Artwork by Køps

Audio Production Guide

Basics, Concepts, Techniques, Resources, Philosophical Aspects

Introduction
The path of learning how to express yourself with music can be very cumbersome and taking on
this effort by yourself is a daunting experience.

Whether you’re a beginner or already a bit deeper in the game, there are always challenges,
doubts and questions gnawing at you. The realization that the mental approach is a huge part of
the process - aside from creative and technical aspects - led me to compile this knowledge base
to hopefully inspire and ease the path for aspiring producers.

On top, seeing a seasoned producer like Phutureprimitive invoke the same concepts I have been
fighting to define in my two years of production is very liberating - a feeling that should be
accessible to anyone.

This is not a 100% scientific collection and is mainly supposed to help your thought process and
quantify your thoughts. This guide summarizes my thinking and metaphors that aided me in
tackling the path to a creative workflow.

Noteworthy keywords, often worth a google search to read up on, are highlighted.

Contents
● Basics Of Sound
● Colour And Timbre
● Concepts
● Philosophical Remarks
● Techniques
● Writer's Block And Frustrations
● Resources
● Final Words

2
Basics Of Sound
What is “Sound”?
First off, let’s start by giving you a way to think about sound and how to deconstruct the things
you hear. For this, we need some basic understanding of how a sound is produced and how we
can alter it. Our tool to see these things is called a Spectrum:

Fig. 1: Snippet of a Saw Wave’s Spectrum in Ableton.

Each of these bumps is a harmonic of the root note being played, increasing in pitch along the
X-axis and amplitude (volume) on the Y-axis. The spectrum is a momentary picture of our
signals frequency composition. The most simple signals are created via oscillation of waves
like sine, saw, triangle and square. The sine wave is the most basic waveform and addition of
several sine waves yields the more complex ones. In theory, any sound can be built with this
from scratch and - in reverse - deconstructed.

Fig. 2: The 4 basic waveforms.

3
As examples from the real world we can imagine didgeridoos, string instruments like violins or
brass as saw-wavy in character as opposed to birds that come off as softer and more as a sine.
Of course there’s examples inbetween.

How these spectrums come together can be extensively explained and has been done so in
many resources. However this would take too much space here. If you’re interested in this,
please refer to oscillators, spectrums of waves, Fast Fourier Transformation (FFT) and basic
waveforms.

A FFT transforms a signal from the time axis (e.g. waveform of a full track with time as X-axis)
to the frequency domain. This means a spectrum is a frame of the frequency composition of the
track at a given point:

Fig. 03: Waveform of a full track (top) and its spectrum at a point in time (bottom). The X-axis in the top shows the
time domain, in the bottom it refers to the frequency of notes. The Y-axis depicts the amplitude in both cases.

4
As we can see in both spectra so far, higher frequencies are generally quieter than lower ones in
finished tracks. Why is that? If we were to have a bunch of frequencies playing at the same time
and the same volume, we would deal with an unpleasant noise. In fact, this kind of sound is
called White Noise.

Volume and Loudness

As humans, our ear functions in a way that we are more susceptible to mid and high frequency
sounds, thus having an easier time listening to them. Of course it makes sense that the
speaking range of frequencies (mostly 800-4000 Hz) is what we are naturally calibrated for.

This means that if we listen to a high pitched sound in the same nominal volume (in dezibel or
dB) as a very low pitched sound, the perceived volume of the high pitched sound would be
much higher. A depiction of this phenomenon is the Fletcher Munson Curve or equal loudness
contour.

Fig. 04: Fletcher Munson Curve (left) and frequency composition of white noise (middle) as well as pink noise (right).
X-axis = Frequency, Y-Axis = Amplitude

This shows the nominal volume needed across all frequencies to achieve the same perceived
volume or loudness. An example of this is Pink Noise. This kind of rather pleasant noise can be
found in nature as rain or when frying bacon.

You will notice that this basic volume to frequency ratio is found in most well produced and
mastered tracks, revealing the path to balance and clarity in your own music.

5
Mono, Stereo and the 3D-Room

Everyone heard the terms mono and stereo, but what does it actually mean? It is quite simple:
any sound that has the exact same signal coming out of both speakers is called mono. Any
difference between the left and right signals automatically makes it stereo by definition. This
can happen via phasing - an offset of waves which for example is introduced by every standard
EQ.

Most people think of something being stereo when it seems very wide and spaced out
throughout the sonic room that is created by a two speaker setup.

Indeed, sounds that are perceived as wider and more diffuse can be traced back to having a
signal that is “more stereo”. This also means - in contrast - that mono sounds are very upfront
and right in your face. A helpful tool later on when you want to give something more presence.

This can be achieved by having a mono layer on top of stereo layers or by

Mid/Side-Equalization, manipulating the balance between mono and stereo signal. Forcing a
signal to mono not only helps with the punch (e.g. on kicks) but also reduces (or reveals)
phasing problems.

Note: Phasing issues (leading to cancellation of waves) are a common reason why the
sub region can lack power by losing volume. To avoid this, the sub is mostly forced down
to a mono signal.

Moving away from a strict mono track to the stereo domain opens up the concept of a
one-dimensional picture towards a 2D canvas, with panning between left and right channels as
X-axis and frequency / pitch as Y-axis.

Note: This concept is not to be confused with an actual X-Y graph like the ones we saw
for the spectrum.

If we add depth, we are able to envision a 3D space in which a sound

can reside.

This Z-axis can be controlled via volume and reverb. All this comes in
handy when trying to craft an experience of your own.

6
This way we can decorate our room - after all it makes a difference if you stack your couch on
top of your bed or alongside it, right?

Looking at the stage setup for a live band reveals that people have cleverly used this concept for
a very long time.

Here we see that the Drum Kit is placed in the center, usually with two guitars to the left and
right. The lead vocalist is in front of the Drum Kit, also placed in the center but representing
higher frequency content than e.g. the kick. The bass speakers are often slightly off center.

Fig. 05: Common stage setup for a live band.

Note: In EDM music we try to keep the sub bass mono for maximum clarity even on club
systems. Logically, the low end of the kick and the sub cannot be stacked on top of each
other without possibly detrimental phasing.

Thus the sub usually ducks away to make space when the kick plays. More on this later.
Higher bass layers (around 100-650 Hz) can take space in the stereo spectrum
depending on context.

Now we know how to furnish our 3D room without our main elements falling over each other.
Setting these priorities tremendously facilitates the later mixing process and helps have a
cleaner signal from the start, esp. when paired with removal of undesired frequencies.

7
You might think to yourself “But I’ve heard things move across the stereo field in a track, all of
the above is static!” and you would be right for both statements. In a compelling track there will
almost always be statically placed elements as cornerstones and mobile earcandy moving left
and right (via auto-panning), back-to-front (via reverb and volume) or top to bottom (pitch
shifting).

Often overlooked but just as important are hi hats: The hi hats on a real life drum kit are spread
out from the center (= the kick) - again in a static fashion. Each hi hat will then come from the
same spot in our room every time it is played.

Taking advantage of this along with velocity will breathe life into your drum tracks very easily!

Fig. 06: Example of hi hats being placed differently in the stereo field.

8
Colour, Timbre And Dynamics
Great, so we now know what waves and more complex sounds consist of and how to place
them in an imaginary 3D room. But how does a sound get its character? How is something
further away, swiveling around our ears or pumping its way through the speakers?

The answer to all those questions is based in the dynamics of a sound. In the digital world,
things are either on or off. The natural or human element is in how the sound ramps up or dies
down, how it is articulated. The human approach is expressive.

