Any phone of a language can be synthesized with 4-5 parallel formant filters

Evidence

The basic configuration of formant speech synthesis is the parallel filter structure shown in Figure 5.14. A finite number of second-order low-pass filters (typically 4–5) are in parallel in the structure, and it can be used to synthesize any phones of a language. A rich set of control parameters is necessary. Each formant has to be controlled by the frequency, the amplitude level, and the Q-value (sharpness of the resonance). Voiced and unvoiced excitation sources are needed, as in Figure 5.11. — Ville Pulkki, Matti Karjalainen, Communication Acoustics- An…, 625

Categorization

• corroborated: 2 independent passages support this claim — Ville Pulkki, Matti Karjalainen, Communication Acoustics- An…, 625