental component used to generate and modify sound. Whereas the GS-1 had four operators, the new DX7 now featured six, allowing it to create much more elaborate sounds. What's more, this revolutionary synth also had built-in functionality for creating and editing sounds, and it allowed these sounds to be stored on cartridge-type memory, all for roughly one tenth of the price of the GS-1. It is hardly surprising then that this new instrument had such a profound effect on the synthesizer world.

photo:The PAMS, the prototype for the DX series

The PAMS, the prototype for the DX series

At that time, a number of Yamaha departments were developing different instruments in parallel, and whereas the GS-1 was preceded by the TRX-100 prototype, the direct forerunner of the DX Series synths was a test model known as the Programmable Algorithm Music Synthesizer (PAMS). In recognition of this fact, the DX7 is identified as a Digital Programmable Algorithm Synthesizer on its top panel.

As its name suggests, the PAMS created sound based on various calculation algorithms—namely, phase modulation, amplitude modulation, additive synthesis, and frequency modulation (FM)—and from the very start, the prototype supported the storing of programs in memory. However, this high level of freedom in sound design came at the price of a huge increase in the number of parameters required, meaning that the PAMS was not yet suitable for commercialization as an instrument that the average user could program.

photo:A prototype DX1

A prototype DX1

In order to resolve this issue, the Yamaha developers decided to simplify the synth's tone generator design by having the modulator* and carrier* envelope generators share common parameters. They also reduced the number of algorithms—or operator combination patterns—to 32. This paved the way for completion of the original DX Series lineup, comprising the DX1, DX-5, DX7, and DX9. Although four models were released at this time, five model codes—DX1, DX2, DX3, DX4, and DX-5—were actually employed during development. The DX1 kept its code upon release, which is quite rare for Yamaha products, while the DX2 and the DX3 together became the DX-5. The DX4 and the DX-5 development models came to market as the DX7 and the DX9, respectively.

photo:From the TX7 catalog (English). Here the TX7 is shown attached to a DX7.

From the TX7 catalog (English). Here the TX7 is shown attached to a DX7.

The DX7 was an instant hit all over the world, and both the instrument and its sound soon became driving forces of the pop music of the eighties. We should note, however, that many of its technologies and features also greatly influenced how synths would be developed thereafter.

The first of these was an LCD screen comprising two lines of 16 characters each. Prior to the DX7, synthesizer parameter values were typically confirmed from the positions of knobs and sliders, meaning that there was no way to accurately check parameter settings or display voice names. With the arrival of this type of UI element, however, it became possible to display all types of information, and the tradition of naming original voices was born. Meanwhile, the fact that individual parameters could be called up and edited one at a time on the LCD screen eliminated the need for a vast array of controllers on the top of the instrument. The neat and tidy control panel of the DX7 would not have been possible without this screen, and this clear distinction from the synthesizers of the past was yet another factor behind its overwhelming popularity.

photo:A connection diagram from the TX7 catalog. The catalog feature explanations with the TX7 and QX7 connected to a DX7.

A connection diagram from the TX7 catalog. The catalog feature explanations with the TX7 and QX7 connected to a DX7.

The next groundbreaking feature of the DX7 was the use of memory cartridges to store and recall voices—a feature that was only possible thanks to the synth's digital design. Whereas the GS-1 had used magnetic-type voice cards, Yamaha decided that cartridges containing digital memory would be better for the DX Series as they are unaffected by the powerful magnetic fields produced by speakers and other similar equipment. The DX7 can store 32 voices internally, but with a ROM cartridge plugged into its cartridge slot, an extra 64 voices become available. RAM cartridges, meanwhile, can be used to write and recall up to 32 original voices. This ability to boost the number of voices is unique to the digital synthesizer, and our highly convenient cartridge-based approach also made the sounds of professional musicians available to all. Back in the era of the analog synth, the only way to reproduce the sounds used by pros was to copy the positions of each and every knob, and even then, it was almost impossible to get the exact same settings. DX7 owners, however, could easily purchase cartridges containing the actual sounds of famous synth players. This novel approach, which made it possible not only to own the same instrument as one's synth hero, but to also play the very same sounds, was extremely attractive to amateur musicians.

Also worthy of special mention are the remarkable improvements that Yamaha developers achieved in keyboard performance—now all the more important in controlling the highly complex sounds that FM made possible. Working in combination with a touch-sensitive keyboard, the FM tone generator can modulate sounds in a myriad of different ways, and in order to get the most out of this technology, we decided to equip the DX7 with our FS Keybed. Although originally developed for the Electone, this keyboard went on to be a standard component in flagship Yamaha synths for over two decades, becoming much loved by a great many musicians.