ADSR-Envelopes
Generally, a short impact of sound is called the transient and it can be anything from the click of
a kick, the plucking of a guitar string or the impact of a snare. The timeframe of a transient is
usually about 30 ms. After this comes the body, the meat of the sound before it dissipates to
silence with its tail. We can see this in audio files but we can also control these parameters
directly in synthesizers with the ADSR-Envelopes.

A = Attack, D = Decay, S = Sustain, R = Release.

Fig. 07: ADSR-Envelopes for three common types of instruments: Pad (left), Lead (middle) and Pluck (right).

To the left we can see a basic envelope for a pad. The volume slowly ramps up to its maximum
value and dies down smoothly when playback is stopped. This greatly influences the feeling of
suspense inherited. Attack times are also used to cause a pumping sound, esp. when synched
to track timing. We will see this later when we look at dynamics.

Note: An easy way to enforce this pumping is by using a sidechain compressor, ducking
the signal based on an external input, e.g. the kick.

9
Leads are often very upfront, with short attack and release times and fully sustained volume
when holding a note, whereas pads and chords often have a lowered sustain. As soon as we
stop holding down the note, the sound usually stops immediately owing to a short release time.

In opposition to pads, Plucks are very short bursts with a fast attack and a quick release. As the
name suggests this is a clicky, plucky transient sound which works well for playing arpeggios.

Note: In principle, any kind of percussion is also a pluck, e.g. hi hats. A cymbal or ride
follows the same theme, albeit with a longer release time.

It is important to know that attack and release times that are too short can cause an unpleasant
clicking. If you encounter this, try tweaking these values.

Movement And Modulation

In addition to the volume over time, we can change the pitch or impose effects on our sounds. In
fact, a kick is exactly this: a short pulse of sound that decays in volume and pitch
simultaneously, similar to the pluck curve in our picture above.

Depending on how quick this pitch envelope moves we get a change from a clicky to a more
laser-like sound. Funnily enough, a bird, laser and a kick are in principle the same sound - a
quickly decaying pulse that also moves in pitch - either up or down.

You can also test this by yourself with a synth - simply map the pitch to decay like the oscillators
release time does. Usually this option is called coarse pitch (Serum) or pitch envelope
(Ableton’s Operator).

Higher pitched waves exhibit a shorter wavelength while low pitched waves are longer in
wavelength. The measure for this is Hertz (Hz). As an example, our western music is tuned to
A4 = 440 Hz. Doubling the frequency goes up one octave (A5 = 880 Hz) while halving it goes
down one octave (A3 = 220 Hz).

10
 Fig. 08: Comparison of the waveforms for two different kicks, consisting of transient, body and tail.

The picture above shows the waveform of two different kicks. In both cases we can see the
transient to the left before we transform through the body to our tail, dissipating into silence.

With what we know by now, we can derive that kick #1 has more click to it since the transient
has more higher frequency (= shorter wavelength) content, as can be seen by the more crowded
starting point. The body of this kick has some mid frequency content to it and smoothly unfolds
towards the sub region (= longer wavelength). This means we have a longer lifetime of the pitch
envelope, evenly going from high to low.

Kick #2 has a lot less high frequency content in the click and basically skips the existence of a
mid frequency body - showing longer, low frequency waves right from the start. This inevitably
means there is little to no pitch envelope and if there was one, it was very quick. We can also
note that the tail looks a lot less like a sine wave and more like a square wave. We will learn later
that this is a direct consequence of distortion and deliberate clipping of our wave. With this,
Kick #2 will sound “more distorted” overall. The small dent in the beginning of the wave also is a
consequence of said distortion.

11
 Note: While we can synthesize a kick with this pitch envelope technique, we can also
additively build it with audio samples, putting a short click as the transient, a thumpy
body and a sub tail together, EQing each. The same principle also applies to Snares.

Generating movement can be done in a million ways but the principle is always the same and is
called modulation:

Changing parameters of a synth or audio effect over time, either by automated Low Frequency
Oscillation (LFO) or manual automation. LFOs follow the same basic waveshapes we already
got to know: sine, saw, square and triangle.

Furthermore we can select a sample & hold (S&H) function, randomly jumping in given intervals
rather than a curve.

Thinking of our 3D space we can manipulate the pitch (upwards movement), the panning
(left/right) and the volume (back/front).

This way we breathe life into otherwise stale bleeps. Combining the basic movement with audio
effects (FX) like delays, filters or distortion further increases character.

Fig. 09: LFO (top left), lowpass filter (top right) and manual automation of this filter (bottom).

Step by step we can impose a tone or timbre. No matter how fancy a FX module seems to be, it
all boils down to three core functionalities: time-based FX, filter-based FX and dynamic
processors.

12
Time-based FX
These audio effects alter the timing of an input signal, thus shaping the depth and dimension. In
addition they are able to stereoize a signal, e.g. by delaying left and right channels differently. As
we learned before, this difference in signal relative to each other is the definition of a stereo
sound.

Note: The amount of FX mixed in with the original (dry) signal is referred to as wetness.
This ratio is balanced with a dry/wet knob.

Common FX of this category are delays, echoes and chorus, reverbs, flangers and phasers,
pitch shifters and harmonizers like resonators or Ableton’s Corpus.

While they all alter a signal over time, the way they do it can vary - thus giving each FX a unique
character.

Delays and echoes are quite self explanatory and do what their name suggests: delaying a
signal in time and spacing it out in the process. A chorus is a very short delay - often times two -
of about 1 to 20 ms that is also filtered and modulated.

Our ear is very sensitive to differences in timing and will start to discern distinct signals at a
delay of about 20-30 ms. Before this threshold of audibility, the delays mix with the dry signal to
form a sort of metallic chorus sound.

Reverberation is the secondary reflection of sounds in their environment and is vital to creating
a lively feel in digital processing. We hear reverb every day and listening to a completely dry
signal can be very uncomfortable. Together with delays they are the most powerful mixing tool
to create a room for your track to reside in.

Note: In the real world, reflections of high pitched sounds dissipate quicker than lower
ones. This means low frequencies travel further. In consequence, reducing the high
frequency content (e.g. via lowpass, hi-shelf) on a signal or reverb makes it sound further
away. A great example of how pitch, volume and delay affect our perception!

13
The size of the virtual room is furthermore controlled by the predelay - the time before the early
reflections first slap back off the wall - and the diffusion of the later reflections.

Flangers and Phasers function by duplicating a signal and delaying one of them in time, causing
phasing that works as a comb filter. They are usually modulated by an LFO to sweep through
the frequency spectrum and accentuate parts of the signal over time.

Note: Flangers and Phasers are also great mixing tools on a group bus. In this case, a
small amount of wet signal is introduced to enhance the stereo image of a collection of
instruments (also referred to as stem), e.g. all melodic elements. See Song Sketches
And Groups for more information.

Another noteworthy mention are resonators, emulating the plucking (e.g. with a mallet, a string
or a tube) and resonance of an instruments body. They are a great tool to retroactively impose a
more melodic and tonal character on an otherwise atonal sample. My personal favourite is
Ableton’s Corpus which can be controlled by external MIDI input like the bass notes (also
referred to as sidechaining).

Filter-FX
Filter effects are tools that can be used to manipulate and alter the frequency content of an
audio signal, changing the tonal colour in the process.

The basic types are low- and highpass, shelves, bandpasses and notches. These can be static
like a standard EQ or modulated to create movement in a signal. Important parameters include
the slope (how strictly sounds are cut), resonance (attenuation at the cutoff frequency) and
bandwidth. Some advanced filters can morph between these modes.

Note: Due to how EQs / filters work they will always introduce phasing into your signal,
unless a linear phase EQ is used. These are special use cases and have other drawbacks
such as introducing pre-ringing.

