Sound Synthesis
Synthesis
The combining of the constituent elements of separate material or abstract entities into a single or unified entity.
Frequency Spectrum
The distribution and relative amplitudes of all sounding frequencies.
Harmonic Spectrum
The distribution and relative amplitudes of only the sounding harmonics (whole-number multiples) of a fundamental frequency.
Filter Envelope
These same concepts apply to the filter envelope—except, instead of modulating the amplifier's volume, these envelopes will modulate your synth's filter and the cutoff frequency of your sound. And the full shape of a synth's sound will be found through a combination of your amp and filter envelopes.
Control Switch
Similar to key switch, but using some sort of MIDI controller like the Modulation Wheel or Sustain Pedal to switch between types of samples.
Amplitude Modulation (AM) Synthesis
A Modulation is a change from one condition to another. In AM Synthesis, the modulator wave controls the amplitude of the fixed-frequency carrier wave. If you turn the volume knob up and down several times per second on an amplifier, you create an amplitude modulation known as tremolo, where you can actually hear the amplitude rising and falling. The concept of AM synthesis is the same as creating a tremolo, except at a much faster rate. The modulator in this case is in the audio-rate range (greater than 20 Hz). When the amplitude of the carrier's oscillator signal rises and falls that quickly, we no longer hear a tremolo. Instead, we hear a new set of frequencies on either side of the original carrier frequency, known as sidebands, that are a result of the peaks and valleys produced by the modulated amplitude. With AM synthesis, the carrier oscillator usually produces a wave at a fixed frequency that is considerably higher than the modulator's highest frequency. The modulator, in turn, modulates, superimposing its shape onto the amplitude of the carrier wave. As it does so, sideband frequencies are produced on either side of the carrier frequency that are the sum of the carrier and modulator frequencies (fc + fm) and the difference of the carrier and modulator frequencies (fc − fm). Because of the limited number of sideband frequencies created, few modern synthesizers implement amplitude modulation as a sound-making technique. With the exception of a specific form of AM synthesis known as ring modulation. Since the use of AM synthesis as a sound source is rather rare, you often see the oscillator designation RM (ring modulation) used instead. In most cases, the ring modulation oscillator will have a control for the intensity, or depth, of the effect. Moving the control to its lowest setting usually produces AM synthesis, while moving the control higher produces ring modulation
Waveshaping Oscillators
A basic wave-type oscillator whose output is manipulated by means of a lookup table. Waveshaping offers the advantage of needing only a single basic oscillator and a lookup table to create a continuous loop Passes a basic oscillator wave through a lookup table, changing the incoming x values of the wave to the y values from the table. The table data can be derived from either a formula or a digital audio sample data can be derived from either a formula or a digital sample.
Linear & Nonlinear Systems
A linear system is one in which the input and output remain proportional. In other words, for every change in input, there is a proportional change in output. A nonlinear system does not maintain the proportional relationship between input
Wavetable Synthesis
A method of generating waveforms through lookup tables. Many software synthesizers use wavetable synthesis where these digitized waveforms are organized in a bank or table, accessed through a sequencer. Wavetable: generates a set of x-y values for a lookup table based on a mathematical formula. As with the table lookup oscillator, the wavetable oscillator repeatedly reads through its table to generate the sound wave.
Subtractive Synthesis
A method of sound synthesis in which partials of an audio signal (often one rich in harmonics) are attenuated by a filter to alter the timbre of the sound. Subtractive synthesis is the type generally used by analog synthesizers.
Ring Modulation
A ring modulator multiplies two signals together to create two brand-new frequencies which are the sum and difference of the input frequencies. Since the use of AM synthesis as a sound source is rather rare, you often see the oscillator designation RM (ring modulation) used instead. In most cases, the ring modulation oscillator will have a control for the intensity, or depth, of the effect. Moving the control to its lowest setting usually produces AM synthesis, while moving the control higher produces ring modulation
Portamento/Glide
A slide between pitches as when a violinist keeps a left-hand finger pressed down on a string while sliding to a new note. On a synthesizer, the portamento control—often called the glide control.
Additive Synthesis
A synthesis method that builds complex waveforms by combining sine waves whose frequencies and amplitudes are independently variable. Overlapping sine waves at different frequencies beat, or cancel each other out. Doing this with lots of oscillators can create unique and interesting sounds.
Modular Synthesizer
A synthesis system comprised of self-contained modules. These modules are specialized electronic circuits useful to producing and shaping sound. Modules are manually linked together using patch cords. Modular systems allow for extreme flexibility and customizability of your system but requires extensive programming knowledge and patience.
Transducer
A transducer converts one form of energy into another form. A microphone is a type of transducer that converts the oscillating air pressure of a sound wave into oscillating electrical energy, and a loudspeaker converts the electrical energy back into sound.
Triode
A vacuum tube that has three electrodes. When the cathode is heated in a triode, it releases electrons into the vacuum of the glass tube. Applying a small positive charge to the grid causes a large flow of electrons to the anode, producing a strong current from the anode's terminal (left). A negative charge on the grid disrupts the flow of electrons to the anode (center). A small, oscillating signal applied to the grid causes the electron flow to oscillate as well, but on a much larger scale (amplification) than the original grid signal (right).
