However, I hesitate to delete them, because some of the class discussions are still available online and removing the post will break the link. Also, they may be of some interest.
Introduction:
This is my 6th and final homework assignment for Introduction to Music Production at the Berklee College of Music, offered via Coursera.This week's topic was Synthesizers. I've never played with a synthesizer before, so I chose something fairly generic for my assignment:
"Explain the usage of the 5 most important synthesis modules: Oscillator, Filter, Amplifier, Envelope, and LFO. "
Lesson:
Oscillator
The Oscillator is what initially creates the sound. The options -- different wave forms -- differ primarily in the sound spectrum produced. The basic waveforms are:
- A sine wave is a pure tone, with no overtones.
- A sawtooth wave produces many harmonics.
- A Square or Triangle wave has overtones, but only the odd harmonics.
- Noise makes a hiss across all frequencies.
The oscillator differs from the oscillator built into a DAW in that it is modulated. Normally, pitch is what is modulated. When you hit different keys on a keyboard, the oscillator responds by changing the pitch of the sound.
To model a specific instrument, it helps to start with the appropriate waveform. For instance, a flute makes a pretty good "pure tone"; a sine wave. The "stick-slip" of a bow across the strings makes a sawtooth style wave. The opening and closing of a reed instrument is appropriately modeled by the on/off form of a square wave.
These are only the basic waveforms; there are others. The Zebralette synthesizer demonstrated in the class material has 16 different waveforms you can choose from.
Filter
The waveform as it comes out of the oscillator is somewhat useless. It has energy all over the place, even up into frequencies we don't really hear. It's very bright (and probably buzzy or piercing or hissy, as well). To make it begin to sound like an instrument, we need to take out the unwanted frequencies. We do this with a filter.
A Low Pass Filter is probably the most important filter; it removes all the hiss and noise above a designated frequency. You can have any of the other EQ-type filters, though.
But remember, in a synthesizer, all these effects are modulated over time. Changing the filter changes the timbre, changes the nature of the sound.
In Zebralette, and in the other synthesizers I've looked at, there are a whole bunch of "spectral effects" filters. These can be very complex and very confusing. For instance, there is one called "Formanzilla", which "Multiplies the wave spectrum with a variable harmonic, resulting in ‘formant’ sounds with a number of strong peaks and troughs." Zebralette comes with dozens of these presets.
Amplifier
This is more than just gain; it is amplification over time. Basically, when you hit a key on the keyboard, that turns on the amplifier, and when you take your finger off that key, you turn off the amplification. The way the amplification occurs, how fast it happens, what levels it goes to, and how it shuts off, are all characteristics that define the sound of an instrument. We control this by defining an envelope for the sound:
Envelope
The Envelope of the produced sound consists of four parts: Attack Time, Decay Time, Sustain Level, and Release Time. (This picture is from the Zebralette manual, and has an additional parameter, F/R)
When you hit a key to play a note, the oscillator starts a wave form at zero volume which then climbs to the maximum set volume. The amount of time it takes to do this is called the Attack time. It the attack time is zero, then the sound starts with an audible "click". Most natural instruments have a short, but non-zero, attack time.
The Decay time is the amount of time for that initial amplitude to drop to a sustained level. So to discuss decay, we have to first talk about Sustain.
Sustain Level is the amplitude at which a sound stays when you hold the note. If you pluck a string, or hit a drum, the sustain will be negligible. But if you bow a string, or play a flute, the note will play as long as you are bowing or blowing. That is the sustain. A piano doesn't have this sort of sustain; it is a plucked string. An organ, on the other hand, keeps playing as long as the key is pressed.
So back to decay for a moment. When you hit a key, the sound goes from zero to max volume during the attack phase, and then the sound decays -- lowers in volume - until it gets to the sustain volume level. This takes a certain amount of time, which you can define as the Decay time.
Finally, what happens when you take your finger off the key? A "note off" signal is sent, which triggers the sound to drop to zero amplitude, over a certain period of time. The amount of time this takes is called the Release time. An abrupt cessation of sound will make a click; there needs to be some amount of time for it to sound natural. A long release can sound like the sound echoing away.
The Zebralette synthesizer (one of the ones recommended for class) has a setting for shape of the envelope - whether the lines curve or not -- and two additional knobs:
The Fall/Rise knob determines a change at the beginning of the sustain phase, sort of an attack into the sustain level.
There is also a Velocity knob which apparently interacts with keyboard velocity. I do not have a keyboard, so I could not play with that.
LFO
The Low Frequency Oscillator is different from the main oscillator, in that its purpose is not to produce a sound. Instead, it applies a cyclic change to the main signal.
A good example is vibrato. Vibrato is a cyclic change in pitch. The pitch goes up and down, perhaps 6 times per second. Remember that cycles-per-second is another name for hertz. We specify how fast the vibrato should be - a low one might be 3 Hz, a fast one 6 Hz, whatever. In Zebralette, this rate is labeled "Sync" and is connected to the song tempo.
We can also say how sharp and flat to make the note - that would be the amplitude of the LFO. This is labeled "Depth Mod" on the LFO1 module of Zebralette.
We can also specify the waveform of the modulation, which might mimic the different types of vibrato - a smooth vibrato might be best modeled by a sine wave, and even up and down. A jazz vibrato might be better modeled with a sawtooth wave - a quick rise in pitch followed by a more leisurely fall.
An LFO is not just a method of adding vibrato, though. You can use an LFO to run any of the sound parameters through a cycle. You can even run LFO effects through other LFO effects... very complicated.
========
Reflection:
I can't believe we are at the end of the course already. I feel that I am only just beginning to get a handle on my DAW. I've never played with a synthesizer until this week. I found that Louden's EduSynth was invaluable in getting a grasp on what they are about, but it was a bit of a shock to try to transfer that to Zebralette. FreeAlpha seems to be a bit closer to EduSynth, but I had trouble getting it to install properly.Thank you so much for reading this, and for your feedback.
Not only that, I would like to thank all the other students in this course. So many of you have done such fantastic work. It's been a fantastic course, and a fantastic opportunity to see and hear some wonderful efforts. -Claire