14
 Fig. 10: Basic types of filter FX: low- & highpass (top), notch & bandpass (middle) and low- & high shelf (bottom.

When EQing a channel, it is oftentimes advisable to work subtractively rather than additively
(surgical EQing). This means cutting undesirable frequencies rather than boosting the ones we
want. A common case is removing the sub of non-bass instruments with a highpass, or small
bandpass dents in the mid frequency region to remove resonances.

Fig. 11: Common EQ shape for a non sub instrument. A highpass smoothly reduces frequencies below 100 Hz, two
surgical bandpass dips reduce resonant frequencies and a broad bandpass emphasizes the mid/high region. Very
high frequencies are slightly tamed with a high shelf.

15
 Fig. 12: Settings of the EQ depicted in Fig. 11.

This way we can bring out the character of our signal or carve out space in the overall mix.
Exemplary for this would be cutting a bit of the main region of our lead from our atmosphere
channel or cutting above ~10k Hz to create more space for the hi hats and earcandy.

Note: The cutting of undesired frequencies also applies to high frequency content.
However we have a hard time noticing the presence of a signal when it is stripped of its
higher harmonics. Thus it is advisable to apply a hi-shelf rather than a hi cut (= lowpass)
and simply dip this area rather than removing it completely in a lot of cases.

Dynamics and Dynamic Processors

In the context of music, dynamics not only refer to the liveliness of a signal. More importantly, in
audio production the difference between the quietest and loudest signal is referred to as
dynamic range.

Dynamic processors are tools to alter and control these differences based on frequency content
and amplitude level, creating a more balanced mix and controlling volume spikes. They are very
sensitive and hard to master as you’ll have to learn to hear the variations between a dry and wet
signal. The most common dynamics processors are compressors, limiters, expanders and
gates.

The average loudness of a signal is called the root mean square (RMS). The peak level of a
channel is the maximum volume reached throughout, independent of the average signal. We can
imagine having a lengthy audio file that is full of silence but has one very loud transient at -1 dB.
In this case our peak level would be -1 dB while the RMS is very low, averaging in the extended
silence.

16
This will be handy when learning to control compressors and deciding whether we want the
average signal to be processed or simply tame the loud peaks that take away our headroom -
the safety space we leave to avoid deterioration of our signal or damaging of hardware.

There’s lengthy discussions on the topic of analog vs. digital and their respective restrictions.
One of the key points is that a digital signal exceeding its point of representation (0 dB for digital
signal) will be chopped off, which is called clipping.

Fig. 13: Effect of distortion and clipping on a sine wave. Top: unprocessed sine - Middle: clipped sine - Bottom:
heavily distorted sine.

As we can see in the image above, clipping the signal alters the waveforms, introducing higher
pitched artifacts (or clicks) and squarifying them. The 3rd is an extreme example of distortion of
the same wave.

17
To understand why a cutting of the wave introduces artifacts we can look at how waveform
addition works. For simplicity’s sake we will focus on the fact that higher frequency waves have
shorter wavelengths than lower frequency signals. The upper frequencies are added to the lower
fundamental frequencies. The resulting waveform we observe is a rippled fundamental, caused
by the overtones.

Note: This is also how saw, square, triangle and more complex waves are formed. This
additive synthesis is not to be confused with Frequency Modulation Synthesis (FM), one
of the basic types of synthesis alongside subtractive and wavetable synthesis as well as
grain synthesis.
In FM synthesis the fundamental as carrier wave is modulated by the 2nd wave acting
as LFO. Please refer to more extensive resources with these keywords if you’re
interested in those topics.

Fig. 14: Comparison of waveform and spectrum generated by additive or FM synthesis..

18
Dents and nudges in a waveform are obviously more complex than the simple addition of an
overtone and come in all shapes and sizes. This basic knowledge will be beneficial for our use
of dynamic processors.

Compressors reduce the volume of louder sounds that exceed a given threshold and / or
amplify quieter sounds, bringing upper and lower ceiling closer together. This “squeezes'' our
signal together, reducing dynamic range and lending the name compression. Consequentially,
the RMS increases while the nominal volume remains similar to before. With this, we are able to
make things seem louder without actually cranking their volume or gain.

Fig. 15: Ableton’s compressor unit.

The effect on our ears is a feeling of a more polished signal that seems more lively. It helps
bring out detail and give a more natural feeling to any kind of signal (e.g. Percussion group or a
singular instrument).

Note: Compressors are often mixed in (= used in parallel) to the dry signal for a more
transparent and natural result. This allows us to squeeze the signal harder and then layer
them together. An example of this is the New York Compression which can also be
combined with an EQ.

A compressor works with an attack and release time, analogous to ADSR-Envelopes for synths
allowing us to fine tune the response to our input signal to prevent (or enforce) pumping.

Used less as a control tool but rather a creative one, Side Chain Compression is a special case.
Here we use a compressor with short attack times that reacts to external input from another
instrument, ducking our side-chained signal whenever the input signal pushes it out of the way.
A great tool to help kick and bass coexist!
19
The ratio determines how much a signal will be pushed down when it exceeds the threshold,
with a ratio of 2:1 meaning that a signal that reaches 8 dB above the threshold will be squished
to 4 dB above. Lastly, we can tell the compressor to watch for peaks or RMS levels. Makeup
Gain can be used to bring the nominal volume back up to its former value.

Fig. 16: Visualization of the ratio’s influence on gain reduction.

Limiters are a different type of compressors, designed to have an extremely short attack and
release time to transparently limit the signal to the threshold. Also referred to as brickwall
compressors, they are used to increase perceived loudness and catch stray peaks, acting as a
safeguard against clipping.

Note: Technically speaking the limiter also clips the signal, albeit in a transparent way. If
the signal were to clip due to hardware restrictions rather than our limiter we would
experience unwanted clicks and artifacts.

Expanders are the opposite of a compressor, reducing signal below a lower threshold or
increasing signal above an upper threshold. This effectively expands the dynamic range and
explains the term for this kind of gear. We can also use them to fix noisy recordings.

If we lower the signal below a threshold far enough we end up with silence - this is referred to as
Noise Gate. It can be used to only let signal through that exceeds the threshold and cut off
anything below. A common use can be cutting off reverb tails of drum recordings though I would
recommend a transient shaper for this purpose.

Other interesting topics in this regard are multiband compression (like OTT), combining a
3-band EQ with a compressor as well as upward and downward compression / expansion.

20
Different gear also has a different tonal character, allowing to introduce subtle coloration with
different compressors.

Distortion
How else can we prevent unwanted clipping or better yet use it to our advantage? While the
deterioration of signals can be an undesired side effect it is also deliberately used to give a
sound more grit and character. The process of forcefully clipping a signal is called distortion.

Here, a signal is pushed in gain towards a threshold. Anything exceeding it will be clipped,
introducing artifacts based on the distortion algorithm. This also means the noise floor will be
pushed, while the maximum volume will remain the same.

Fig. 17: Examples of distortion units in Ableton 10.

Different distortions emulate varying types of analog gear like vacuum tubes, transistors or
digital circuits. The audible effects of this are a distinctive warmth and bite being added. It can
be described as adding upper harmonics to a sound which also explains the observation that
distortion of a lower frequency signal leads to a more pleasing result. Another neat example of
very pleasant distortion is a Saturator.

However, driving a signal too much basically brings up all harmonics to the same level (lifted in
gain, clipped at threshold), creating plain noise so we have to be careful not to overdo it. This is
in direct relation to the Fletcher Munson Curve and how we perceive sound as mentioned
earlier. Learning to properly control distortion is very tricky and takes a lot of experience!

21
Concepts
Music - A fractalized collection of variables
With all this knowledge in hand we can lean back a bit, start looking at the bigger picture and
how we can mentally approach even remotely handling it. This part is of a bit of a philosophical
nature but if you are anything like me it will be of a great help.

