The inspiration behind Carefully Introducing Problems is best described by what Scott “Knobs” Harper wrote for the intro to his video Fantastic flaws (& where to find them):
All music is full of wrongs.
Mistakes that have become the norm over time.
Mistakes and discoveries and mistakes.
There’s also this quote from Tony Rolando of Make Noise, about his modular synth gear:
Often, at trade shows, people will ask me, 'What problem is your product solving?' Typically I say that it's creating them. This product does not solve a single problem, unless you say it solves the problem of inspiration.
Among singers and violinists, vibrato was once considered gauche — but it became the norm because it helped overcome the limitations of gramophones before electronic amplifiers were invented. Since the early 50s, rock guitarists have taken advantage of damaged speakers and overloaded circuits. In the 60s, Reed Ghazala invented circuit-bending, the art of shorting out electronic circuits to produce “wrong” new sounds. John Cage experimented with variable-speed turntables, Christian Marclay intentionally damaged vinyl records, and DJ Kool Herc made record scratching into virtuoso performance. In the 90s came glitch music based on intentionally broken digital audio technology. The chiptune and synthwave genres explore the nostalgia of 80s video game sound and synthesizers. Producers run recordings through analogue consoles, tape recorders, old samplers, old delays, guitar amps — or plugins that emulate them — because “clean” audio just does not sound right to us, and all those flaws add character and emotional weight. It’s a sort of wabi-sabi involving technology and technique.
My own music tends toward minimal voices and simple structures, but it has an overall organic messiness thanks to adding these “wrongs.” In fact sometimes if a piece of music just isn’t working out, making it wronger will make it right.
Throughout the album
As usual, Eurorack modular is my main instrument, with Soma Lyra-8, Yamaha Reface CS, with a few software instruments and a lot of software effects. My DAW is Bitwig Studio, and controllers are the Make Noise 0-Ctrl, 16n Faderbank, and occasionally the Expressive-E Touché SE.
My musical process is a little different than what many musicians are used to:
- Building. During this phase, I patch synthesizers, assign controllers, add effects, create sequenced loops or algorithmic sequences, and tune and balance different voices to work together. I also make some general plans for how the performance is going to go. The Building stage includes elements that would traditionally be called sound design, composition and mixing, but it doesn’t complete any of those elements.
- Performing. I perform the piece, recording the full mix with effects baked in. This is almost entirely improvisational, and mainly involves manual control of mix levels and timbres, transposition, starting or stopping clocks or sequences, and playing any manual parts. Once I’m happy with it — usually after the first take, if I don’t interrupt myself to change or fix a glaring problem — I will unpatch my synth and close the DAW project. No going back!
- Editing. Alternating between Sound Forge and Bitwig, I’ll clean up the start and end of the recording, adjust the overall dynamics and equalization, and fix any clicks or glitches that are wrong in the wrong way. Often I’ll apply more effects to glue the recording together or tame excess harshness. Sometimes I’ll cut and splice parts or make more radical changes, and this can be as much a creative part of the process as the first two stages.
- Mastering. This stage is a little more conventional. I work to balance the levels of each track on the album and (gently) maximize them overall, check phase alignment, and make minor adjustments so things sound their best. My mastering chain typically consists of EQ (CraveDSP CraveEQ or Toneboosters Equalizer 4), DDMF MagicDeathEye compressor, u-he Presswerk used as a gentle limiter, and Toneboosters Barricade 4 as a final limiter.
(Patch notes for each track below don’t include all utility and modulation modules, nor basic EQ and dynamics plugins.)
01 Drop Bears
Voice 1: The Harvestman Zorlon Cannon into Mutable Instruments Rings, with LFOs into Structure & Position, through Native Instruments Supercharger GT and Wavesfactory Cassette.
Voice 2: Zorlon Cannon into a second Rings, with an LFO into Structure.
Voice 3: Same as #2, but also through Mutable Instruments Beads.
