The Yamaha DX7 is an amazing synthesizer with a dirty secret. It’s no fun to program.
By way of its fabulous presets, the DX7 made Frequency Modulation synthesis one of the de facto sounds of ‘80s pop. Its bass thumped all over hits by Michael Jackson, Wham!, and Tears for Fears. Its simple electric piano sounds blended subtly into the background on tracks like Tina Turner’s “What’s Love Got to Do With It?” Its leads and pads define the aesthetic of the Giorgio Moroder-produced Top Gun soundtrack along with its numerous hits.
But practically no high profile producers or musicians explored the DX7’s sound design possibilities. Many synth heads make the very good point that FM is nowhere near as intuitive a program as traditional subtractive synthesis. Even the luminary synth engineer Dave Smith likes to go on record half–joking that he doesn’t understand how to program with FM.
Analog subtractive synthesis operates on similar principles as an EQ or guitar effects, with twists of knobs shaping the produced sound in a straightforward way. The early and enduring appeal of those huge Moog and Buchla modular systems is that players could conjure any imaginable sound by learning how to tweak a big system with immediate visual and sonic feedback.
But what’s sauce for the subtractive goose is not sauce for the FM gander. Yamaha likely knew it couldn’t make an interface as intuitive as the Minimoog’s or the Juno-60’s for its FM synths, so it gave the DX7 some buttons that programming diehards could use to navigate menus and called it a day.
Couldn’t some other company have come along and devised that perfect interface for the FM synth that made programming as accessible as an analog synth filthy with knobs? The short answer is no.
The long answer starts with Stanford University.
Stanford’s Cash Cow
In 1964, computer music pioneer Max Matthews from the Bell Laboratory helped the burgeoning composer John Chowning set up a computer music program using the computer in the Stanford Artificial Intelligence Laboratory. Three years later, Chowning would develop the Frequency Modulation protocol for digital synthesis and Stanford would quickly patent it.
Frequency Modulation’s is based on a principle that’s elegantly simple in theory, and mind numbing to the uninitiated in practice. In its simplest form, FM uses two sine waves. One is the carrier wave, which is the one you hear, and the other is a modulator wave, which shapes the carrier to create a complex wave shape and, in turn, a variety of tones.
The combination of a carrier and a modulator is referred to as an operator. The operator takes the place of the oscillator on FM synth. In all modern FM synths, multiple operators (sometimes identical, sometimes programmed very differently) are chained together to create FM’s complex tones.
That last part is what makes programming FM synthesis seem like alchemy to many. If using one wave to change the sound of another is a confusing process, then using a modulated wave to modulate a second wave, and those two combined to modulate a third, is mind boggling.
There’s something quantum and, in turn, brain-bending about the rapid mutations that sound undergoes when you start playing around with the interaction between those oscillators.
In 1973 Yamaha became the sole licensee of Stanford’s FM patent. It would reportedly go on to be Stanford’s second most lucrative intellectual property, bringing the university over $25 million in total. Chowning, on the other hand, would get passed over come time for tenure.
Yamaha’s licensing primed the company to be at the forefront of the digital synthesis revolution. And lucky for Yamaha that revolution was already in full swing when it finally came time to release a confounding sound-sculpting technology to the public.
The earliest innovation of the digital revolution in music — and one of the most lasting — was the simple synth preset. In the early ‘80s, analog synths like the Sequential Circuits Prophet line and Roland’s Juno line used presets to give users the easy ability switch between a wide variety of timbres and textures with the quick click of a button.
This gave studio musicians the ability to change sounds quickly in the middle of a recording session, and allowed the slightly adventurous to tweak, say, a brass preset ever so slightly to help it cut through a mix.
This preset was also a godsend for Yamaha, which needed an easy way to market the revolutionary new sounds that FM synthesis offered.
Yamaha’s most successful FM synth, the DX7, featured 42 buttons for navigating the digital architecture and provided a diagram of all of the different routing options for the FM operators. But as the DX line (and rack–mountable TX line) would get developed, Yamaha would pare these programming options down further and further to just about non–existent.
A similar trend would take hold across all digital synthesizers throughout the ‘90s and 2000s. Especially on the lower end of the market, synths like the Alesis Micron and microKorg had minimal interfaces and menu–heavy programming structure. This type of interface makes all synths hard to program, not just FM.