I like to think of making music as a system with a vast amount of variables. It is a fractalized
process where we can indulge in as much or as little detail we want.

Like a microscope, we can zoom in and out to view singular samples, phrases , a whole track or
the countless options in just one plugin. We can focus on percussion or bass, texture, ambience
and atmosphere. On top there is (chord) progression, tension and release, call and response.

How many layers are playing overall, is the reverb on this instrument mixed in (called parallel
processing) or part of the original synth sound?

Fixpoints and Choice Paralysis

When a human is given too many choices, the critical thinking part shuts down and the ability to
make decisions is hindered. To be able to possibly muster all the fascinating input thrown at us
when constructing and deconstructing music we have to limit the variables available to us in
order to not be struck with choice paralysis.

22
This means looking for fixpoints, giving ourselves a way to freeze some variables in time to
create meaning and context. This can mean preparing drum racks, MIDI patterns, chord
progressions or full song sketches - mostly with the help of reference tracks to learn from the
best - the artists we look up to.

Song Sketches And Groups

Below we see a very basic deconstruction of a track I recently started to analyze to get an
overview of what it might look like.

Fig. 18: Rough deconstruction sketch of a track. The original file is on top, separated into phrases / points of interest.
Below are different dummy channels used to mark where the respective elements play.

We can copy the drum patterns with dummy notes, find the note progression of the bass, where
crashes are used to emphasize a new downbeat or mark the length of elements with empty MIDI
clips. This way we train your analytical and decision making skills and are able to nail down
exactly how many elements are playing or how they are spread out.

Note: There is no need to copy the actual sounds to write down the pattern we hear. If
we desire we can try and find similar sounds (e.g. via synth presets or deconstruction)
later once we put the idea down.
For the deconstruction try thinking about how a sound was made. Is it a quick, delayed
pluck or a pad with filter movement? Where is it in the stereo field? Is the snare noisy or
an acoustic clap? All these things also play a huge role in the overall vibe of a track!

23
It is important to know that grouping multiple audio tracks based on their function is a
tremendous tool to see the bigger picture both in our own productions and when deconstructing
finished tracks. Once we manage to focus just on the percussion or ignore other groups to listen
to the atmosphere, we will have a much easier time to pinpoint your points of interest.

Note: When mixing finished tracks, artists often use stems which are basically all the
groups of a track bounced (= printed) to their respective audio track and put in a new
project.
Not only does it save CPU but it also creates another fixpoint, taking away the possibility
to endlessly tweak 1000s of FX and pushing the track towards a finished work of art.
These groups can further be merged with group busses (e.g. Drum Buss, Mono/Stereo
Buss) to apply specific processing to related elements.

In the example below we see the sub notes of the reference written down and a vocal group with
three tracks, referring to spoken words and different vocal chops. There is no audio information,
simply a different length of clips to mark where and for how long these things play, plus a short
description in the clip name.

Fig. 19: Closer look at the deconstruction of the track shown in Fig. 18. There is no audio information, only notes and
structure have been marked.

24
Visualizing this gives us an easier time to make sense of what is happening and paves way to
notice conventions. The DJs among us probably have done this via their DJ software without
even noticing. Doing it in a production mindset creates a fixpoint we can refer to. Over time you
will learn what you need to do to unveil how an artist creates a compelling progression while still
portraying a formidable patience throughout their track.

Obviously our own ideas remain fluid and diffuse for the most part while creating a track. To
give you a picture of how it might look after cleaning up and ordering everything I attached two
pictures of finished tracks of mine (titled Nemesis and Versus) below.

Fig. 20: Visual representation of two finished tracks.

In the second picture, Tracks 61 and 62 are references, namely an older mix of the project and a
mastered track.

25
“The Rules” And Perspective
Know this: while there are “rules” to music and I encourage you to find them there are also no
hard rules which means as soon as we discover and quantify them we can start breaking them.

The idea here is challenging yourself to find out what other people commonly do, investigate
different ideas and concepts you think are used and then go ahead and see if it actually works
for YOU.

While a certain technique might work for someone else it might be completely illogical to you or
your working process. However if you put in the work anyway you will give yourself the benefit of
perspective, thus carving your decision making skills. It is OK to disagree with someone and
think “Why on earth would anyone do this?”

Again, this is a fractalized process so we could be looking at how people process their kick, how
they build atmosphere or see a track as a whole - among others. Ideas come in all shapes and
sizes and can be a very abstract thing. Investigating them from your point of view will enable
you to create your own inherent rules, forging a signature part of your very own style.

Juxtaposition And The Space Between The Notes

When making music we paint a picture on silence. What does it tell us? Well, as much as actual
sound is needed to build a track it cannot exist without the space between the notes. Like
shadow and light, both sound and silence are coexisting.

Contrast is what keeps things interesting and engaging and we

can use this to our advantage when trying to create tension,
suspense or release. A great example on how to think of this: if
everything is loud, nothing is loud. And if everything is wide,
nothing is wide.

However by juxtaposing loud and silent elements, wide and

focused layers or low and high frequencies we actively create a
3D room where everything has its carved out spot, working with and towards each other. The
same goes for risers or up / downbeats.

26
A break in a track is mostly just this: turning off some of the elements from before while
introducing something new that converges into the new phrase, e.g. by switching off the
percussion, changing the melody, filtering or automating the width to 0% so the new downbeat
feels wider when it jumps back to full width.

If we make a habit of having a group structure we have access to the possibility of these ideas
with a mouse click by muting them. Never be afraid to throw something out or turn it off to give
yourself space to think outside the current pattern!

Note: The more space between notes, the bigger the effect of pads, atmospheres, field
recordings, reverbs and delays (time-based FX!) So if something feels empty, try slapping
those on it while possibly filtering some of the harsher elements to emphasize the mood.

Sample Choice And Noise Floor

One of the best pieces of advice I received (credit to Pluvio) - specifically for drum work - was
that sample choice is extremely important. The higher the production quality of a sound we use
- be it our own or from a sample pack - the less fragile it is when doing post processing. The
reason being that lower quality samples have more involuntary imperfections which will
inevitably be emphasized, for example by compression (which as we know now brings out
details in sounds while controlling peaks!).

What makes a “bad” sample and how can we circumvent this pitfall? As I said it comes down to
involuntary imperfections. These add up all over the frequency field to form a noise floor,
meandering below the main idea of a sound we created. An example would be unadepted use of
distortion, adding unwanted noise that is now part of our signal.

While we might have tried using distortion to re-emphasize high frequency content we filtered
out with an EQ befor, it could have been too much, thus tampering with the overall balance of
harmonics. Any FX following will then magnify this mistake, especially filters. For this very
reason, the proper use of distortion is a tricky thing to get a hold of and takes a lot of trial and
error.

The reese bass commonly used in Dubstep, DnB and Garage is a very good reference to
elaborate on this concept: The basis consists of two slightly detuned saw-waves which naturally
have a high amount of overtones. A combination of filtering and distortion, working in

27
conjunction, is then used to control the sound and impose movement via automation. Starting
with a lowpass to tame the high frequencies and adding some (time-based) FX after gives us a
good basic signal.

The detuned waves will phase cancel each other in the low frequency region, further causing
movement of the sound. By using distortion and compression we can now emphasize the loss
of frequency content caused by the filter. Especially OTT, as well as several notch filters
automated with LFOs are responsible for the typical reese sound.

Remember: the point of distortion and compression at this stage is re-emphasizing the loss of
high frequency content as well as bringing out the movement of the filters. However, if we
overdo said dynamic processing we will end up with a flat, lifeless signal barely having any
dynamics, despite it being extremely loud. Additionally, all filters will start sounding the same -
an unpleasant whistling that is creeping below the main signal.

