A legend reborn
FM synthesis – especially as implemented by Yamaha via its legendary DX7 – has a rep for being difficult to program. However, FM’s legendary impenetrability is due primarily to two factors. First, the DX7 was released in 1983, when cheap(ish) analogue polysynths reigned. Preceding digital synths were very expensive, meaning the DX7 wound up in the hands of users exposed only to analogue synthesis – which had remained all but unchanged for nearly two decades. With its newfangled tech, the DX7 might as well have been dropped from a passing UFO.
The other big obstacle to the would-be FM programmer was the DX7’s interface. Yamaha swept the DX7’s front panel clean, leaving only a tiny LCD and a single data slider. Most users simply called up the presets and got on with it. Quite unsurprisingly, the DX7 is often credited with having kick-started the entire third-party patch industry.
Yet if the user is allowed a clear view of things, FM synthesis isn’t really that difficult to get (yes, we know the DX7 was, in reality, based on phase modulation rather than frequency modulation, but this is the terminology Yamaha chose, so we’re sticking with it). In the DX7, Yamaha presented a synthesis engine with six sine wave ‘operators’ (erm, oscillators) per voice. These operators could modulate one another to create more complex tones. The modulators themselves could be modulated by envelopes, LFOs, and various performance controls, producing ever more complex timbres in the process.
The DX7’s envelopes often leave synthesists scratching their heads. Rather than the familiar ADSR envelopes of the typical analogue synth, those on the DX7 were complex time/rate jobs. Yamaha graciously printed graphics across the front to help explain them, but to little effect. Fortunately, modern software GUIs make them a lot easier to understand.
The subject of FM synthesis is complex, and would take up an entire magazine to explain in depth. Luckily, it only takes a few short steps to get a feel for the potential of FM – let’s dive in.