Last but far from least is the DX7's support for MIDI—a technical standard introduced in 1982 to enable musical instruments to exchange information digitally with one another. In addition to information produced by playing the keyboard, this also includes data generated by operating the sustain pedal, the volume pedal, and many other performance-related controllers. The simple fact that Yamaha adopted this standard so soon after its release was another reason why the DX7 attracted so much attention at the time, yet the functionality it provided was just as inspired. For example, when controlling the DX7 using a MIDI sequencer—a device that can play synths automatically by transmitting MIDI data—one could recreate the performance of another musician note for note, and effortlessly create robotic sounding parts or high-speed phrases that humans would find very difficult to play continuously. Another feature that got the DX7 noticed was its ability to produce innovative, cutting-edge music such as the dance and techno sounds born in the eighties—music that was achieved by combining robotic MIDI performances with hard synth bass that possessed a distinctive FM sound.

With these and other trailblazing features, the DX7 digital synthesizer shook up both the performance and business sides of the music industry and greatly influenced both the pop music of the day and the shape of synths to come.

The changing world of the synthesizer

Following the introduction of the DX7, the world of the synth underwent major change. The addition of MIDI support not only made it possible for musical parts to be played automatically, but it also gave rise to the concept of tone-generator expansion for real-time performers. For example, two DX7s could be made to play exactly the same electric piano part, and if the pitch of one of these instruments was raised ever so slightly, the result would be a chorus-type effect, making the overall sound much richer. This approach was just as applicable to greater numbers of synths, but because nobody could actually play three or four synths at the same time, we realized that DX7s being used exclusively in an expansion capacity had no need for a keyboard. Our response was the TX Series of keyboard-less tone generator modules.

The DX7 was followed by many more of these products, such as the rack-mounted TX-816 tone generator capable of producing extremely rich sounds, and the TX-7, which packaged a DX7 tone generator in a highly distinctive case. The luxurious FM sounds they delivered through tone-generator expansion also became an indispensable part of the music scene of the day, gaining these products an excellent reputation in the process.

The synths of the Yamaha DX Series continued to evolve together with advances in technology. The DX7 II had an aluminum body for reduced weight and greatly improved portability, and a disk drive was later added to cater for the 3.5-inch floppy disks that were in widespread use at the time. Further development of the series saw the introduction of even more creative features, such as dual output channels with support for stereo panning, and micro-tuning functionality that allowed musicians to use tuning systems other than equal temperament, such as Arabic musical scales. Meanwhile, the DX100 mini keyboard model (which was released slightly earlier) featured a number of innovations specifically for the player: for example, the pitch bend wheel was moved to the top-left corner, and when played standing up using a strap, the instrument's pitch-bend direction could be reversed so that notes could be bent in the same way as with a guitar.

Not only did it take the music world of the eighties by storm, the DX Series was also an engine of development of the modern digital synth's UI and primary functionality as an instrument.

Onward to home music production

Up until the eighties, amateur musicians certainly performed their music live, but recordings were exclusively made in studios by professionals. During this decade, however, the multitrack recorder (MTR)—a device capable of recording four individual tracks on a standard music cassette—became extremely popular, and this made it possible for anyone, regardless of ability, to produce multitrack recordings from the comfort of home. In the beginning, the standard MTR process was to record the rhythm first using a drum machine, and to then layer bass, guitar, and keyboard tracks on top to complete the song. With the growing popularity of MIDI-compatible instruments, however, musicians were able to synchronize their sequencers and drum machines, and MIDI synths such as those of the DX Series were regularly used for both bass and chord-type tracks. Yet the DX7 could only produce one voice at a time, meaning that two of these synths would be required if, for example, bass and electric piano had to be played at the same time.

Yamaha's solution was the multi-part tone generator. MIDI data can be assigned to specific channels, and if a MIDI sequencer such as a QX Series product is used to transmit performance data organized into different channels, then bass, piano, and marimba voices can, for example, be played by the data on channels 1, 2, and 3, respectively. A multi-channel tone generator receiving this data would assign a different voice to each channel, and in our example, the equivalent of three individual synthesizers would need to be integrated into a single tone generator. The product that Yamaha developed in line with this approach was the TX81Z tone generator module—a revolutionary device containing the equivalent of eight FM synthesizers, each with four dedicated operators. All eight of the FM tone generators could also be set to the same channel to produce richer, denser sounds. Furthermore, the operators were equipped with waveforms other than the sine wave for the first time so that a more diverse range of sounds could be generated, and for this reason, the TX81Z it is often considered a hidden gem among synth modules.

Around this time, music producers started playing all parts—be they rhythm, bass, or chord-type instruments—simultaneously using one synthesizer, and synths with built-in MIDI sequencers began to appear on the market. Developed to address this need, the Yamaha V-50 was the ultimate FM synthesizer, combining the TX81Z with a keyboard, a MIDI sequencer, a rhythm machine based on PCM tone generation, and digital effects processors. This instrument transcended the boundaries of the digital synth, propelling it into the workstation era.

In less than a decade, between the GS-1 of 1981 and the V-50 of 1989, our digital synthesizers evolved from basic performance instrument to full-fledged music workstation. Without doubt, the eighties was one of the most exciting and dynamic periods in the history of the Yamaha synthesizer.

photo:V50

V50

photo:DX9

DX9

photo:DX7II D

DX7II D

photo:DX100

DX100

photo:TX816

TX816

photo:TX81Z

TX81Z