This whistling is what the moving of a filter on a white noise sample (esp. with a bit of
resonance) sounds like. In principle we are now not emphasizing the movement of our main
signal anymore and instead get the sound of filtered white noise as noise floor, muddying up our
sample.

Below we see a close-up of a reese bass as well as a zoomed out picture of the overall
waveform. Note the dents on the fundamental sine wave from the distortion and how their
presence changes over time. We can also observe the phasing in the overall wave.

Fig. 21: Closer look at the waveform of a reese bass. The fundamental sine wave is dented by distortion. The top right
is more distorted, resulting in a more crowded appearance. Below we see the effect of phasing on the wave.
28
Strong Fundamentals
Another concept is having strong fundamentals in your track to emphasize the tonality overall
and have a full and rich sound, courtesy of Culprate.

This applies all over the place, from sub notes to the kick’s thump or the chonk of our snare.
Chords also act as fundamental notes compared to atonal field recordings like birds chirping.

Fig. 22: Example of a snare’s spectrum. We can clearly see the fundamental G2 at 193 Hz.

It is also advisable to tune our drums and percussion to the key we are working in (or forcefully
introducing dissonance by defying this).

Keeping this in mind will allow us to add to a sample that is lacking in a targeted manner rather
than trying to cram it all into the perfect synthesis of a sound, e.g. by layering a missing
harmonic in form of an extra sine.

Layers
The term has been mentioned several times by now but deserves its own spot in this list.
Layering is a core component of creating rich textures and a lively room in full tracks and their
elements. The basic principle is enhancing a sound by adding more complementing or
enhancing elements to it.

This could be adding a clap or snap shortly before a snare hit, ramping up to it (remember: we
can hear differences in timing as small as 20-25 ms) or adding a click to the transient of a kick.

We could add a reverb layer to our drum kit, have a synth play 2 octaves at once for more
harmonics or reverse the audio of a pad to use it as a riser towards our next downbeat.

29
 Fig. 23: Example of layered synths in a project.

Another neat example is restricting a voice to speaking range with a broad bandpass, then
forcing it to mono and having this upfront dialogue be surrounded by FX-altered stereo layers of
the same voice all around (e.g. Vocoder, Bitcrush, Delays or Reversed samples):

Fig. 24: Mono (top) and stereo (bottom) layers of a voice sample.

30
These sounds will then have to be balanced with each other, depending on which element we
want to lead or support. This is again done with volume and reverb, frequency (EQ!) and
panning.

Synesthesia
We can expand the visualization of a full tracks waveform (not to be confused with basic
waveforms like the sine) to a colour code: Bass, Mids and Highs.

Fig. 25: Visual appearance of 4 exemplary tracks in Serato DJ software. The frequency content influences the
coloration of the waveform.

This is something I picked up by DJing with Serato software, later realizing it was a very good
tool to visualize the quality of my own mixes in comparison to other artists and reference tracks.
With this we not only see the transient, the length of sounds as well as the space inbetween. In
addition we get mixed colours based on the frequency (= colour) composition.

Deep red means mostly sub / bass. The louder a frequency is, the more prominent its colour will
be. Yellow (red + green) and orange means it has a lot of melodic mid frequency content
together with bass and purple is what happens when there’s lots of high frequencies together
with bass.

In our first example above we see a green break, meaning melody and / or vocals, the blue parts
are most likely sibilances from the voice sample. Then comes a drum roll before the kick and
bass hit again. The purple colour suggests a lot of high frequency content on top of a bass, like
neuro layers, a cymbal or hi hats that are quite present in volume.

The 2nd example is a trappy, very bass heavy track with chopped robotized voices and noisy
sizzles, resulting in a mostly red visual that is occasionally shifted to mixed colours based on
which chop plays.

31
The intro of track #3 is made of organic percussion, a heavy kick and a kalimba like pluck and
we can clearly see the sequencing: a mostly red kick, the blue snare and the green pluck notes.
When percussion and pluck play at the same time the colours mix again.

The last example is a future garage track, driven by a drone like bass playing permanently and
strong atmospheres in the mid region. As a result we get the orange / yellow waveform we see.
Additionally, we can see the very short snare, that is more mid-heavy and thus acoustic in green.
The bass seems to be filled with a lot of movement, either by pitch or phasing that makes the
wave change form. If we were to compare to an 808 trap bass, the latter would be almost static,
like in our 2nd example.

We can also derive the volume and dynamic content of elements with this method: In all
examples the kick is the loudest element, often followed or matched by the snare. They are both
formed kind of like a brick, meaning they have been pushed with compression quite a bit. The
third, more organic example has less of an extreme kick, with a smoothly decaying tail
compared to the heavier tracks.

The melodies and bass are both lower in volume and more dynamic, exhibiting less of the
“sausage” style of a brickwalled compression. We can also see that the kick often completely
pushes away the other instruments (esp. the bass in example #4), giving it space to punch
through the mix. This is a direct consequence of side chain compression.

To help you envision the possibility to go from unsatisfactory early mixes to more easily
accessing the power of clean mixes early in a project I’ll show you two versions of my very first
track compared to an early version and the master of my track Versus.

Fig. 26: Comparison of different versions of tracks. Top: my very first project. Bot: Track of my most recent EP.

A first glance shows us that the limited and mastered versions (row 2 and 4) are louder, which is
desired for a finished release. The orange colour in the first track showed me that the balance I

32
wanted was not what I envisioned it to be, exhibiting too loud of a melody and too little low
frequencies.

Note: The volume of unfinished tracks and premasters is generally lower to allow some
headroom for the mastering engineer to work with your track without clipping -
commonly -3 to -6 dB - while a mastered track peaks at about -0.30 dB. However, making
our work in progress louder helps us to see whether the distinction and sequencing of
single hits works as intended or is muddied up.

For Versus I was very happy with the distinction between elements from a very early version
that was in fact more enjoyable than the finished (and later mastered) version of my first track
ever got to be.

Fig. 27: Example of a track that would be too quiet to properly judge. Being a finished mixdown / premaster this track
was deliberately given some headroom before reaching the mastering engineer.

We can refer to the same terms that are used in photo manipulation or digital art: Hue, colour,
contrast and saturation. Wait, saturation? As a matter of fact a Saturator distortion module has
the same effect on audio as increasing the saturation of a picture does, making it more vibrant
and pleasant.

Even the resolution of a picture can be brought in context with audio: We could be looking at the
greatest movie or digital art we have ever seen in 160p or 4k UHD resolution. The motif will be
the same and we will get the idea in both cases but the experience is so much more engaging
when we get to see the higher quality version.

The same applies to the clarity and execution of a mix - the most beautiful and heartbreaking
melody means little to nothing when it is blurry and cloudy like a cheap JPEG. Healthy
production and sampling choices will automatically lead to a better end result and are a matter
of experience and failing thousands of times.

With these ideas in mind I compared tons of tracks and artists over the years throughout
different styles and forged my own idea of what good sound is and what it looks like on paper -
completely separated from the creative / artistic quality of a track.

33
Session View vs. Arrangement View
Ableton gives us the opportunity to work either in Session or Arrangement View. Both have their
advantages and disadvantages depending on how your workflow functions but both are worth a
shot for different applications.

The general idea behind Session View is looping instruments, clips or samples in a timeless
fashion, allowing us to jam and change ideas on the fly. This is also very handy for preparing and
performing live sets. Here we have the possibility to play with all kinds of building blocks, from
full tracks to singular instruments and groups both as audio or MIDI.

This can be very helpful when wanting to jam with ideas separated from an external timeframe
other than the tempo of our track.

Fig. 28: Tweaking of drum loops in Session View.