Wavetable Oscillators
A wavetable oscillator is similar to a table lookup oscillator. It too reads through a lookup table containing a single cycle of a waveform in a continuous loop. The true wavetable oscillator reads a user-generated array, or table, of x-y coordinates to create a wave cycle in real time, while the table lookup oscillator reads a prerecorded digital sample of a wave cycle. Wavetable oscillators can generate complex waves that create the effect of combining multiple waveforms together.
Linear & Nonlinear Soundwaves
With sound waves, the term linear is used to describe the natural interaction of sound wave energy where the composite sound is simply the sum of all the individual sounds. In the purest sense, the parallel combination of oscillators on the left in Figure 3.1 produces a linear sound response. Nonlinear sound refers to sound waves that are combined in such a way that the output sound is quite different from the input. Nonlinear sound usually contains a number of frequencies that are not part of the original waves. It should be noted that the human ear is also nonlinear in responding to sound. When sound energy consisting of combined frequencies strikes our eardrum, especially at high amplitudes, we frequently hear so-called combination or Tartini tones as a result of the nonlinear distortion produced in our ears.
Low Frequency Oscillator (LFO)
is a time-variant controller which by outputting frequencies in the subaudio range can induce change in a generator, a filter, or an amplifier depending on which module is the destination of an LFO control signal. Low-Frequency Oscillator, or LFO. Although these are, in fact, oscillators, they usually operate well below our range of hearing (less than 20 Hz) and are not used to make sounds but to control other aspects of the synthesis process. LFO, audio-rate modulation oscillators are not used for sound, but to control other elements in the synthesizer. LFOs and audio-rate modulation oscillators are important tools in synthesis, and we will discuss them in great detail in coming chapters.
Multisampling
recording more than one sample per note of an instrument. These might be different articulations or timbres, or even multiple recordings of the same note and articulation, to provide some variety.
Table Lookup Oscillators
samples of sine, triangle, sawtooth, and pulse waves stored as tiny digital files loaded into a quickly accessible memory location called a lookup table. By simply changing the speed with which the instrument reads and loops through the table, it can synthesize digital representations of analog oscillators. For example, if you load a single cycle of a triangle wave into a lookup table, and then read through that table two hundred times a second, you create a 200 Hz triangle wave. The part of the synthesizer that reads through the lookup table hundreds of times per second, creating a sound wave based on the values in the table.
Bandpass Filter (Peak)
A bandpass filter lets frequencies pass that fall within a narrow range. All other frequencies are stopped.
Fourier Series
The series of sine waves added together to form a specific complex wave.
Oscillator Combinations
Although raw materials in the physical world can be quite valuable on their own, their true worth often comes from being combined with other raw materials to create complex and exotic compounds. The first involves simply adding or mixing the oscillator outputs together to create a more complex sonority. The second method uses one oscillator to control, or manipulate, another oscillator.
Analog
Analog comes from the same root as the word analogy. In common use, an analog is a continuous, uninterrupted representation of something else. For example, an analog clock represents the passage of time with continuously moving hands interconnected by gears so that, as the second hand sweeps around the face of the clock, the minute and hour hands slowly advance as well. With analog audio, the compression and rarefactions of air pressure are represented by continuous positive and negative changes in voltage. These changes can be recorded as grooves on a record, or as changing densities of magnetic flux on audiotape.
Parametric (Peak/Notch)
The parametric filter, sometimes called a "peak/notch" filter, is a bit of a hybridization of the bandpass and bandreject filters. Like the bandreject filter, the parametric filter has a control to adjust the width of the stopband. However, it has an additional control that allows the user to adjust the amount of attenuation, or even boost, in the stopband.
Sample-Playback Sources
As good as the hybrid oscillators were at improving the "realism" of sounds, their reliance on oscillator waveforms for the bulk of the sound still presented problems for users wanting to recreate authentic acoustic instrument sounds. Technically, these instruments are no longer using "oscillators" but are instead using sample-playback devices. Some manufacturers acknowledged this difference and began calling the internal sample-playback device something like a "source" or a "sound source." With the use of digital audio files, or samples, as their source material, a whole new breed of devices sprang up generally referred to as "samplers." Technically, the word applies only to devices that can record their own samples. Instruments that can load and play prerecorded samples are more correctly known as "sample players" or "ROMplers" (referring to read-only memory, the portion of the instrument's internal memory where the digital audio files are stored)
Partials, Harmonics, and Overtones
As opposed to sound waves produced by oscillators, sounds in the natural world are generally quite complex, consisting of a rich spectrum of frequencies known as partials. Generally the loudest, and most prominent, of these frequencies is also the lowest one, called the fundamental. The partials above the fundamental, often called overtones, are usually much quieter, and typically they get even softer as ascend in frequency until becoming inaudible. Some of these overtones have a distinct mathematical relationship to the fundamental frequency and are called harmonics. A sound that has a greater emphasis on harmonics tends to sound more "pitched," while a sound with more emphasis on the other partials tends to sound "unpitched," or as a "noise."
