Synthesizer Music: Types of Synthesis Used in Modern and Electronic Music
There are many types of synthesis used by electronic musicians to create new and unique sounds and unnaturally complex timbres. Additive, Subtractive, Frequency Modulation and Wavetable Synthesis are the most frequently used but other types include Phase Distortion, Scanned, Bezier and most controversially Circuit Bent. This paper will discuss the different types of commonly used synthesis as a basis for new and current types of synthesis and whether they can be considered synthesis at all.
Additive Synthesis is by far the most common form of synthesis. In 1906 Thaddeus Cahill invented the first instrument to include additive synthesis. His instrument was knownas both the Dynamophone and the Telharmonium and produced music using alternating current running dynamos and weighed over 200 tons. The theory of additive synthesis comes from Joseph Fourier. He formulated that sound was the summation of sine waves and that by adding waves together new sounds could be created. Current instruments that use additive synthesis are the Kawai K5000 and software versions like Camel Audio.
Subtractive Synthesis is the process of taking a complex waveform and subtracting simpler waveforms from it. It is most commonly done by filtering frequencies out of the complex wave. It is still based on Fourier’s equation but takes the opposite approach. A common instrument that used subtractive synthesis was the analog Moog synthesizers.
Frequency Modulation was first discovered by John Chowning in 1967-1968. Frequency Modulation or FM uses two waveforms. One is the carrier waveform and one is the modulating waveform. The modulation waveform must be in the audio range of the carrier waveform. The modulating waveform causes the carrier waveform to go above and below its frequency. This is called the deviation. The more amplitude the carrier waveform has the more deviation it creates. The deviation creates sideband frequencies or partial frequencies that may or may not follow the overtone series. These sideband frequencies are heard along with the carrier frequency to create a new timbre. The more deviation the more emphasis is given to the sideband frequencies over the carrier frequency. FM is very good at creating bell-like and metallic type sounds. By far, the most popular FM synthesizer is the Yamaha DX7. Yamaha bought the technology from Mr. Chowning and held its patent until it expired in 1995. This made it basically the only company that could make an FM synthesizer.
Phase Distortion Synthesis was created by Casio in 1984. The technique uses a sine wave and the phase angle is modified to create a new waveform. In this sense a sine wave can be turned into a saw tooth or square wave just by adjusting the phase angle. Casio Used this type of synthesis in the CZ and VZ synthesizer models starting in 1984.
Wavetable Synthesis simply uses a bank of waves called a wavetable to create the timbre of a sound. This wave is then sped up or slowed down the create individual pitches. A single instrument can use as few as one wave over the entire range or many on just a single key, determined by velocity. Wavetable Synthesis is the most common form of synthesis in modern synthesizers and computer software synthesizers.
Scanned Synthesis was developed by Bill Verplank, Rob Shaw and Max Mathews from 1998-1999. It uses a sine wave below 15 Hz to generate the timbre and the pitch. The timbre is a scan of the wave from a wavetable by the sine wave and the pitch is the speed of the scan. This method allows control over wavetable pitch without changing timbre.
Bezier Synthesis uses sounds that are created using the Bezier curve. It uses four control points to determine the shape of the curve. These four points can be moved together or independently. These is like having four points that can be modulated unlike only one in Frequency Modulation. The Bezier curve is from the 1960 and was developed for Renault car body design, but Bezier Synthesis was first presented in May 2004 and so far there are only software implementations of it’s capabilities.
Circuit Bending is not the newest but maybe the most controversial of all synthesis. Originally, when the first creators of synthesizers came up with a concept it was based on the predictability of the circuit and therefore could create the same sound every time. The original inventors relied on this to create their circuits and sell their designs. Circuit Bending can often have the opposite approach to synthesis which would lead people to disregard it.
What is Circuit Bending? Every synthesizer, whether analog or digital, is made of electronic components. These components were strategically placed to create the desired sound of the instrument. But are all of the sound possibilities realized by the original design? Circuit bending is the process of selectively short-circuiting an audio creating device to create a new sound that it was not originally made to create. A very common example of this is the Texas Instruments Speak and Spell. The sound chip in an 80’s Speak and Spell is made to say words that are typed into the toy. You may, however short-circuit the toy to create a synthesizer. It can be made to create very lush, ambient moods or chaotic noise. You may also fit it with MIDI controls so that it can be controlled by a common MIDI keyboard and therefore easily seen as a musical instrument.
The real creativity comes not from altering toys but from altering actual working electronic instruments. The most common and valuable being the Roland TB-303. This Bass accompaniment made by Roland was a flop for consumers from the beginning, it didn’t sound like a bass for one and it was somewhat expensive product aimed at an audience that didn’t have a lot to spend. Needless to say, not many were sold. Many years later, however, people began circuit bending these instruments and adding options. The most valuable being the Devil Fish which is as expensive as three non-altered TB-303s. By adding things like LFOs and ADSRs along with variable pitch control this is not an intricate part of modern electronic music. It is where the term Acid Techno came from. One of the presets on the original was called acid drop and is about the only thing used in it’s current for.
The reason that circuit bending is not considered synthesis is for this simple reason. On a digital chip a sine wave may be read in binary as 1000, 1010, 1100, 1010, 1000 and on and on where a 1 is 1 volt and a 0 is zero volts but when you bend the circuit it will send distorted code to the chip like 1000, 2000, 3000 etc. the binary chip has no idea what to do with more than 1 volt so when it sees 2 or 3 volts it acts erratically and unpredictably. It will still spit out audio but what will the audio sound like? It may be different every time. Imagine if middle C on a piano sometimes played middle C but sometimes played A1 and sometimes sounded like a cat screeching and sometimes did nothing and on occasion caught fire. This is what circuit bending is. But sometimes the short-circuit that you find may be the same every time because it is still within the normal operating range of the chip. This would normally be a more useful “bend” than a chaotic one but that depends on the operator. Basically, the way circuit bending modulates, unlike any other synthesis, is be digitally altering the signal before it is converted back to analog instead of modifying an analog signal.
Every synthesis technique was developed to create never before heard sounds. So in a way they are all different means to the same end, unique creativity. People will continue to want something new and there will always be new types of synthesis emerging.