While we can also loop in Arrangement View, there is a timeline associated with the order and
sequencing of clips and instruments. This might be useful as a fixpoint but can also prevent us
from having the proper perspective to go ahead and change something radically. We can record
directly into arrangement view from session view but most artists solidify their track and all the
small details sprinkled across in the actual arrangement.

34
 Fig. 29: Arrangement View of clips jammed and recorded from Session View.

Knowing these two ways of working exist and practicing with both of them will be very beneficial
to finding your workflow and when to use which. For example, jamming percussion beats
oftentimes works better in Session View while sound design and layering might work better in
Arrangement View, as well as crafting breaks and progressions.

Later on we will see some fancy ways to make use of both of them.

35
Philosophical Remarks
Waves Are Everywhere!
We’ve all heard the rather esoteric expression “Everything is vibration”. While this might sound
like an exaggeration we can actually compare audio waves with other kinds of waves quite
nicely. For example, light (or photons) can be described as electromagnetic waves, exhibiting
their own wavelength in Hz that can be transformed to their respective frequency domain (and
thus spectrum) via FFT. Each wavelength / frequency refers to a specific colour (in the case of
visible light) and energy. To elaborate, X-ray radiation has a very short wavelength and thus a
high energy associated with it - capable of disrupting chemical bonds - while radio waves have
long wavelengths up to multiple meters and are low in energy.

Additionally, quantum particles like protons and electrons, the very building blocks of matter,
behave both as particles and waves (called wave-particle duality). Their movement and position
can be predicted by use of very complex wave equations like the Schrödinger Equation.

The vibration of molecules is just a small portion of this energetic spectrum, residing in the
infrared (IR) area. This vibration or oscillation of molecules is what we can observe as heat,
triggered e.g. by the IR portion of sun rays. The better expression would then be “Everything is
oscillation” or “Everything is waves”.

Another fascinating parallel is that the upper

harmonics of a given note are multiples of
their base function. Atoms and molecules
exist in different energetic states and can
only exist in discrete states that also are
multiples of their fundamental frequency.

Funnily enough the depiction of this is called

the Harmonic Oscillator. This describes the
energetic states a vibrating molecule can
exhibit - with higher states oscillating faster,
at a shorter wavelength and thus with higher
energy.

36
The faster oscillation of the molecules means they drift apart (see X-Axis in Fig. 30), breaking
the bonds in extreme cases which is known as thermal decomposition.

Last but not least, same as quantum particles and light, both audio waves and waves of water
(in a macroscopic sense) show interference, the addition and subtraction to a new wave by
combining them. A consequence is phasing, which the double-slit experiment demonstrates
nicely.

Fig. 31: Depiction of the interference of waves.

37
Instruments As Voices
There are also other, less complex parallels I like to invoke. One of them is the thought that
instruments are voices and thus inherently carry information, whether we understand their
language or not.

Imagine the following: we humans use our voice to communicate with each other, spreading
soundwaves to share information. Obviously, this requires us to speak the same language or the
information will go missing for the most part. However, emotions can still be transmitted even if
we don’t know which words are spoken (e.g. yelling).

Fig. 32: A bird of paradise, as published by BBC.

Animals also communicate with their voices yet their language is even further from being
understood by us than a foreign human language. Additionally, us humans as well as birds tend
to sing to each other, be it for mating or entertainment. This means voices are an instrument
and if we reverse this statement we might go as far as claiming an instrument is a voice.

The combination of sound, melody and - dependant on the case - information invokes
something in us. While a lot of it is psychoacoustic effects or setting off an emotional response
we can also have direct biochemical responses to audio input.

The most prominent example would be the excretion of adrenaline in response to a scary noise
like a lion roaring behind us in the wild. For this very reason I am open to the idea of music being
a language we mastered to use for a purpose, without necessarily knowing the vocabulary.

38
Emotion In Music
Our reactions to different music are highly subjective so there is no general scheme to reach
everyone. At the very least we are able to access emotions and - esp. while learning to produce
music - will inevitably find instruments, atmospheres or techniques that will feel like something
that was made for us, speaking to us on the deepest level.

Fig. 33: Artwork created by Rodolphe Parfait as a depiction of his emotional state when listening to
Phutureprimitive’s track Kinetik. Title: Kinetic Coronation.

The combination of certain elements, a synth’s colour and timbre, a specific chord progression,
a rhythm or atmosphere might reveal themselves to be descriptive of our musical values and
this is a very liberating experience. Keep in mind that this comes through trial and error,
constantly reapproaching hardships until it will retroactively be unveiled to you!

39
Techniques
Variations On A Theme
More than just an expression, creating variations on a theme is a helpful technique to quickly
generate multiple ideas from a simple starting point. Whether we elaborate on a basic MIDI
pattern by expanding or inverting it doesn’t really matter. We can make use of this both in sound
design and in our actual arrangement of a track, for example by warping, pitching or stretching
our material.

Warp Modes And Stretching

This kind of manipulation is done via Warp Modes in Ableton. Warp modes are different
algorithms with which the software tries to stretch or warp an audio signal in time. We can use
this to quantize the transients of a drum pattern to the rhythm of our track or mangle some
crazy samples we have just created to get more source material with a similar vibe.

There’s different types of warp modes suited for different uses, depending on what we want to
achieve. As soon as we stretch or pitch a sample away from its original state, our selected
algorithm will start working.

A quick overview looks as follows:

Beats focuses on the beat structure via transients.

Tones and Texture modes work with grains to preserve the tonality or grit
of samples - great for things like detuned samples (think Neuro Basses).

Re-Pitch simply pitches our sample depending on how much we

stretched it, e.g. doubling the length halves the pitch.

Complex and Complex Pro are very handy for stretched, tonal samples
like pads. The latter is especially useful for working transparently with
voice samples. Feel free to experiment with them to find your own rules!

By stretching and pitching we can create a pad from a percussion sample or make a bass stab
from a chord - the sky is the limit! This can be further enhanced by using Ableton’s sampler,
giving us the opportunity to impose pitch envelopes, filters and LFOs on a sample.

40
Comparing Waveforms
We’ve extensively looked at waveforms of tracks and samples and for good reason: By
comparing waveforms we can deconstruct the content of a given signal, copy the waveform of
a kick or analyze the patterns in a track.

Fig. 34: Analysis of a tracks waveform to deconstruct its contents. An impressive example of transparent sound
design!

As you can see here, we are easily able to identify the position of different percussion hits by
knowing what they might look like and listening to the track. This way we can write down the
MIDI pattern to copy the groove of a track that inspires us, giving us a nice starting point to work
with.

Remember: Any percussion usually has a transient that will be easy to spot, the kick and snare
will have some body to follow and basses tend to be sausages with long wavelengths! The kick
exhibits longer wavelength content compared to the snare.

Additionally, we can compare our mix to a professionally mastered track with a spectrum and
find the right balance - how loud are the sub and kick, how much high frequency content is
present? By looking at the sub we can also easily identify the notes being played!

Fig. 35: Spectral analysis of the track shown in Fig. 34.

41
In the example above the sub peaks at -22.7 dB at around 36 Hz - representing a D0 note. The
first harmonic is a D1 at ~ 73 Hz, about 7 dB louder than the sub with -15.2 dB.

Note: For our examples here we are only using Ableton’s stock spectrum plugin but there
are better alternatives like the free VST SPAN that give you a more precise idea of what
you are looking at. Learning to use this is worthwhile and takes some practice!

As mentioned we can go as far as copying actual waveforms to learn how to synthesize them.
Below you can see the zoomed in kick of our previous example and a rough sketch of it, made
with the VST Kick2. Quite obviously they are not the same and we can also explain why:

Fig. 36: Closeup of the kick from Fig. 34 (top) and rough reconstruction of its waveform (bottom).

On the one hand, it is a bit of a tedious process to perfectly match the frequency content at any
given point of a wave.

