1Physics 1240: Sound and Music Lecture Notes Mon 3/9/06 [email protected] Exam II Next Thurs 3/16/06 Next time: Modular synthesizers, review for Exam II Today: Electronic Synthesis Outline • Historical Perspective: from the Theremin to analog, to digital, then sequencers, then back to virtual analog • Synthesis of the "steady tone" waveform o Digital sampling o Additive synthesis o Subtractive synthesis • Attack Decay Sustain Release (ADSR) • Frequency and Amplitude Modulation (AM, FM) • Modular synthesis o Voltage controlled oscillator o Voltage controlled amplifier o ADSR unit A musical instrument that foreshadowed much of the electronic music revolution was the Theremin, developed in the early 1920's by the Russian physicist Leon Theremin. Theremin was also a cellist. The player moves her hands through two sensors, which work by sensing changes in the capacitance of an electrical circuit. The Theremin became quite popular internationally and was the original source of that spooky science fiction sound. RCA and Moog manufactured them. In fact, this is how Robert Moog began his career in electron synthesizers. The Theremin not only was the first instrument to produce a purely electronic sound, it was also a foreshadow of future interfaces for controlling electronics and computers that go beyond the more conventional keyboard interface. In the 1970's modular analog synthesizers became popular, which are a collection of interconnected voltage controlled oscillators, voltage-controlled amplifiers and filters. We will go into some detail how sound is synthesized by these analog synths because it gives a good understanding of how more modern synthesis is done. The analog synthesizers of the 70's were limited in what acoustic sounds they could realistically reproduce. But, in some ways they were revolutionary, like with the electric guitar, they produced sounds that were artistic and interesting and became quite widely used. In the 1980's digital technology had progressed to the point that digital sampling became virtually the only technology used for synthesis. A sample of an acoustic sound is made, and then modified. To get different pitches, the frequency of the sample is simply changed. Digital synths allow a range of modifications to the original waveform (we will talk about what can be modified later). A good digital synth uses different samples at different frequencies to properly model how an acoustic instrument's waveform changes as it plays higher or lower notes. A cheap synth does not have more than one or two samples for a given instrument. When you play high or low notes it, becomes clear that the synth sound is fake. On the first page of the popular book "The Science of Musical Sound" the author, John Pierce, he says, "Today analog sound is obsolete." He wrote this in 1992, and it is ironic that since around 1997, or maybe a year or two earlier, the analog sound is hot. There is a huge market for what are called vintage analog synths. The mini-Moog is a classic. People pay way too much for these old synths. The digital synthesizer industry has responded and now new (since 1998 or so) digital synths have analog emulation. Why? Well, a2few things. First, the old analog sound was a pleasing sound. Second, it was a truly unique instrument. The "fat Moog" sound, for example, has its own unique identity. It is not in any way, a reproduction of an acoustic sound (maybe a cross between an electric bass and an organ). Third, it is a creative medium for the musician. She is not simply sampling; she is creating new sounds. Synthesis We will talk about four common forms of synthesis. For simplicity we will only discuss steady-tones for now, and come back to how we handle the transients of the attack and decay of the note. 1. Digital sampling 2. Additive synthesis 3. Subtractive synthesis Digital sampling we have already talked a lot about. A digital sample of the actual waveform is saved in the synthesizer's memory and the frequency can be changed to produce any given note. Experiment: digital synth playing a sample of an acoustic instrument. Additive synthesis is using the addition of sine waves to produce a given timbre. Remember that we talked about how any steady tone that produces the sensation of a distinct pitch is made up of sine waves with various amplitudes that form a harmonic series. You can think of this as simply connecting a bunch of signal generators producing sine waves and adjusting the frequencies to form a harmonic series, then varying the amplitudes of each (the volume) to produce any given timbre. This seems so easy in theory, but it gets hard to control, since you have so many amplitudes to set. In fact, when digital sampling came along, it was such a sensation, because the additive synthesis was pretty much a failure, except in the laboratory. Subtractive synthesis is the technique most widely used for analog or virtual analog synthesis. Here, a waveform with a lot of spectral content (many higher harmonics with large amplitude) is passed through a filter. In many ways like how the vocal tract shapes the spectrum of the waveform generated by the vocal cords. The filter "shapes" the frequency spectrum and this gives different timbres. This is a very effective way to control and create sounds, but it is still hard to actually reproduce an acoustic sound. The filters are fairly complex, and adjusted by ear, not by viewing the frequency spectrum. Although visual spectral shaping could be done in principle. I'll demonstrate this subtractive synthesis using a "real-time virtual synth" called Vaz. Vaz is not all that high-tech as far as virtual synths go, but it is simple and cheap (around $39). The better ones seem to be Buzz, Generator and Reality. You can check them out and download at least demos on the "virtual synth page" which I have a link to in our course web site. Buzz is free and looks pretty neat, but it is 98 based and I use NT. You can hook a keyboard through a MIDI (Musical Instrument Device Interface) cable which connects to the joystick connector on the sound card. There are also parallel port interfaces, but they are more pricey, ($100-200). Attack Decay Sustain Release One thing we have mentioned is that transients play a critical role in how musical instruments sound. Pierce mentions how to produce a realistic timbre of a given instrument; you need to accurately account for the fact that different harmonics fade
View Full Document
Unlocking...