Voice 4: A drone from Plogue Chipsounds, with Sugar Bytes Wow2 and Native Instruments Raum.
Voice 5: Mutable Instruments Stages set up as a sequential switch, clocked at audio rates, and fed by 3 outputs from Zorlon Cannon plus the SynthTech E352. Through Make Noise Mimeophon (modulated with noise from the Zorlon Cannon) and e-phonic XPressor. Pitch sequencing is from Mutable Instruments Marbles.
The entire mix runs through Denise Bad Tape, XLNAudio RC-20 Retro Color and Wavesfactory Cassette.
02 Activity Indicating Activity
The Harvestman Hertz Donut mk2, gently lowpass filtered by Random*Source Serge VCFQ processed by Rings, Valhalla Delay, and Bitwig pitch shifter (on low frequencies only).
Marbles sequences Rings’ pitch, and modulates wavefolding on the Donut.
03 A Chindogu Must Exist
Voice 1: Both outputs of ALM Busy Circuits Akemie’s Castle, crossfaded at audio rate in stereo by the E352’s two outputs (using Expert Sleepers Disting EX), through Beads. Castle and E352 are sequenced by Marbles.
Voice 2a: Synprez FM on an Android phone, through Plogue Chipcrusher, Valhalla Supermassive & Valhalla Delay — sampled, and then looped in Bitwig Grid with Sonic Charge Echobode.
Voice 2b: Same as 2a, but the sample looper is in Cycle mode rather than Repitch.
Voice 3: A secondary crossfader output from Voice 1, through Rings, with its Position LFOd and Structure sequenced by Make Noise 0-Ctrl. With Bitwig Pitch Shifter, Supermassive, and Valhalla VintageVerb.
Voice 4a: Both oscillators of the Harvestman Kermit at audio rate, with waveshape sequenced by 0-Ctrl. Ring modulated together, through Raum, Valhalla Delay and Supercharger GT.
Voice 4b: Same as 4a, but the two Kermit outputs are run through Bitwig comb filters in mid/side stereo.
04 Brick Eye
This is Arturia Buchla Easel V through Sugar Bytes Wow2, four instances of Gigahertz Lossy, Chipcurusher, and Audio Damage Rough Rider. The full mix runs through Beads (with the mix level under envelope follower control) and Supermassive.
05 Ten Heartbeats
Voice 1 (the drum) is Rings in modal mode, with Supercharger GT and Audified STA Enhancer.
Voice 2 (the chord drone) is Akemie’s Castle A output, through VCFQ, Audio Damage Ratshack Reverb, u-he Runciter and Lossy.
Voice 3 (another drone) is the E352 detune mode set to uni/phase, with Kermit modulating both phase & morph at audio rates, and Valhalla Delay.
Voice 4 (squeaky bits) is Hertz Donut’s B output, through Happy Nerding FM Aid, Xaoc Tallin, Red Panda Tensor, Catalinbread Adineko, Mosky spring reverb and Beads.
Voice 5 (another drone) is Akemie’s Castle B output, through Mutable Instruments Ripples lowpass, Tallin do add post-filter drive, Wavesfactory Spectre, and Arturia Plate-140.
Voice 6 (plucky sounds) is Hertz Donut A output through Rabid Elephant Natural Gate and Mystic Circuits 0HP LPG triggered by different patterns, through Mimeophon.
Clocking and sequencing are done by Zorlon Cannon, with additional clock processing for the drum by Monome Teletype.
06 Legitimate Salvage
Voice 1 is Rings through Mimeophon and Cassette. Rings’ pitch and Mimeophon’s zone are gated by Marbles, and Mimeophon Hold by Zorlon Cannon.
Voice 2 (noise/drone) is Zorlon Cannon triggering Maths at audio rate, through Joranalogue Filter8’s BB-N filter, with Valhalla Delay, Bitwig pitch shifter and Lossy.