Key switch
Assigning a set of keys on the MIDI keyboard (usually the very lowest octave) to switch between types of samples. This is often used to change between articulations, as with legato and staccato, for example.
Velocity Map
Assigning multiple samples recorded at different dynamic levels to a single key. As you strike the key harder, not only does the sound grow louder but the playback source switches to a sample that has a "louder" tone quality to it.
ADSR
Attack: the amount of time it takes for the sound to go from silence to full amplitude when the synthesizer key is pressed, usually expressed in milliseconds or seconds Decay: the amount of time it takes for the sound to transition from full amplitude to Sustain level amplitude, usually expressed in milliseconds or seconds Sustain: the level at which the sound maintains as long as the synthesizer key is held down, usually expressed as a percentage of full amplitude, or as an audio level in dB Release: the amount of time it takes for the sound to completely die away after the synthesizer key is released, usually expressed in milliseconds or seconds Note that attack, decay, and release are all time values. Only sustain is an actual level.
Audio Filters
Audio filters are the most complex elements of almost any synthesizer. Audio filters work on the frequency spectrum of a signal, and like their mechanical counterparts they stop the undesirable frequencies from passing while allowing the desired frequencies to pass. Audio filters are used to refine the timbre of oscillator waves to produce the wonderfully rich sounds we have come to expect from a synthesizer. A basic audio filter has a specified range of frequencies, called the passband, that are allowed to pass through unaffected, and a range of frequencies that it stops from passing, known as the stopband. Because no audio filter can make a complete and abrupt change from passband to stopband, there is always a transition area between the two bands in which frequencies are allowed to pass, but with their amplitude gradually reduced, or attenuated, as the transition area approaches the stopband. Filter types are often displayed as an xy graph showing the passband and the stopband. Moving from left to right along the x-axis represents frequency from low to high, while moving up and down along the y-axis represents a gain change in the output amplitude of the frequencies. Passband values above the x-axis represent a boosting of frequencies, while values below the x-axis indicate an attenuation of frequencies. Most basic audio filters are designed so that the upper edge of the passband runs along the x-axis and drops away to infinity as it approaches the stopband.
Crackle Noise
Crackle noise is a special type of noise made up of randomly spaced bursts of noise at varying amplitudes. "analog" type of sound.
Hertz (Hz)
Cycles or waves per second, a measurement of frequency
Analog to Digital Converter
Device for sampling analog data and producing a digital sample of it. With analog audio devices, digital devices first convert the acoustic energy of sound into electrical energy via transducers. Next, the electrical signal is passed through a process known as the analog-to-digital converter (ADC), where the signal's amplitude is measured thousands of times per second. Each individual measurement is recorded in digital words, or bytes, consisting of groups of eight bits (an elision of the words binary and digits), representing a single amplitude value at a discrete point in time. Since computers and other digital devices cannot "read between the lines," they quantize the sample measurement by using the closest amplitude value on their internal scale when measuring the voltage.
Digital
Digital represents the concept that, instead of a continuous stream, the data are individual, momentary measurements. For example, a digital clock may show only the hour and minute, and remain in this state for sixty seconds until the clock updates at the next minute. With digital audio, the compression and rarefactions of air pressure are measured thousands of times every second to capture a momentary value. These measurements are recorded as groups of bits that can be stored as alternating high and low voltage on a digital tape, or as alternating pits (small indentations) and lands (flat spaces) on the surface of a plastic disc like a CD or DVD.
Digital Synthesizers
Digital synthesizers offered the distinct advantage that, since every sound value and setting on the synthesizer was an individual number, the data could be easily captured, recorded, and edited
Pulse Width
Most pulse wave oscillators also have a control to adjust the percentage of time spent in the positive phase versus the negative phase. If the pulse width (sometimes called the "duty cycle") is set so that both halves are equal to each other, the pulse wave is known as a "square" wave. The position of the pulse width is indicated variously by synthesizer manufacturers. Pulse width adjustment can have a significant impact on the timbre of the wave's sound.
Amplitude Envelope Controls
Envelope editor: use this window to manually position a breakpoint by dragging it with your mouse. When a note is triggered, the envelope proceeds through all the breakpoints until it reaches the penultimate one. At that point, it will either sustain until the key is released, or loop back to the first breakpoint if the Loop button is illuminated. The final breakpoint will always have an amplitude value of zero. Scale adjusts the time scale of the envelope editor window. When the Sync button is off, the grid values are seconds and fractions of seconds. When the Sync button is on, the values are quarter notes and subdivisions of quarter notes. Points menu allows you to select the number of envelope breakpoints in the editor window. Envelopes may have as few as four breakpoints to as many as nine. Loop button causes the envelope to repeatedly loop from the penultimate breakpoint back to the beginning of the envelope as long as the key is held down. Sync button synchronizes the envelope to the tempo of the VST host application and changes the scale of the envelope editor window from seconds to quarter notes. Function/Preset menu allows you to copy and paste envelopes into Curve positive creates a greater amount of change at the end of the transition between breakpoints. Curve negative creates a greater amount of change at the beginning of the transition between breakpoints. Pulse (1.0, 0.9, 0.75, 0.5) causes the amplitude to instantly jump to the numeric value and halfway between breakpoints to instantly jump to the next breakpoint value. Spike (1.0, 0.9, 0.75, 0.5) is similar to the pulse option, but instead of staying at the numeric value it immediately begins a curve toward the value of the next breakpoint. Flat remains at the breakpoint value until the next breakpoint and then instantly jumps to that value. Inverse pulse down remains at the breakpoint value until halfway to the next breakpoint, where it drops to zero until jumping instantly to the next breakpoint value. Inverse pulse up is similar to inverse pulse down except the value rises to maximum at the halfway point and then drops down to the next breakpoint value.
