X 66

The Formant System

In theory at least, as well as in practice for some tonal effects, the formant system is the better of the two. Furthermore the formant system exactly, or at least very closely parallels what happens in many musical instruments including people's voices and a host of other basically musical sounds.

We'll look first at a saxophone, as that is an instrument that just about everybody knows and likes. When the musician blows into the mouthpiece of the saxophone, the blast of air sets a reed into vibration. Unlike the free reeds of the Wurlitzer electrostatic instruments, the saxophone reed is called a striking reed because the reed is bigger than the opening in the shallot over which it vibrates and during part of each cycle of vibration, it hits the shallot. Shallot in this context is actually a pipe organ term. In the saxophone I believe the equivalent of the shallot is called the mouthpiece of the saxophone. When a reed is larger than the opening over which it vibrates, it produces an entirely different sound from that of the free reed, which is very slightly smaller than the opening over which it vibrates.

Regardless of what term we use, the pitch at which the reed vibrates is determined mainly by the fundamental resonant frequency of the air column enclosed by the body of the saxophone. The "keys" or valves along the sides of the saxophone open or close ports which tune the resonant air column to different frequencies. Because of the strong standing wave that gets set up in the saxophone by the reed, it easily synchronizes its vibration rate to match the fundamental frequency existing in the saxophone for whatever valve combination the musician uses at any instant.

The sound wave which this reed generates in the air is an approximate sawtooth wave. It is somewhat like the sawtooth wave illustration below. It also sounds completely different from the soundwave of a free reed.

In the saxophone, the approximate sawtooth wave from the reed enters the body of the instrument. The air column within the saxophone body has certain frequencies at which it is naturally resonant; these change according to what valves are open at any moment, but for each note, there will be certain bands of frequencies which the saxophone body will emphasize. Likewise, there will be certain frequencies which it will suppress somewhat. Therefore, the shape of the soundwave from the reed changes inside the saxophone. Some harmonics are emphasized, and others are suppressed. The tone which we finally hear when we hear a note played on a saxophone is no longer a sawtooth wave and it is no longer a nondescript buzz (which is the best way I can describe the sound of a sawtooth wave). It emerges from the bell of the saxophone as a recognizable sound, that of a saxophone.

In an electronic instrument which uses the formant system, the tone generator should output a fairly decent sawtooth wave for each available pitch on the instrument. The sawtooth waves then pass through sets of electrical filters which are designed to approximate the action of the various musical instruments to be imitated. That is, the electronic instrument maker designs the formant filters to have the same bands of resonance and suppression as the body of the real musical instrument which he wishes to imitate. When the sawtooth waves from the tone generator pass through suitable electrical formant filters, they emerge as recognizable instrument tones. In this way, the electrical formant system exactly parallels the acoustical formant system of a real musical instrument.

When duplicating those instruments which have significant upper harmonics and strong bands of resonance such as the trumpet, trombone, English Horn, oboe, bassoon, and bowed strings such as violins and cellos, the electrical formant system can give some fairly realistic results. This is not surprising, because both the acoustic and the electrical systems are doing essentially the same thing; that is, receiving a sawtooth waveform at their inputs and then modifying it into a recognizable musical instrument sound.

The formant system does not do as well, however, at imitating sounds which have very little harmonic development such as certain flute-type tones and it is also hard to create the interesting novelty effects and pseudo-percussion effects that you can get from harmonic synthesis. Many bell tones for example have not only harmonics but also partials. A partial is like a harmonic but the frequency of a partial is not an exact integral multiple of the fundamental. Likewise in many bells, there may be a fundamental, and then no other harmonics at all, only a few upper partials. Therefore, the convincing duplication of instruments such as the glockenspiel, celesta, chimes or bells is not possible with the formant system. Additive harmonic synthesis is the only way that these types of instrument tones could be simulated by analog electronic organs. As a side note, the glockenspiel imitation in the X66 percussion sections is the most realistic synthesized glockenspiel that I have ever heard, and it is synthesized from several individual sine wave frequencies.

Another potential drawback of the formant system (as it exists in low-end electronic organs) is that most of the instruments that these electronic organs imitate by means of the formant system are monophonic instruments. That is, they are capable of sounding only one note at a time. You can't play chords on an individual saxophone, for example. If you want chords with saxophones, you must have as many individual musicians with saxophones as you have notes in a chord.

When you play a chord on an electronic instrument using a simple formant system, the formant filter tends in a sense to "unify" the chord, making the individual tones slightly lose their identity and making the chord itself sound almost like a single sound. A second even more serious problem is that when the fundamental pitch of a note approaches a frequency that the formant filter likes, it becomes too loud. Then as you play along from one key to the next, certain groups of notes will be disproportionately loud. This fault is glaringly evident on some of the early (ca. 1950s) Baldwin model 5 organs where, if you play the oboe stop on the upper manual, you will find that when you are about an octave and a half over Middle C, it gets way too loud in comparison to the rest of the instrument.

As we will see, in the X66, the Hammond engineers divided the formant system into a number of different note groups, and scaled both the volume levels and the formant filter characteristics accordingly, thus significantly eliminating what would otherwise be a major defect.

The Hammond people also knew that if you wanted to have a really versatile instrument that could do everything, then you needed to develop an electronic organ that used both systems, that is, deriving some tonal effects by additive harmonic synthesis and others by the formant system. Do this successfully, and you can essentially eliminate most of the limitations that are inherent in either system.

From this above description, you will infer [correctly] that the Hammond X66 instrument does indeed use both systems It has sets of harmonic drawbars controlling sine waves to develop tonal effects by additive harmonic synthesis as does a traditional Hammond, and it also generates sawtooth waves which then pass through formant filters to create tonal effects which require formanting for best results.