On the other hand, we are looking at a full track - not just an isolated kick sample - meaning we
will have additional dents and ripples in our waveform, originating from other high frequency
content like crashes, hi hats or melodies playing at the same time and the distortion on the kick.

However, in theory we can reproduce any kick by layering different clicks and distortions as well
as tweaking the pitch and volume movement. Kick2 can be a handy tool to do this but we can
also use a synth like operator or refer to sample banks if we don’t desire the hassle of getting
lost in synthesis.

42
Fig. 37: Pitch and volume automation of the 2nd kick from Fig. 36 in the Kick2 VST. With this tool we can freely shape
a kick to our desires.

By visualizing what we hear we have an easier time pinpointing problem areas and concepts.
Combined with the use of groups and song sketches we create fixpoints that reduce the
amount of active variables and prevent choice paralysis.

43
Be Random
A good way to create interesting source material is to break free from context and the idea of
creating a meaningful piece of art like a full track. By keeping the sound design process mostly
separate from our creative process we enhance both: We are less critical to whatever sound
might be coming out of our speakers when it is not associated to a track - also we avoid the
pitfall of tweaking settings for 30 minutes when we should be working on a tracks progression,
thus killing our creative flow.

This paves way to doing highly random stuff in the sound design
phase and saving it for a later point in time - this might be months
or years down the road. Using our own premade samples will also
get rid of doubts whether or not we are “allowed” to use stock
samples (Spoiler: you are!).

Keep in mind that incorporating lots of movement will lead to more

interesting and less predictable sounds so automation in this stage
goes a long way.

One of the best ways to capture interesting bits is recording idea

jams where we tweak on a synth’s or sample’s FX knobs for some
time while recording the whole procedure.

The result is a file some minutes long from which we can curate
interesting bits or further process them before saving what has
been worthwhile.

Fig. 38: Examples of the naming syntax for curated samples (top right), an idea jam as well as curated chops
(bottom). For tonal samples it is advisable to note the root note.

44
Another great technique to create randomized sound can be seen below. By throwing a bunch of
samples of any nature in session view we can create a randomized playback that jumps
between those samples:

.
Fig. 39: Creating randomized sounds with follow actions in Session View.

To achieve this we select all samples, put them to Loop and Legato mode and create Follow
Actions. These follow the timing and algorithm we select. Recording this generates a more or
less random piece of audio. Any kind of FX we can imagine can be imposed before or after the
recording.

45
We can also emphasize artifacts of time-based FX by pushing them into a limiter, boosting them
to the same level as the main part of a signal (courtesy of my friend and mentor Køps). A
convenient way to do this is using 2-3 utility tools with gain cranked all the way up (+ 36dB) and
a Limiter. DO NOT FORGET THE LIMITER! OTHERWISE YOU MIGHT DAMAGE YOUR EARS AND
/ OR EQUIPMENT!

Fig. 40: FX setup to record artifacts of time-based FX and distortions. Mapping parameters to an audio effect rack
gives way to different sound colours and interesting movement.

We feed this setup with a pulse of a signal that is followed by a time-based FX unit and / or a
distortion (e.g. Waveshaper, a special case of a Saturator). By turning up the gain while limiting
the louder parts we brought up the artifacts created by our FX unit and the distortion, giving us
interesting clicky and foley sounds. Automating parameters like Delay times or others is very
helpful.

Note: We can also feed this with longer sounds but will end up with a big fat sausage of
noise in most cases, owing to the stacking and limiting of too many sounds rather than
the accentuation of a signals tail.

Bouncing And Reverting

As we see over and over, printing what we do to audio samples (referred to as bouncing) gives
us additional tools to tamper with our source material opposed to simple MIDI instruments. This
can also be very helpful when we have a melodic or chord progression in a track and need some
way to create movement and texture in the background.

For this, we can simply bounce our instrument to audio and then
manipulate it by warping, reverting etc.

An easy way to do this is using the Freeze function in Ableton. Simply

freeze your (MIDI) track, drag it to a new audio track while holding down
Ctrl and you will have an audio version of your MIDI file including all the

46
FX you used. As a nice bonus, by doing it this way you get to keep the original in case you need it
later.

Fig. 41: Freezing arrangements gives way to new processing and ideas, fixing variables while also saving CPU.

Note: We can use the same process not only on MIDI but also on audio tracks, rendering
down the FX. There’s also the option to Freeze & Flatten which replaces your former
channel with the new audio track. This means you will not have access to your original
source down the road so be careful!

Fig. 42: Reversed snippet of a frozen clip used as a riser (top). This riser looks similar to a reversed kick sample
(bottom).

Above we can see the use of a stretched and reverted snippet created from our frozen clip, used
as a riser.

Fun fact: The waveform looks surprisingly similar to a reversed kick. Without listening to it we
can postulate being able to make a kick from this sample!

47
Remix Yourself
Once we have a more complex track structure with some phrases and breaks we can
experiment with it by changing their order. With a few breaks and a progression in hand we
might notice that our third phrase makes more sense appearing earlier in the track or that we
need more patience leading up to a specific part in our track.

Keeping in mind that our work in progress is a fluid idea we can go ahead and remix ourselves,
swapping building blocks like legos until we feel it is right. A group structure and separation of
phrases via locators facilitate this process. We can also insert empty time with the shortcut
Ctrl+I.

Fig. 43: Overview of phrases and their contents in a more complex track structure. We can swap these building blocks
by choice.

Again, we could also be bouncing the output of a full group to audio and substitute some of our
melodies’ notes, e.g. with its reverse part to invoke more interesting push and pull movement.

Fig. 44: Substituting notes with their reverse within a pattern can create interesting movement.

48
The key factor here is generating a Call and Response between elements on a micro and macro
scale. If we combine this with FX, the space will easily be filled out with interesting textures.

A great example is recording a sample with a reverb tail, then reversing it and applying another
reverb. This way we have the reversed reverb of the original sample acting as fade in - like the
attack time of a synth - while also smoothly having it decay with the second reverb. Whispering
ghost-like voices can be easily generated this way!

Fig. 45: Two main layers of a track enhanced with gated chops, a textural sample and a riser, all made from audio

49
Reverting is just one example, feel free to experiment with anything you can think of. In the
context of movement we can also automate the pitch and filtering of such samples to ramp up
to following elements or chop an atmospheric sample to be gated (staggering effect!).

Fig. 46: Pitch envelope on a riser (top) and gating of an audio sample with the Beats warp mode (bottom).

Time Signatures
An honorable mention in our technique section is the use of different time signatures. Never be
afraid to experiment with something like a 6/8 signature to impose a feel different from the
classic four-on-the-floor type beat. Topics like swing and quantization also come in handy.

A creative way to use time signatures without deviating from the 4/4 pattern we are used to can
be alternating between 3/4 and 5/4 timing, effectively switching the attenuation of the downbeat
every so often while still summing up to be a 4/4 beat (courtesy of Maitri).

Fig. 47: Time signature control in Ableton enables a lot of creative possibilities.

50
Watch How Others Work
The best way to sharpen your decision making skills while navigating your software is by
watching how other people work. This doesn’t mean just watching “how to XY” videos which
mostly explain one single type of sound.

Our goal is to educate ourselves in a way we can pick up knowledge and adapt it in a different
context by also trying to understand why things are done the way they are. This converges with
the concept of finding conventions. Whenever you watch other people work try to pay attention
to the details between the lines.

Watch out for EQ settings, the order of FX, the grouping of the project, how much dry/wet ratio
is on a FX module etc. Most of these things will not actively be covered in the context of the
tutorial but the information is still on your screen, ready to be analyzed and learned from.

Actively asking yourself why an artist does some specific processing will tremendously
increase your confidence in making your own decisions. Repetition, trial and error will then add
up to experience.

