Chapter 2: Synthesizer basics

Synthesizers are actually not particularly complicated devices. Neither do modular synthesizers! A lot works the same way as with other instruments: A tone – maybe just a noise – is produced, and then we can change the sound and volume of this tone. While we play it! With a “conventional” instrument, we would do this by blowing harder, changing the pressure on the violin bow, changing the tension of the lips, releasing a foot pedal, etc.

A synthesizer allows you to change many different parameters of a sound – pitch, range of overtones, mixing ratio of individual sound components, etc. In contrast to “conventional” instruments, this is not only possible manually (lip pressure, etc.), but also automated: Almost every synthesizer has a number of building blocks that can be used to control sound parameters. This shifts the work from lips, hands, and feet to electrical circuits that we have to tune, or “program” (what a terribly “unmusical” word, isn’t it?) to do the work.

Structure of a “normal” synthesizer

Let’s take a concrete look at this using a very simple synthesizer:

Simple audio path in the synthesizer.

The oscillator creates a basic sound, then there is a filter that can change the “color” of this sound, and finally there is an amplifier that is responsible for the volume of the sound. Of course, timbre and volume can change while playing, mostly even within a single note.

However, the oscillator, filter and amplifier do not exist on their own, but are controlled by various other modules: A keyboard that determines the pitch, an envelope curve for the progression of sound and volume with each keystroke, etc.

Here you can already see on a very minimal synthesizer that there are quite a lot of mutual connections:

A minimal synthesizer with oscillator, filter, envelopes and LFO.

There is (at least) one oscillator that produces a basic sound. Usually we can play the pitch of this oscillator with a keyboard. An oscillator often already offers different timbres – for example, a “triangle wave” with very little overtones, a “square wave” with a somewhat hollow sound and a rich “sawtooth wave”. Or noise, which then has no “playable” pitch – noise simply rustles.

The sound from the oscillator can then be further modified with a filter. This is often a “lowpass filter” that attenuates (more or less strongly) all source material above an adjustable cut-off frequency. For example, other types of filters attenuate below a cutoff frequency (highpass filter), or attenuate a narrow range at the cutoff frequency (notch filter), etc. We’ll get to know the different filter types in detail later.

Then it goes to an amplifier, which is responsible for the volume of the sound: Controlled accordingly, this amplifier gives us percussive or soft oscillating, long or short oscillating tones or even a tremolo.

In our simple synthesizer, the filter and amplifier are controlled by a so-called “envelope generator“. It is restarted with each keystroke and then runs through an adjustable transient phase (“attack”) until the sound reaches maximum volume, followed by a decay phase, the duration of which can also be set, until it remains at a “sustain” level (in this case, a medium volume that can be set in advance). When the key is released, the decay phase (“Release”), which can also be adjusted in duration, starts until the sound finally stops. After these four adjustable parameters “Attack” – “Decay” – “Sustain” – “Release” such envelope generators are also often called “ADSR” generators. There are other types of envelope generators besides ADSR, which we will look at separately later.

And because the envelope generator was so practical for the volume curve of the tones, our simple synthesizer gets a second one for the filter. In this way, each tone also has a separately adjustable timbre progression.

Finished? Almost: A very slow oscillator “Low Frequency Oscillator” (LFO) is a pretty useful building block. You can use it, for example, to influence the pitch – and you get a vibrato or a wild howling if you have set the intensity of the modulation very high. By the way, we call the controlling of the parameters of a component by another component “modulating”, just as the envelope generator previously modulated the parameters of the amplifier or the filter. Of course, the LFO can also modulate parameters of other components – the corner frequency of a filter (“wahwah” effect) or the volume of the amplifier (tremolo) etc. If there is a greater need for complex modulations (and a somewhat larger purse), there can of course also be several LFOs, perhaps also those that are themselves voltage-controlled, which would then be so-called VCLFOs, i.e. “Voltage Controlled” LFOs.

The simple synthesizer is already finished with that. Maybe some handwheels, joysticks or ribbons that can be used to manually influence certain parameters. Or even an input for external signals in the filter. Another oscillator or another filter are also welcome. You can make decent music with that.

Analog synthesizers – what is analog here?

