Questions: Frequency Modulation Synthesis Theory in Composition

Question 1

A composer wants to synthesize a bell-like, gong-like timbre with inharmonic partials. Which FM parameter setting best achieves this?

• A. A carrier/modulator frequency ratio of 2:1 with a low modulation index
• B. An irrational carrier/modulator frequency ratio (e.g., f_c = 200 Hz, f_m = 141 Hz) with a high modulation index
• C. A carrier frequency equal to the modulator frequency with a modulation index of zero
• D. Any integer carrier/modulator ratio, since integers always produce inharmonic spectra

Answer: (B)

Inharmonic spectra arise when sidebands (f_c ± n·f_m) fall at non-integer-multiple frequencies, which happens when the ratio f_c/f_m is irrational. Integer or simple rational ratios produce sidebands at harmonic-series positions, giving a pitched, musical tone. A high modulation index spreads energy across more sidebands, adding density and complexity. A modulation index of zero gives a pure sine wave at the carrier frequency alone.

Question 2

As the modulation index d increases in FM synthesis, what happens to the output spectrum?

• A. The output stays the same — the modulation index only affects pitch, not timbre
• B. The spectrum collapses to a single frequency, since higher modulation cancels sidebands through phase interference
• C. More sidebands acquire significant energy and the spectrum grows more complex, transitioning from a nearly pure tone to a dense, wide-band sound
• D. The carrier frequency disappears and only the modulator frequency remains audible

Answer: (C)

Sideband amplitudes are given by Bessel function values J_n(d). At low d, only J_0 and J_1 have significant values — nearly all energy stays at the carrier with faint first-order sidebands. As d grows, higher-order Bessel functions become significant, spreading energy into many sidebands. This is why FM can model everything from a flute (low d, nearly pure) to a metallic bell (high d, dense spectrum) — modulation index is the primary timbre-complexity control.

Question 3

In FM synthesis, the carrier/modulator frequency ratio determines whether the resulting spectrum is harmonic (pitched) or inharmonic (bell-like, metallic).

• T. True
• F. False

Answer: True

When the ratio f_c/f_m is a simple integer or rational number, the sidebands land on integer multiples of a fundamental frequency, producing a harmonic series. When the ratio is irrational, the sidebands fall at inharmonic frequencies, producing the bell-like or gong-like timbres that made FM synthesis famous. John Chowning identified ratio classes as defining timbral families precisely because of this relationship.

Question 4

Once a modulation index is set for an FM instrument, the timbre it produces is fixed for the duration of the note.

• T. True
• F. False

Answer: False

FM parameters, including the modulation index, can be made time-varying using envelopes or other control signals. Time-varying the modulation index is one of FM's most musically powerful features: it simulates how real instrument timbres evolve over time (e.g., the bright, spectrally dense attack of a brass instrument followed by a more muted sustain). This is central to how FM synthesis was used to model realistic instrument sounds on digital synthesizers, and it is what allows composers to create timbral morphs and dynamic textures.

Question 5

Explain in your own words why a small modulation index produces a nearly pure tone while a large modulation index produces a complex, dense spectrum.

Model answer: The output spectrum of an FM signal contains the carrier frequency plus sidebands at f_c ± n·f_m for n = 1, 2, 3, … The amplitudes of these sidebands are given by Bessel functions J_n(d), where d is the modulation index. At small d, only J_0(d) ≈ 1 is significant — nearly all energy stays at the carrier, producing a nearly pure sine tone. As d grows, higher-order Bessel values J_1, J_2, J_3, … become significant, spreading energy into more and more sidebands. At large d, energy is broadly distributed across many frequencies, producing the dense, buzzing spectrum of metallic or noise-like timbres.

The modulation index d controls the depth of the frequency wobble — how far the carrier swings. A small wobble barely departs from the carrier and generates few new frequencies. A large wobble creates rapid, wide frequency excursions that decompose into many discrete spectral components via Bessel function mathematics. Timbre complexity is therefore continuously controllable through a single parameter.