As a nice bonus you might be inspired by others taking a different path to the same goal or
doing something you never considered inside the framework of your own rules. The perfect time
to fire up your DAW and explore!

Production Via Ambience

By working in the digital realm we are working on a blank page, meaning that when nothing is
playing we will be dealing with complete silence. While this may be desired at some points to
create tension it also makes it harder to sketch, owing to the lack of context.

We can circumvent this with a neat trick: implementing background noise that acts as
non-intrusive static to help us stay in the zone. This can be vinyl crackles, a recording of pink
noise like a thunderstorm or field recordings such as birds chirping or a river streaming down a
valley.

51
 Note: Implementing these layers directly in our tracks arrangement again means there
will be absolute silence as we stop playback.

When jamming and testing ideas it can be advantageous to have said ambience running
externally, for example a 10 hour long recording of a thunderstorm playing on Youtube while we
work in our software. This way we can keep a steady headspace! Ableton’s vinyl crackle has the
benefit of always running as long as the module is turned on.

Fig. 48: Ableton’s vinyl distortion gives us the opportunity to have a static crackling running over our track.

Not only does this facilitate creative work but it also is a great way to breathe life into a track,
emulating real places and organic sounds in which our track resides.

52
Writer's Block And Frustrations
If You Want To Change The Outcome, Change The Process!
With great power comes great responsibility - being at a point where anything could possibly
turn into a great idea or technique it becomes just as easy to feel mentally blocked, preventing
us from expressing ourselves or experiencing joy during it.

The countless variables at our disposal when interacting with music, from the tiniest detail to a
full track act as pitfalls popping up left and right. Plentiful are the examples of how we can drift
into feeling frustrated and meaningless, like we will never be as good - or good enough - as we
hoped to be.

One of the most memorable demonstrations from personal experience was copying the
complete percussion pattern, song structure as well as the amount and tonal quality of
instruments with their distinctive roles from a reference track. Even with a similar atmosphere,
progression and sonic texture the reference track always had the impression of being more
lively - despite being of a rather loopy nature.

Fig. 49: Referencing other people’s work can both be helpful or paralyze us when trying too hard to emulate them.

Being hellbent on creating the same experience coming out of my speakers, put together by a
seasoned producer, was bound to fail given the fact my production experience was limited to
just about 8 months. Inevitably, the project was laid to rest for a while before new inspiration
manifested itself from different sources (specifically a voice sample and a harp).

Later on it became the title track of my debut EP - Salix Lacrimosa, the reference being Sigil -
Camel.

53
Conclusion
It is important to note we are not the artists we try to emulate, instead we are a sum of all our
musical influences, inherent rules and decisions along the way. Our way to break free then
becomes as easy as looking for new inspirations from other places and contexts.

Accepting a change of outcome, we circumvent our usual thinking and workflow either by
chasing an idea or testing a new technique we picked up along the way. Whether it be a melody
popping in our heads or an atmosphere touching us - the possibilities are endless and abstract.
Expect to be watching a TV show thinking “This quote will fit my track’s mood perfectly!” out of
the blue.

Whenever we are stuck it is advisable to step off the standard path and try new things and
change our perspective. Be random, chase ideas, indulge in atmospheres. Basically, follow the
concepts and techniques elaborated on earlier.

There are no stupid questions - as long as we keep asking “Why?” and “How?” there is always a
way out, revealing itself sooner or later. Every question we possibly have can be answered by
analyzing reference tracks and their composition.

Another great example is the idea of warmth. Having looked for what creates this feeling
seemingly forever, the answer to it will always be in the bass region (specifically 100-250 Hz) as
anything below (the sub) will mostly be felt rather than heard and anything above will already be
too bright.

Inspirations can be direct or abstract and referring to multiple sources of material helps us see
the bigger picture. With the ideas presented in this work I hope to have offered some guidance
for you to find your own rules and workflow.

Enjoy the process!

Much love,

54
Resources
Below you will find an extensive list of resources to follow up on some of the concepts
mentioned here. As producers we are sound designers, composers, mix engineers and
multi-instrumentalists so there is a lot to learn. Happy producing!

Track Analysis Technique by Pyramind

General Warp Vocals In Ableton
Ableton Manual
Mastering Ableton Live Software From Scratch
Shortcuts In Ableton From The Ground Up With Mubali
Nine Components Of Sound Starting A Track With Frequent
BBC Sample Library Writing Ambient Trap Beats
Writing Experimental Music
Must See Writing Future Bass Track
The Art Of Mixing
Writing Grime Footwork
Mixing Masterclass
Writing Hip Hop Beats
Punchy Drums With Operator - Au5
Writing Trap Beats
Drum Synthesis
Writing UK Grime
SPAN In Mixing by AHEE
Sampling Silence Sound Design
Bass In Ableton’s Analog
Tutorials By FabFilter Feedback Sampling
Beginner’s Guides
Granular Simpler - Woulg
Plugin Introductions
Growl Guide
Mixing And Mastering
Growl Guide #2
Kick & Bass Tutorial
Techniques + Workflow
Audio Alchemy Webinar Series Neuro / Reese Bass - Theory & Workflow

Chain Selectors Neuro Bass Guide

Drum Fill Chain Selector Phase Modulation - Au5

Follow Actions With Vespers Psytrance Leads

Group Structure And Busses With AHEE Psytrance Grid Synths

Indian Rhythms Randomization In Sound Design With Halfred

Lorn Style Tutorial Sound Design Tutorials With Frequent (e.g. Sine

Making The Most Of A Sample Pack - Au5 Compression)

Musical Storytelling Squishes And Squelches

Recording Screen Audio WIth Audacity Vocal Processing with Subaqueous

Sound Morphing Tutorial by Pyramind

Strong Fundamentals - Culprate Masterclass
Culprate Masterclass #2

55
Mixing & Mastering
Importance Of Mono In A Stereo Mix
Metering Subs With Spectrum - Mr. Bill
Mixing Sub BassWorking With Sub Bass
Rough Mix vs. Final Mix
Loudness Is In The Mix
Mixdown Tutorial
iZotope - The Mastering Workflow
Thoughts on DIY Mastering
Get The Most From Your Master

FX
The first source should always be the
manual to understand the fundamentals!
Understanding Audio FX
Arpeggiator
Surgical EQ Tools
Introduction to Corpus Effect Tap BPM
How To Use A Compressor BPM To Time Calculator
Multiband Compression Tutorial Spleeter
Sidechain Compression
New York Compression Talks & Lives
Finding A Place Through Music
Distortions in Ableton
Translating Feelings Into Sound
SAUSAGE FATTENER
Neon Vines Live Performance

In The DAW
With Au5 Downloads
Black Octopus Free Pack
With Bassline Drift
Drum Kit Reddit
With Frequent
Kontakt Komplete
Krush Bitcrush
Wider Stereo Tool

56
Final Words
This guide was put together during the production course of Phutureprimitive which gave me a
feeling of being able to connect beginner’s problems directly with the concepts and ideas I have
been developing in the first 2 years of my own production path.

Being able to see the bigger picture and connecting the dots unleashed a creative workflow that
feels incredibly liberating.

When I started my DJing and Production project the goal has always been to give back to the
community that welcomed me with open arms.

Thus this guide will be available for free to anyone who might benefit from it, from complete
beginners to seasoned producers looking to accumulate perspective or new patterns of thinking
about music.

While I adapted a lot of ideas I found, noticed or have been taught over time this is merely my
view on the matter of production and in no way claims to be fully technically correct.

If you felt inspired by this guide I welcome you reaching out to me at any time.

The guide will be released together with my 2nd EP “Versus / Nemesis” which will also be
downloadable for free. Each download will include a copy of the guide.

57