Envelope Generator Behavior Modes
Free-Run Mode: If an envelope generator is in free-run mode (sometimes called one-shot mode), then the envelope always completes the full attack and decay stages before jumping to the release. Since attack and decay are so vital to our perception of a sound, having free-run mode helps maintain the sound quality even with extremely short notes. Retrigger Mode vs. Multi-Trigger Mode: If your envelope generator is set to retrigger mode, then each new note will trigger a new envelope starting at zero amplitude, even if the old note is still sounding. If multi-trigger mode is on (this might be called mono mode or simply retrigger-off mode), then each new note's envelope begins wherever the previous envelope's level is when the new note is triggered. Retrigger mode (left image) starts the next note's envelope at zero even if the previous note is still being held. Multi-trigger, or mono, mode (right image) begins the next note's envelope wherever the previous envelope is when the new note is triggered. Single-Trigger Mode: Sometimes you want to create a sound that, when played legato, produces no discernible attack for the subsequent notes. It sounds as if one note just morphs into another. To accommodate this smooth type of articulation, some synthesizers implement single-trigger mode (often called legato mode). With single-trigger mode, for as long as a note is held down, all subsequent notes begin at the sustain level of the held note and do not have any attack or decay segments associated with them Once all notes have been released, the next note played will get a full attack-decay segment to its envelope. Single-trigger mode, designed for smooth melodic lines. There is some inconsistency in the use of the word legato to describe envelope generator modes. Some synthesizers use legato mode to describe the multi-trigger envelope, while others use it to describe single-trigger envelopes. If your synthesizer has a setting called legato mode, you will need to check the instrument's documentation to determine which version of legato mode is actually employed.
FM Synthesis
Frequency Modulation. One of the early forms of digital synthesis. FM is a type of modulation in which the frequency of a continuous carrier wave is varied in accordance with the properties of a second (modulating) wave in order to generate complex waveforms. When the modulating wave is in the audio range (above 20Hz or so), FM is perceived as a change in tone color, and that used in FM synthesizers to create their unique sounds. Modulator oscillator controls the amplitude of a carrier oscillator. As the name suggests, frequency modulation synthesis uses a modulator oscillator to control the frequency of a carrier oscillator. Almost all of us use a subtle amount of frequency modulation—typically at a rate of about 2-5 Hz—when we sing or play an acoustic instrument. We just call it vibrato. We use the modulator oscillator to rapidly raise and lower the frequency of the carrier oscillator, and just as in AM synthesis, the modulator is operating in the audio-rate range (greater than 20 Hz). With FM synthesis, as you raise the level of the modulator's amplitude, the frequency of the carrier goes up. As the modulator amplitude goes down, so does the carrier's frequency. In FM synthesis, as the modulator forces the carrier wave to rapidly speed up and slow down, a complex set of sideband frequencies is created. Depending on the amplitude of the modulator, some of these sidebands can be quite strong and even mask the original carrier wave frequency (again, the modulator is not heard). As with oscillator sync, the modulator and carrier are usually set at a fixed ratio to each other.4 If their ratio is a harmonic one (e.g., an octave, a perfect fifth, a perfect fourth, etc.), the carrier produces a strongly pitched sound. However, if the modulator is tuned to some inharmonic ratio, the carrier tends to produce a noisy, clangorous type of sound. It is this latter category that has been particularly attractive to synthesizer users as these tones work wonderfully for things like bells, metallic sounds, basses, electronic pianos, and all manner of special effects. The two primary controls in FM synthesis are the harmonicity ratio, the ratio of the modulator's frequency to that of the carrier (fm/fc), and the so-called modulation index, the ratio of the modulator's amplitude to its frequency (am/fm). Think of the speed with which you move the wheel up and down as analogous to the harmonicity ratio, while the distance you move the wheel up and down, away from its center point, is like the modulation index. With FM synthesis, we frequently start with an existing sound and start making small adjustments to the harmonicity ratio and modulation index rather than trying to start a sound from scratch. Also, most FM synthesizers allow users to modulate the modulators. Both AM and FM synthesis are types of audio-rate modulation synthesis, and on some instruments they are actually labeled as such rather than as AM or FM.
Blue Noise
Halfway between violet and white is blue noise, the opposite of pink noise. Here, the energy of each successive octave is doubled. Because blue noise provides a little bit of low-frequency content, with a lot of high-frequency emphasis, it can sometimes add "shimmer" and "sparkle" to things like cymbal sounds and reverberation.