We just found out that modulation of parameters is quite useful. In fact, that’s what basically characterizes a synthesizer. With analog synthesizers, the sounds are controlled, generated and modified with the help of analog circuits. The modulation is done simply by generating appropriate voltages, which in turn affect other circuits. And with analog synthesizers, the tonal material also basically consists of voltages: a tone of 1,000 Hz is nothing more than a corresponding alternating voltage with a frequency of 1,000 Hz.

The American physicist and electrical engineer Robert A. Moog built the first prototype of a synthesizer together with Herbert Deutsch in a basement in 1964/65. An article by Moog for the 1964 AES Convention then explained the principle of voltage control for the first time (you can read more about it at: www.moogarchives.com).

Moog is legendary today, in particular because of its filter design (transistor cascade, “Moog filter”) – BUT: In comparison, the principle of voltage control was a much more consequential and influential invention! Modern synthesizers would be unthinkable without it! The ingenious idea is simple: Let’s replace the manual turning of knobs (how many of them can we possibly manage at the same time and with what precision and speed?) with processes that can be automated!

Why modular? – Why not!

Analog? OK, checked. But what distinguishes a modular synthesizer from “normal” analog synthesizers? A modular synthesizer allows great freedom of configuration! This applies to both the number and type of modules (oscillators, sound benders, controlling modules, etc.), as well as the respective connection of the modules to each other, which is usually done with the help of patch cables. Incidentally, even modular systems are no longer necessarily completely “analogue”. There are now many modules that have digital “innards” (e.g. Bitcrusher or digital effects such as Hall & Delay, but also digital oscillators, MIDI interfaces, sequencers with memory functions, etc.). Of course, the controlling of these modules remains analog.

From time to time there are (often quite academic) discussions about “semi-modular” synthesizers. With them, for example, the modules are already defined or certain connections are already wired internally as standard, etc. The transitions are fluid, flexibility decreases, but sometimes that’s an advantage – you don’t have to laboriously wire everything “from scratch” yourself. A prominent example of such a semi-modular system is the ARP 2600 or the Korg MS20. Both are great synthesizers that are extremely flexible, but always have the same – unchangeable – combination of oscillators, filters, etc.

Before deciding for or against a modular system, you should think carefully about your needs in terms of complexity and flexibility of an instrument. Modular synthesizers, even in smaller versions, cost quite a considerable amount of money and it is a pity when you have to realize afterwards that your own way of working is rather opposed to the modular approach and that you then don’t make music with the expensive device.

Another risk of buying a modular synthesizer is getting bogged down in some form of collecting hobby. If you don’t have this or that (exquisite, difficult to deliver and also hardly affordable) module, then you can’t make music at all! Meanwhile you browse websites, forums and test reports for hours, design ever fatter and gigantic systems with online module planners and forget that you actually already have a pretty cool system at home that could also be used to make pretty cool music instead.

Another “trap” with modular synthesizers is a stubborn clinging to measurable but rather irrelevant properties of modules in practice. An example is the octave purity of filters in self-resonance. Many filters generate a sinus wave with high feedback values, which can in principle also be played melodically via the keyboard and the filter’s voltage control. This is basically nice, but you may not need it very often and certainly not with all filters. And it doesn’t work for all filters (for various reasons). If you need it, you will do some research and then select a suitable filter module separately. Otherwise there are much more exciting things that you can do with modern filters. But if you’re obsessive enough, you can make a big deal out of this small (side) feature, in the sense of “what an incredible sloppiness that not even that works, in the past every filter (of the two available at the time) could do it”, and so on. If you’re prone to this type of compulsiveness, a modular system may not be right for you either.

So check in advance whether a modular synthesizer suits your inclinations and needs:

