Music synthesizers

Moore, 1985] Moore, F. R. (1985). The FRMBox - A Modular Digital Music Synthesizer. In Strawn, J., editor, Digital Audio Enginnering An Anthology, pages 95-108. William Kaufmann. 

Recognizing the limitations of fixed PCM samples in music synthesizers, the Downloadable Sounds (DLS) specification was added to MIDI. This provides a means to load PCM into a synthesizer and use MIDI commands to play it back. Essentially any instrument/percussion sound can be replaced with arbitrary PCM, making it possible to control large amounts of recorded sounds. The emergence of software synthesizers on PC platforms has made the use of DLS more feasible than in earlier days when the samples were contained in a static ROM within a hardware synthesizer. 

There are an infinitely large variety of acoustic frequency sequences or melodies which satisfy the definition of musical sound. The all possible melodies make up a statistical set or ensemble. Existing pieces of music are members of this statistical ensemble, but still there are many, many other possible acoustic waveforms (or melodies) which can be musical sounds. Some of them are waiting for human composers and others will never been composed by human composers because they cannot be played by musical instruments or they are not so attractive. We tried to pick up these members mathematically and played them with musical synthesizers after proper arrangements made by a musician. Some of them became very nice background music. When this music was played with a synthesizer in such a way that the tempo was slightly fluctuated as 1//, it sounded milder than when it was played in a constant tempo. The 1//-fluctuated tempo simulates the music played by human players as will be described in 2. 

Electronic Sackbut (Hugh Le Caine) Le Caine builds the first music synthesizer, joined by the Special Purpose Tape Recorder in 1954, which could simultaneously change the playback speed of several recording tracks. 

The basic operation of a sampling synthesizer is to playback digitized recordings of entire musical instrument notes under musical control. Playback of a note can be triggered by depressing a key on a musical keyboard, or from some other controller, or from a computer. The simplest samplers are only capable of reproducing one note at a time, while more sophisticated samplers can produce polyphonic (multi-note), multi-timbral (multi-instrument) performances. 

Moorer, 1981] Moorer, J. A. (1981). Synthesizers I Have Known and Loved. Computer Music Journal, 5(1) 4—12. 

Synthesis can have arbitrarily high expressivity, since the sound is generated from fully parametric descriptions of a synthesis process. However, synthesis typically has very poor accuracy. FM, a powerful and popular synthesis method, is notoriously unsuccessful in synthesizing the sound of an acoustic piano. But FM has very simple methods for varying the brightness, harmonicity, and other parameters of a signal under musical control. 

Effect of storing attacks. Storing the attack of a musical instrument is needed because musical instruments are identified to a large extent by their onset characteristics [Winckel, 1967], To illustrate, when the onset of one musical instrument is grafted onto the sustain or steady-state portion of a second instrument, listeners usually identify the instrument based on the attack, not the sustain. Synthesizing the attack of an instrument accurately with other music synthesis methods is very difficult. 

Alternatively, a more realistic music attack or speech plosive may be obtained by splicing into die synthesized deterministic component using the method described in section 3.6. Splicing of the actual attack has been shown to significantly improve the sound quality for a number of musical sounds including the piano and marimba [Serra, 1989]. Hybrid sounds can also be created by matching the sine-wave phases of the attack of one sound with the phases of the deterministic component of a second sound at a splice point. 

Rave A type of loud, synthesized dance music, often played in nightclubs where patrons take ecstasy. 

Nintendo releases the hand-held gaming system, Nintendo DS Carbon nanotubes are synthesized in bulk, and spun into a yarn iPod Nano and a video-capable iPod are introduced by Apple Motorola releases the ROKR El phone, capable of music downloading High-definition DVD players become commercially available Apple computer introduces MacBook Pro, MacBook, and iMac product lines that contain Intel dual-core chips - the first to contain over one biUion transistors... 

In order to synthesize music, speech, or sounds in general, accurate and flexible control of pitch is necessary. The most accurate pitch control is necessary for music synthesis. In sampling synthesis, this is accomplished by dynamic sample rate conversion (interpolation). In order to convert a sampled data signal from one sampling rate to another, three steps are required bandlimiting, interpolation, and resampling. 

---