Pink Noise
Pink noise is noise that continuously reduces its energy level by half as you pass upward through each octave. "pink" because it is halfway between red and white noise. It sounds a bit like a heavy rainfall Pink noise also emphasizes lower frequencies, but the drop in amplitude at higher frequencies is not as extreme as red noise. Since every octave has twice as many frequencies as the octave below, and pink noise drops by half with every ascending octave, we hear it as having an equal energy level across the entire frequency spectrum.
Triangle Wave
If the oscillator wave's voltage is allowed to transition from one pole to the other at a steady rate of speed, and then abruptly change direction and head toward the other pole at a steady speed, a triangle wave results. Because of the quick change of direction at the top and bottom of the wave, the sound tends to have a brighter timbre than the sine wave.
White Noise
If you randomly combine all frequencies of the audio spectrum equally, you create white noise. White noise has the same amount of energy at every frequency. White noise actually has a consistent energy level for all frequencies. However, since there are twice as many frequencies in each ascending octave, we hear white noise as having a high-frequency tilt.
Lookup Table
If you split a single cycle of a sound wave into 256 slices (remember, 0-255 is actually 256 units) and store the amplitude for each individual slice in the appropriately numbered memory slot, you make the wave cycle easily accessible. The table sees the incoming number from the count and outputs the stored amplitude value from the numbered location. The wave needs to be generated (stored) only once. Loads a set of values (y) at individual memory locations (x). When it receives a number that matches one of its memory locations, the lookup table outputs the value stored at that memory location.
Red Noise aka Brown Noise
Red noise (sometimes called brown, or Brownian, noise) exhibits a substantial drop in energy level as the frequency ascends. Because the emphasis of this noise is at the low end of the frequency spectrum, it is named red, like the low end of the light spectrum. Red noise is often described as having a low "roar," like a distant waterfall. It comes in handy for things like "boomy" sound effects, bass drums, and other low-frequency sounds. Red noise emphasizes the lower frequencies and rapidly drops in amplitude at the higher end of the frequency spectrum.
Pulse (or Square) Wave
The pulse wave is nothing more than a two-state wave: it's either fully positive or fully negative. There is no perceptible transition between poles; rather, it jumps abruptly from one pole to the other. Because of the two abrupt transitions per wave oscillation, pulse waves also tend to be rather bright-sounding, and because they produce only odd-numbered harmonics they have become popular for creating a number of woodwind-type instrument sounds.
3 Primary Characteristics of Oscillators
In addition to its wave type, the typical oscillator has three primary characteristics that may be easily controlled: wave phase position, wave amplitude, and wave frequency. All three have been manipulated as a synthesis technique to create complex sounds.
ADADSR, ADSHR, AHDSR, and DADSR
In the first example, a second attack and decay is added before the sustain segment. This design helps create the double-attack sound commonly associated with brass instruments as seen with the "trumpet" envelope. Other variations add another sustain segment known as hold (labeled H) to the envelope. Some envelope generators use this hold section to delay the release stage when the key comes up, and others put the hold between the attack and decay stages to create two sustained levels. The initial D on the DADSR envelope is a delay that offsets the beginning of the envelope by a specified time value after the key is pressed.
Hybrid Oscillators
It is possible to create a decent representation of the middle, or sustain, portion of many of these instruments' sounds using only oscillators, especially wavetable oscillators. However, they still do not sound very realistic. It has been known for some time that the first few milliseconds (abbreviated ms) of a sound's attack is critical to our perception and recognition of the sound. With this idea in mind, some synthesizer manufacturers began experimenting with a system that first played a very short sample of a sound's attack and then smoothly transitioned into an oscillator waveform. This hybrid oscillator approach, sometimes called Sample+Synthesis (or S+S), did a nice job of using limited memory capacity for maximum effectiveness, while still relying on a basic oscillator to produce the majority of the sound.
Key Tracking
Key tracking simply refers to the physical location from lowest note to highest note on the synthesizer's keyboard. We commonly think of key tracking as controlling the pitch of the synthesizer. Every time we go up from one key to the next adjacent key, the pitch rises by one semitone. Like velocity, the notes on the MIDI keyboard are numbered from the lowest note of 0 to the highest note of 127. The beauty of both of these data types is that in addition to controlling loudness and pitch, we can also use them to control an envelope generator. *The eighty-eight notes of the piano keyboard are equal to MIDI note numbers 21-108. When we assign velocity data to the envelope's amplitude, the overall height of the envelope changes with the velocity. As the envelope becomes taller, the slopes are steeper, giving the sound slightly more percussive quality for the louder notes. In general, we expect higher velocities to produce taller and louder envelopes. Many instruments also provide the capability of inverse scaling. In this case, the harder you strike the key, the quieter the sound gets. Such an effect can be quite handy when layering sounds. You can create a normal envelope for the loud sound and an inverted envelope for the soft sound. Then, as you strike the key with more and more force, the soft sound fades into the background while the loud sound becomes more prominent. Envelope scaling often uses an arbitrary set of values such as 0-100 to determine how much scaling occurs. In most cases, a value of zero means no velocity scaling. As you increase the scaling value, the effect becomes more pronounced. Key velocity may also be used to adjust the time, or rate, values on many synthesizers. With most acoustic instruments, the louder you play them, the more percussive their attacks become. However, as happens on many instruments, the louder you play them, the longer their release becomes as well. For this reason, most synthesizers that allow velocity to affect envelope times let you set the value for the individual envelope stages. For example, you may want high key velocities to shorten the attack and decay times so that the sound grows more percussive. However, you might also want the release time to increase so the sound takes longer to die away. To accommodate this, you will find that the controls allowing you to manipulate envelope stage times usually have both positive and negative values. As with velocity scaling, many synthesizers implement key tracking for their envelope generators as well. Just as you can use velocity data values to control an envelope's amplitude and times, you can use the data values of the MIDI note numbers to control the envelope. Typically, when an instrument uses key tracking to manipulate an envelope, middle C is the center point. Notes above middle C increase the amplitude, and notes below decrease the amplitude of the envelope. You can use amplitude envelope generators to create articulations and shape synthesizer sounds into musical notes.