Which takes us to a small piece of forgotten FM history in the early years of home computers.
Enter the Matrix
So much of the analog synth revival of today is centered around knobs. When Moog introduced its Little Phatty monosynth 10 years ago, it required a decent amount of menu diving. But it was eventually succeeded by the Sub Phatty and Sub 37, which almost exclusively rely on knobs for programming.
Likewise, Korg’s newest analog synths — the Volcas, the Minilogue, the Monologue — are heavy on the hands–on tweakability. Then, of course, the Eurorack movement sees musicians using knobs and sliders as the main forms of expression on their instruments.
Programming a synthesizer demands something to look at and play around with. But not all forms of synthesis are created equal and neither are their interactive needs.
The fact of the matter is that while the hardware on the DX and TX series was incredibly poor for programming purposes, there was a simultaneous wave of digital programming cresting for Yamaha’s synths.
With primitive personal computers like the Atari videogame system and early Macs, sound designers were using YSEditor and other programs to develop sounds for the Yamaha DX and TX synths using a big graphical interface. In retrospect, YSEditor’s GUI looks more like a meteorological analysis system than a music interface, but it did make programming Yamaha’s synths, well, possible.
YSEditor laid it all out right there for users to see. There were clearly labeled sections for the oscillator (or operator section), envelope generator, LFO, and more. Sliders gave some visual sense to the strange ways that the sound was being mutated when, say, the carrier signal on the third operator was being tweaked.
Sure, nothing is as easy is turning down a lowpass filter on an analog synth to make a bass sound. But at least YSEditor gave FM synthesis some visual logic for people outside of academic computer music labs.
This graphical computer interface would go the way of the Dodo for a couple of decades, through a combination of people leaving their Ataris behind, Yamaha’s FM synths falling out of favor, a newer digital synths developing using the so–called virtual analog protocol, which really meant approximating subtractive synthesis using digital architecture.
FM in the DAW Age
The truth is that today is the best time for the common synth lover in the history of FM synthesis. For people who simply want a newer version of the DX7, Korg has released its ultra–portable Volca FM, which effectively runs on the same design and can import all of a DX7s sound. It even has a couple of sliders that work somewhat like filters and envelopes on an analog synth.
But for the people who want to get down and dirty with FM, today’s diverse menu of VSTs provides an absolute wonderland for tweaking.
Ableton Live Suite comes with a great FM synth called Operator which makes the interaction between those titular FM operators really logical. It basically puts FM operators into a subtractive synth, meaning that you can just tweak presets’ filters, LFOs, envelopes, and so on.
But it also makes the operators straightforward. You can drag knobs to adjust the carriers and modulators in a way that helps you understand how these elements are interacting. In fact, seeing FM operators alongside subtractive shaping techniques will provide a familiar space for many people used to more classic synthesis to try out something new with FM.
There are great options for players who want softsynths that faithfully recreate the architecture of the DX7 yet allow users to get under the hood easier.
Then there’s the Dexed DX, a full–on DX7 emulator that goes big where the DX7 went home. It provides a screenful of dials and envelope diagrams that finally give you a visual for everything going on inside of the DX7 every time you pull up or play around with a preset.
The SONiVOX FM Piano stands as a faithful recreation of the DX7’s beloved electric piano sounds while also offering the ability to easily tweak the sound to fit whatever context or mix you need it to fit.
One of the most acclaimed FM VSTs on the market today is the Native Instruments FM8. Like so many of the company’s soft synths, this one has incredibly high sound quality. But more importantly, this is a synth that final offers an accessible interface which mitigates the confusing nature of FM, helping the more run of the mill tweaker get down and dirty.
FM8 lays out out the operator structure of its FM mechanism in an intuitive matrix, giving you a direct visual representation of how you’re sculpting your sound. It’s like YSEditor with a significantly better GUI.
There’s still an alchemical aspect to how those waves transform as you adjust sliders in the synth, but the FM8 strikes that perfect balancing of hiding nothing and keeping sound design as simple as possible. FM8, of course, comes with presets, one-upping Yamaha’s classic DX and TX synths by making it relatively for you to put your signature on stellar pre–programmed sounds.