Synthesis

For this week's assignment in the Introduction to Music Production class on Coursera, I'm going to demonstrate the usage of various controls on a synthesizer.  I will be using the Analog synthesizer in Ableton Live.

First is the oscillator - this is the part of the synthesizer that generates the sound.  The oscillator generates sound by producing one of four waveforms: sine, sawtooth, square, or white noise.  In the following video, I demonstrate the first three waveforms (I will use the white noise setting in the next section).  Then, I show how the frequency of the waveform can be modified at three different levels of granularity by turning the Octave, Semitone, and Detune knobs:

Next is the filter - this component shapes the sound coming out of the oscillator.  The four most common filter types are:

• Low-pass: filters out sound above a given frequency
• Band-pass: filters out sound above and below a given band of frequencies
• Notch: filters out sound at a given frequency
• High-pass: filters out sound below a given frequency

In the following video, I play some white noise and filter it using one of these filters at a time.  After selecting each filter type, I sweep the filter frequency to demonstrate the filter's effect.  Then, I demonstrate the resonance control, which uses feedback to add a peak in the spectrum at the filter's frequency:

Finally, we have the amplifier - this component modulates the amplitude of the sound.  The amplifier applies an envelope to the sound that has four attributes:

• Attack Time: the amount of time it takes for the sound to reach its peak amplitude after a note is played
• Decay Time: the amount of time it takes for the sound to reach its sustain level after the attack phase
• Sustain Level: the level at which the amplitude will remain from the end of the decay phase to the release of the note
• Release Time: the amount of time it takes for the amplitude to reach zero after a note is released

In the following video, I start by increasing the attack time in order to make the sound fade in slowly when a note is played.  Then, I decrease the sustain level to make the sound more percussive, and I demonstrate how the decay time affects the transition from attack to sustain.  Finally, I play around with the release time, initially making the sound end abruptly when a note is released, and then making the sound fade out slowly when a note is released:

Thanks for watching!

The Delay Spectrum

For this week's assignment in the Introduction to Music Production class on Coursera, I'm going to demonstrate the effect of short delays on the frequency spectrum of a sound.  I will also demonstrate how to effectively use medium and long delays to enhance a piece of music.

When a sound is combined with a slightly delayed copy of itself, the interaction between the waveforms creates an interference pattern that results in notches in the output signal.  This effect is known as comb filtering.  In the following video, I start by playing an unprocessed white noise signal.  Then, I turn on a delay set to one millisecond with a dry/wet mix of 50%.  The result is a series of notches in the output.  Finally, I slowly increase the delay time to two, three, four, and five milliseconds.  With each increase in delay time, more notches appear, and the notches are closer together:

The feedback parameter in a delay effect is used to send the output of the delay back into the effect, causing the sound to repeat.  When this is done in a short delay effect, the relatively high frequency of the repeated sound is perceived as a pitch.  In the following video, I start by playing an unprocessed drum sound.  Then, I apply a one millisecond delay and progressively increase the delay time by one millisecond at a time.  Each increase in delay time causes more overtones to appear, and the perceived pitch drops:

A medium-length delay can be used to achieve a slapback effect, which simulates what happens when a sound bounces off of a hard surface.  In the next video, I start with an unprocessed drum track, and I apply an 80 millisecond delay.  Initially, the dry/wet mix is set to zero, so I increase the amount of wet signal until the effect adds a sense of space to the drum kit.  Then, I slightly increase the delay time in the right channel in order to emulate being a different distance from two walls in the space:

A long delay can be used to add rhythmic interest to a melodic line.  In the final demonstration, I play the opening guitar part from The Electric Co. by U2 (I apologize for the lame guitar patch - I didn't have time to set up my real guitar).  Without the delay, it sounds dry and uninteresting.  I then turn on a 3/16-note delay that is synchronized to the tempo of the song.  Feedback is set to 50%, and a filter is applied to roll off the low end of the wet signal, differentiating the echoes from the original notes:

Thanks for watching!

Compression

For this week's assignment in the Introduction to Music Production class on Coursera, I'm going to show you how to apply compression to an audio track in a Digital Audio Workstation (DAW).