Allpass filter
Let's all frequencies pass but inverts the phase of the frequencies at some user-specified point in the frequency spectrum. When audio goes through an allpass filter, it begins perfectly in phase. As you move up through the frequency spectrum and approach the designated changeover point, the phase of the frequencies begins to shift until they are 180° out from their original position. Most people can't hear this phase shifting so, In most audio devices, these filters are not used to alter the timbre of a sound, but to compensate for signals that have been shifted in phase by some other aspect of the audio process.
Sawtooth Wave
Like the triangle wave, the sawtooth wave begins at one pole and transitions smoothly to the opposite. However, when it gets to the opposite pole, it abruptly and nearly instantaneously jumps back to the first pole and starts the process over. With its abrupt directional change, the sawtooth wave tends to be considerably brighter, and "buzzier," than the triangle wave.
Logarithmic and Exponential Curves
Logarithmic curves have a greater rate of change when the input value is lower, while exponential curves change more when the input value is higher. Logarithmic curves have a greater rate of change when the values are low and are commonly used for ascending stages in an envelope. In contrast, exponential curves have a greater rate of change when the values are high and are normally used for descending envelope stages.
Segment Curves
Many envelopes have straight line segments. However, some synthesizers allow you to curve these segments. Most people, when making a musical transition like a crescendo, decrescendo, accelerando, or ritardando, do not make the change in a straight, linear fashion. We often begin with a greater rate of change at the beginning, and then lessen the rate of change as we approach the end of our transition. Using segment curves, instead of straight lines, we can apply a bit more of a human touch to our envelopes. On some instruments, the attack, decay, and release segments are curved by default, and on other instruments they can be individually curved by the user. Flat segments like sustain, hold, and delay usually remain flat and do not curve.2
Multiple Envelope Generators
Multiple envelope generators, especially ones capable of generating multistage envelopes with onset delays, allow wonderfully complex sound shaping, and they are particularly effective when working with additive synthesis where we use the delays and complex envelope shapes to cross-fade between partials, creating a richly evolving sound. Some instruments have separate envelope generators for the individual oscillators and another final envelope generator for the combined sound.
Amplitude
Neither our ears nor our audio equipment captures a sound's frequency. They capture a sound's changes in amplitude. Our brain interprets those changes in amplitude (large changes = loud, small changes = quiet, frequent changes = high frequency, infrequent changes = low frequency), but it is amplitude change to which we physically respond and that we capture with audio equipment. It is extremely important to remember, then, that when we discuss the way sound waves behaves, we will be talking about how their amplitude values add to, subtract from, and interact with each other.
Nyquist Theorem
Nyquist theorem, states that the digital sampling rate must be at least twice as high as the highest analog frequency to be recorded. In other words, since most humans can hear sound up to about 20 kHz, the Nyquist theorem says our sampling rate must be at least 40 kHz in order to capture the frequency range. That "highest analog frequency to be recorded" (half the sampling rate) is similarly known as the Nyquist frequency.
Digital to Analog Converter
On playback, the individual samples are read into the digital-to-analog converter (DAC) and turned back into voltage levels to be transduced into sound by the loudspeakers. Although you and I might connect the dots with direct, diagonal lines, a computer cannot see the next value until it happens, so the voltage remains at the current sample value until the next sample, at which point it immediately jumps to that value. The resulting differences between the original sound wave and its digitized version are known as the quantization error.
Wave Phase
Oscillator waves are measured in degrees through their cycle. Like starting at the top of a compass and traveling all the way around the circle, the beginning of a wave is measured at 0°. Halfway through the wave is 180°, and the end of the wave (beginning of the next wave) is 360° (or 0° again). At first glance, the starting position of a wave may not seem important, but as we will see in the next chapter the starting phase position of a wave can have a huge impact on a sound when oscillators are combined together.