  • You get maximum flexibility in creating sounds, but that flexibility comes with a price tag.
  • The processes of creating both sound and music are closer together than in any other instrument. That sounds great, but it can also require you to make more decisions than you want or are able to make.
  • However, you have to spend more time on this than with a “normal” synthesizer. Both in the concrete creation of a sound and in general when familiarizing yourself with the instrument. And last but not least, also in the run-up to the planning of your system.
  • You need to do a bit of basic physics / electrical engineering stuff. You don’t need too much of this electrical engineering knowledge, but e.g. “voltage” and “resistance” should not be terms that put you off…
  • There are no “factory presets” to conveniently start with.
  • At some point, you have to remove even the most ingenious patch in order to be able to continue working. There is no storage. (I have said this before, did I?) Don’t assume that you can document the patch anywhere near exactly to recreate the sound 1:1 “from the cookbook”. Instead, learn to master your instrument in such a way that you can spontaneously implement your tonal ideas.
  • A complex patched modular system is clearly, and I really mean clearly more confusing than a conventional synthesizer with a well-structured surface. So you also need to remember significantly more to make music with it. (“Where is something patched right now and where does this long green cable lead to? Ah, that’s not plugged in at all on the other side…”)
  • Modular analog synthesizers are usually monophonic. For a long time, polyphony was very expensive (multiple VCOs, VCFs, etc.!) and also quite cumbersome because you had to manually match all modules such as filters and envelopes. A lot has changed here.
  • Expect “surprises” that you no longer know from digital or hard-wired systems, for example: Cables can break or sockets can get bad contact over the years, power supplies can add hum to the audio path, some modules produce noise and other strange background noises, most oscillators should be “preheated” for a while in order to be in tune, some modules distort the input signal from a certain level (this can also be desirable with filters, for example), etc.
  • Not everything that a module could theoretically be able to do (see octave purity of filters) it will also do in practice. Learn to be creative with what you have and don’t cry when something isn’t possible.
  • Before you buy the instrument, you have to think about which functions and options you want and what you can do without (for the time being). This website is intended to provide a little help with that.
  • Beware of buying a big, impressive (who to impress?) monster system unless you fully understand what you’re doing, what the modules can and can’t do, and how it even aligns with your personal approach to music. Large monster systems are practically no longer salable “in one piece” (and very, very laborious module by module). I’ve met musicians who enthusiastically invested huge sums in one go and then spent years selling the useless junk at great loss.
  • Expect your modular synth to get bigger over time than you planned, but beware of becoming a “collector”.

Well, still motivated? Great, then we’ll take a closer look at such a modular system in the next chapter. The Doepfer A-100 System is not the only modular synthesizer on the market, but I don’t know of any system with even remotely comparable variety of high-quality modules. It sounds good and the craftsmanship is more than solid. The basic descriptions of oscillators, filters, etc. can of course be applied to oscillators, filters from other manufacturers, but the tonal properties will differ in individual cases.

The design of the user interface is also often very different between different manufacturers – and should represent a meaningful working environment for you. The individual needs are very different: sockets always on the side? Always at the bottom? Octave Switches for Oscillators? Optical design rather “freaky” or rather conservative / laboratory-like? Blue or red LEDs? Smooth-running or stiff knobs?

Thanks to the standardized sizes and power supply of the modules, there is of course no reason not to put together a system from the offers of very different manufacturers. Actually, most people do so.

Audio signal or control voltage?

One of the great advantages of modular systems like the A-100 is that there is no fundamental difference between audio signals and control voltages. You can easily use the output signal of an oscillator to modulate a filter cutoff frequency, use white noise to modulate the frequency of an oscillator, or use the control voltages of a fast-clocked sequencer to generate an audio signal (“graphic oscillator”). And ring modulators, for example, can be “fed” with audio signals as well as with control voltages from LFOs (or a combination of these).

However, there are a few technical exceptions to this principle. Some modules are more optimized for audio signals, others for much slower control voltages.

Audio only

Only slow control voltages

  • A-163 VC Frequency Divider (can be adapted to audio signals via jumpers on the board).
  • A-129/3 Vocoder Slew Limiter.

On the “taxonomy” of the modules

A taxonomy is an attempt to create and define a classification scheme for a set of “objects”. The best known one is probably the taxonomy in biology, a widely branched tree of living things in which we humans classify ourselves as a species in the genus Homo within the subfamily apes within the family primates within the subclass of higher mammals within the class mammals within the subphylum of vertebrates within the phylum of chordates within the taxon of multicellular animals within the animal kingdom within the domain of eukaryotes.

Despite the now very large number of Doepfer modules, my “taxonomy” is a bit simpler, but some modules are at home in more than one category. I restrict myself to two levels (three for the filters), the number next to the category shows the number of modules currently described here:

Previous chapter:

From a book to a website and why

Next chapter:

Generating tones and noise