I'll demonstrate compression with a drum track.  First, while playing the track and observing its waveform, I decrease the threshold until it is below the peaks I woud like to compress.  This threshold determines the amplitude at which the compressor is activated.  Then, I increase the compression ratio - this determines the amount of compression that is applied.  Initially, I set it too high, and the drum track loses all of its dynamics.  I then dial it back until I achieve the desired balance between all of the drum kit elements in the track.  Note how the display shows the input sound as a faint waveform in the background, and the output sound as the waveform in the foreground:

The compressor has removed the transients from the beginning of the kick and snare drum hits, and I would like to bring them back.  To do this, I adjust the attack time - this determines how quickly the compressor is activated after the input rises above the threshold.  I turn the attack knob to the right, increasing the attack time until I can hear enough of the transients - this restores some of the punchiness to the track.  Note that the first display (gain reduction) shows a slight change in the speed at which the compressor reduces gain (see the yellow line at the top of the display.  The second display now shows that the output contains those transients, which rise above the compressor threshold:

Finally, I want to set the release time of the compressor - this determines how quickly the compressor is deactivated after the input falls below the threshold.  Initially, I set the release time too high and the compressor releases too slowly, flattening the dynamics of the track.  I then bring it back down to a speed that achieves smooth transitions and avoids the pumping effect that is caused by very short release times:

There is no one-size-fits-all recipe for applying compression - every project will present different requirements, so the desired results can only be achieved through experimentation.

Thanks for watching!

Recording Automation

For this week's assignment in the Introduction to Music Production class on Coursera, I'm going to show you how to record automation in a Digital Audio Workstation (DAW).

I have recorded a bass track, and I would like to add a resonant filter to it that varies in rhythm with the bass line.  Here's the original track:

Next, I add the filter effect to the bass track, arm the track for recording, and set the record mode to Overdub so that I don't erase the bass line when I add the automation data.  Now all I have to do is manipulate the filter effect while the bass line plays.  Here's the next take - watch as I move the filter frequency around in the bottom right part of the screen:

Alternatively, you can draw the envelope of the effect onto the arrangement.  To do this, I set my DAW to Drawing Mode and select the Frequency parameter of my filter effect for editing.  I also set the grid to 1/32 so that I can draw a smooth curve.  Here's a take where I draw the curve and play back the resulting track:

In recording this lesson, I found it awkward to use the drawing tool to draw such a complex curve.  I would recommend using it in simpler situations, such as when you need to fade a track.

Thanks for watching!

Recording and Quantizing MIDI Instruments

For this week's assignment in the Introduction to Music Production class on Coursera, I'm going to show you how to record and quantize a MIDI track in a DAW (Digital Audio Workstation).

To record the track, follow these steps:

1. Select an instrument and assign it to a track.  Often, this involves dragging an instrument from an instrument browser onto a track in the arrangement view.
2. Arm the track for recording - this prepares the track to receive and record MIDI data when recording starts.  Some DAWs automatically arm the track when you assign an instrument to it.
3. Set the tempo, turn on the metronome, and set the metronome count-in to at least one bar.
4. Press the Record button, wait for the count-in, and start playing.

The following video demonstrates these steps using Ableton Live.  I chose to try to record an entire drum kit in one take, in the hopes that the performance would be in need of some quantization.

As expected, the performance was a bit sloppy.  To quantize the track, follow these steps:

1. Select the clip in the arrangement view.
2. If you plan to quantize to the grid, make sure that the grid is set to the correct spacing.
3. Select all of the notes to be quantized.
4. Set the quantization amount to a low number (such as 20%) and perform the quantization.
5. Listen to the quantized track.  If it still sounds too sloppy, perform quantization again.
6. Repeat step 5 until you like how the track sounds.

The following video demonstrates quantization in Ableton Live:

My hard-learned lesson this week is that screen capture software (I'm using Camtasia) can have a significant effect on the performance of the application you're recording.  I was trying to record the track using my computer's keyboard, and Camtasia introduced enough latency to make that impossible. Switching to an external MIDI keyboard solved the problem.

I hope this was helpful - thanks for watching.