Compression & Refraction
Sound energy consists of two phases, called compression and rarefaction. When two sound waves are at the same point in their compression or rarefaction phase, we describe them as being in phase with one another. When this happens, their two energy levels complement each other and are added together (linear combination) to create a larger amplitude of compression or rarefaction. By contrast, when one wave is in its compression phase while the other is in its rarefaction phase, we refer to the waves as being out of phase with each other, and the positive energy from one wave cancels the negative energy of the other wave. In fact, if you combine equal amounts of positive and negative energy—as can happen when a sound wave is combined with a 180° out-of-phase copy of itself—the two waves completely cancel each other out, and you have total silence.
Oscillator Tuning
The FineTune slider raises and lowers the oscillator frequency within a single semitone.
Amp Envelope
The amp envelope, by affecting your synth's internal amplifier, will control the volume. The Attack parameter here will determine whether your note will come in immediately or gradually, while the Decay and Sustain knobs will set how present the sound will continue to be after the initial hit. The Release will determine how quickly or slowly the sound expires after your release the key.
Bandreject Filter (Notch)
The bandreject, or "notch," filter allows all frequencies to pass, except for those in the stopband. Bandreject filters are used to remove a specific range of frequencies, and they are sometimes used to isolate and control unwanted noise in a signal.
Sine Wave
The basic type of electrical wave created by an oscillator is one in which the voltage level changes smoothly and continuously between positive and negative poles without any abrupt changes in either direction. Because of its relationship to the motion around a circle, this type of energy wave is commonly known as a sine wave, although its technical name is a sinusoidal wave. Because sine waves have a smooth motion without any abrupt changes in direction, they produce only the tone of their root frequency and tend to be rather dull and dark sounding. Pure sine waves rarely occur by themselves in natural sounds, but are quite common in electronic devices like synthesizers.
Controlling One Oscillator with Another Oscillator
The other common method of combining oscillators is to have one (or more) oscillator(s) manipulate, or control, another oscillator. With this approach, we usually do not hear the controlling oscillators, but only the sound of the controlled oscillators after being manipulated. Wonderfully rich, fascinating sounds can be created with only a small number of oscillators thanks to their nonlinear response, whereas to create equally exotic sounds with a linear approach such as additive synthesis would require a huge number of oscillators.
Quantization Error
The error that is introduced during digitization. Also known as quantization noise. Although you and I might connect the dots with direct, diagonal lines, a computer cannot see the next value until it happens, so the voltage remains at the current sample value until the next sample, at which point it immediately jumps to that value. The resulting differences between the original sound wave and its digitized version are known as the quantization error. One way to reduce the quantization error is to increase the bit depth, or bit resolution (the number of zeroes and ones used in the amplitude measurements). Another way to reduce the quantization error in digital audio is to increase the sample rate, or the number of samples captured per second. The higher the sample rate, the closer together the measurements, and the more accurate the digitized version of the wave.
Allpass Filter
The filter inverts the phase of the frequencies at some user-specified point in the frequency spectrum. When audio goes through an allpass filter, it begins perfectly in phase. As you move up through the frequency spectrum and approach the designated changeover point, the phase of the frequencies begins to shift until they are 180° out from their original position. Our ears are not particularly sensitive to phase positions in a sound, and most people will not hear any difference in a sound processed by an allpass filter. In most audio devices, these filters are not used to alter the timbre of a sound, but to compensate for signals that have been shifted in phase by some other aspect of the audio process.
Fundamental Frequency
The loudest, and usually most intense, frequency of a complex sound; most often perceived as the sound's basic pitch.
Envelope Generators
The many changes that occur from the beginning of a sound until it completely dies away make up what is known as the sound's envelope. Within that envelope there will often be changes in elements such as amplitude, pitch, and timbre. There might be envelopes for a sound's amplitude, pitch, timbre, or any other element that changes through the sound's duration.
Amplitude Envelope Generators
The most prominent of the various envelopes, though, is the amplitude envelope, which gives a sound its beginning, middle, and ending shape. The amplitude envelope is so prominent that most people, when using the phrase sound envelope, actually mean the amplitude envelope. There are four distinct sections of a musical tone's amplitude envelope: the attack (the section of time in which the tone is first heard) The attack section had two subsections: the initial, upward slope that usually leads to the peak amplitude of the sound, and the decay, a downward-sloping section that descends from that peak into the sustained state followed by the release of key being held down. AEG go by a number of names and abbreviations, among them EG, AmpEG, ADSR, VCA (voltage-controlled amplifier), and DCA (digitally controlled amplifier). Think of the amplitude envelope generator as a remote controller for an imaginary volume knob on an oscillator. When a synthesizer key is pressed, the envelope generator's control signal causes the "knob" to quickly turn up from zero to full amplitude (attack), and then down (decay) to a predetermined level (sustain) until the key is released. At that point, the control signal turns the knob down from the sustain level back to zero (release).
Frequency
The number of times the sound wave completes a cycle (both compression and rarefaction phases) in a second determines the sound's frequency. Frequency is a by-product of changing amplitude. The faster a sound wave rises and falls, the higher we perceive it to be in frequency.