Voice 3a (another drone) is Hertz Donut’s secondary output feeding Starling VIA Sync3 with phase modulation from Akemie’s Castle, through VCFQ FM’d by the other Sync3 output.
Voice 3b: Donut’s primary output. (3a and 3b both are through two instances of Ratshack Reverb with alternating short/long delays on left/right channels.)
Voice 4: Yamaha Reface CS through Elektron Analog Drive, with Adineko and Mosky Spring Reverb in feedback loop, fxPansion Maul, Bitwig Rotary, and Wavesfactory Spectre.
Voice 5: E352 with wavefolding modulated by Zorlon Cannon, through RIpples, Lossy and Valhalla Delay.
Voice 6: Zorlon Cannon pulses through two Bitwig comb filters and Sonic Charge Permut8.
07 Let’s Play Wormball
Voice 1: Madrona Labs Aalto through Valhalla Delay, Plogue Chipcrusher, STA Enhancer, and Cassette
Voice 2 is the Soma Lyra-8, with Supercharger GT, Runciter, and Valhalla Delay.
Voice 3 is Plogue Chipsounds through Runciter.
08 Cosmo DNA
Voice 1: Both Rings with gates into their audio inputs, into FM Aid, through Beads, Adienko and spring reverb feedback loop, and Raum.
Voice 2: Two triangle ouputs from Hertz Donut mk2, controlled by 0-Ctrl’s dynamic envelope and main gate, into Disting EX’s precision adder. Sent through a radio transmitter and receiver. Freakshow Industries Mishby, Melda MTransformer, and Valhalla Delay.
Voice 3: Ripples self-oscillating into Schlappi Engineering Angle Grinder, which is modulated by Kermit. Natural Gate and Mimeophon.
Both 0-Ctrl and Rings are clocked by Marbles. The full mix went through Bitwig’s pitch shifter (in a narrow band) and then was downsampled to a lower pitch.
09 Cling Bitterly
Voice 1: E352 in cloud mode, with both outputs through FM Aid. Through Audiothing FogConvolver, Valhalla Delay.
Voice 2: d16 LuSH-101 through Supermassive.
Voice 3: A loop from a radio receiver captured by Beads, and resampled and looped in Bitwig. Through Supermassive, Chipcrusher, and Lossy. Pitches are selected via a fader, quantized in Bitwig Grid.
Voice 4: Zorlon Cannon into Rings input, through Beads (triggered by rhythms from Teletype), through Mimoeophon.
Voice 5: Akemie’s Castle drone (with Teletype rhythms into the V/OCT input), through Runciter and Supermassive.
This is a very heavily edited and processed copy of a rejected song, which originally involved some odd rhythmic pulses with Angle Grinder, and Sync3 attempting to follow E352’s cloud mode.
But it’s basically been turned inside out, liquified, fermented, and composted, and what you hear is the mushrooms that grew from the compost.
11 This Header is Deprecated
Voice 1: Kermit’s two oscillators cross-AMing each other, through Supermassive.
Voice 2: Two outputs from Angle Grinder, ring modulated by Disting EX, bitcrushed (also Disting EX), through Ripples lowpass filter, Valhalla Delay and Supermassive.
Voice 3: Sync3 driven by Angle Grinder and phase modulating itself, through VCFQ lowpass and bandpass outputs, with Ratshack Reverb
Voice 4: voice 2 processed through Beads, RC-20 and Spectre.
Sequencing from Marbles & Zorlon Cannon, not synchronized. Zorlon modulating Angle Grinder’s grind inputs affects the pitch due to its feedback loop.
12 Iteration Zero
This is Hertz Donut’s two triangle outputs — one with its internal waveshaper, one shaped by FM Aid — and Kermit through Angle Grinder, into two Disting EX crossfaders, through Bitwig’s pitch shifter and Supermassive. Pitch and waveshaping controlled by 0-Ctrl.