In a method of synthesizing voiced speech from pitch prototype waveforms by time-synchronous waveform interpolation, one or more pitch prototypes is extracted from a speech signal or a residue signal. The assembly includes a first fixation element, a second fixation element, and an adjustable flexible member construct. The device can have the reflective material layer form each reflector and can use a wavelength converting layer having different densities, and therefore can emit a wavelength-converted light having a high light-emitting efficiency and a uniform color tone from various small light-emitting surfaces. 5 cm in diameter when fully open.