Violet Noise
The opposite of red noise (low-frequency emphasis) is violet noise (high-frequency emphasis). This type of noise has an extremely high-frequency tilt, with little or no low-frequency content. Violet, sometimes called "purple," noise can be useful for creating sibilance, or "s" sounds, when judiciously applied to a sound. Violet noise emphasizes the higher frequencies and rapidly drops in amplitude at the lower end of the spectrum.
Articulation of a Sound
The shape of the amplitude envelope creates the articulation of a sound. It determines whether a sound is short and percussive, long and legato, or anything in between. In general, the shorter the times (or steeper the slopes), the more percussive and staccato a sound becomes. As you lengthen the time values (or flatten the slopes), the quality of the sound becomes more legato and sustained
Delay Spread
The time delay between the reception of the main RF signal and a reflected RF signal.
Granular Synthesis
The use of tiny fragments of digital sound recordings (usually less than 50 ms in length) known as grains. Each grain is so short that when heard individually it often sounds like a click or a short hiss. Combine it with hundreds of other grains at the same time, and you can create a wonderful, swirling "cloud" of sound.use of tiny fragments of digital sound. When played slowly, the grains tend to produce an atmospheric cloud of sound that evolves over time, and when played quickly they tend to produce pitched notes with exotic timbres. One popular technique in granular synthesis is to slowly move the section window around while the grains are playing, to produce these sonic changes.
Sound vs Audio
The word sound refers to the natural acoustic phenomenon of vibrations moving—usually through air—to our ears. Audio, by contrast, refers to the capture, storage, and reproduction of sound through electronic means. Audio equipment represents the changing air pressure of sound with a changing electrical voltage inside the components. Sound radiates away from the sound source in a sphere, moving outward and then recoiling inward, then outward, and inward, and so on. Every sound wave consists of two parts, or phases: compression (when the air pressure is greater than ambient air pressure) and rarefaction (when air pressure is less than ambient air pressure).
Timbre/tone color
Timbre is often defined as the tone quality, or tone color, of a sound, although it is sometimes described more by what it is not than what it is. A sound's timbre is primarily the result of the relative amplitudes of all the sounding partials, what acousticians call the frequency spectrum.
Beat and Combination Frequencies
We also have an additional rising and falling amplitude as a result of the linear interaction between the two waves producing a combination tone. If the resultant rising and falling amplitude of the combination tone occurs more than twenty to thirty times a second, we sometimes hear a new frequency. On the other hand, if the combination tone produces a rising and falling amplitude fewer than twenty times a second, we usually hear a pulsing or throbbing in the overall amplitude of the two original sound waves, known as beats. These beat frequencies occur whenever there are two similar tones at a close frequency interval; just like the difference frequency, they can always be calculated with the formula fbeat = f1 − f2. Beat frequencies are one way musicians tune to each other. If you listen to two instrumentalists tuning, you will notice that they both play the same note, and if they are not perfectly in tune a beat frequency occurs. Usually, one player (the principal) will hold steady while the other player slowly adjusts the tuning on their instrument until the beats disappear. At that point, they are perfectly in tune.
Aliasing
When the sampling rate is not high enough to capture two or more samples per wave period, artificial lower frequencies called aliases are produced instead of the original high frequencies.
Oscillator
Without oscillation, there is no sound. Strike a tuning fork, pluck a guitar string, blow through a clarinet, sing, clap your hands, or even slam a door, and you create oscillations. These oscillations, also called vibrations, happen first in the device (the tuning fork, string, clarinet, throat, hands, door) and then transfer into the surrounding air, creating oscillating changes in air pressure. As that oscillating air pressure travels outward from the sound's source, it reaches our ears where the rapid changes in air pressure cause our eardrums to oscillate, sending nerve impulses to the brain. If the oscillations happen more than about twenty times per second (20 hertz), we perceive them as sound. In the audio and recording world, oscillations go through several intermediate steps where they are converted to oscillations of electrical voltage by a microphone, or other type of transducer. Once the wave is converted to an electrical signal, it can be stored on electromagnetic media such as audiotape or a computer drive. When the audio is played back, the oscillations in voltage are converted back into changes in air pressure by another type of transducer called a loudspeaker.
Tartini Tones
are a psychoacoustic phenomenon that varies greatly from person to person and is also highly dependent on the amplitudes and frequencies of the original sounds being produced. Although some composers and electronic musicians have attempted to create music that produces such additional frequencies, the result tends to be rather unpredictable, since it depends on the listener.
Keymap
assigning various samples to specific keys on the controller (usually a MIDI keyboard).
Multiwave Oscillators
combine similar, or even identical, oscillator sounds, but at slightly different frequencies? The multiwave oscillator is usually created with a minimum of three oscillators, The second oscillator is tuned a few cents higher than the first oscillator, and the third oscillator is tuned the same number of cents lower than the first oscillator. For even more richness, add two additional oscillators. Tune one of them a different number of cents higher than the original oscillator, and tune the other the same number of cents lower than the original. Again, these two new outer oscillators should be equidistant in cents from the original. The more you add to the mix, the richer the sound becomes.
Round-robin
the ability of a sample-playback synthesizer to switch between multiple versions of the same sample so that repeated notes do not all sound alike.
