Wednesday, July 27, 2022
For this sketch, I went back to a loop-driven project after watching a neat YouTube video by Polarity about how to more reliably timestretch clips in Bitwig's sampler device. It was the advice missing from my last attempt at working from loops, and the technique helped this attempt feel much better:
I worked on this hours after returning from a long trip to Kauai, so virtually digging through crates of hapa haole songs was a way to give voice to my wistfulness.
Activities that seemed to make a difference:
I was absolutely attempting a "lo-fi beats to miss your vacation to" kind of thing. Evaluating the project after the fact, I can tell that I wanted to "validate" the work on the sample by having it be very present, but I think I should have at least tried EQing the sample down to more of an AM radio feel. I suspect it would sound better overall and still stand out.
Wednesday, July 20, 2022
Most of my time at the keyboard today was spent on piano-playing drills, but I did manage to load up "deep experiment" briefly to rework the chord progression once again. I ended up largely undoing what turned out to be shoddy work at the end of my last session with it. By "shoddy work," I mean that I had erred in identifying the 7th intervals for the chords—so the chords were simply wrong, and after fixing them, the inversions sounded more chaotic and didn't really provide the cohesiveness I thought was there.
I returned the chords to their simple forms and shuffled the progression around to a better-sounding-at-this-time iv - v - III - i (was previously III - iv - v - i). It feels a little like a bridge part, so I feel compelled to revisit it again later—but I'm left again with the feeling that I'm in a very naive place with this effort, and a little doubtful that I can "experiment" my way into something compelling.
I tried to read more about scale degrees and chord functions, but the advice I've found so far seems either too simple (like the introductory guidance about 2-5-1) or too far ahead of where I am now (I found what seemed like undergrad-level academic material describing the different functions of the third scale degree in different keys).
Sunday, July 17, 2022
My kid learned that I've been trying to make music and has adorably been asking to hear the recordings. "Deep experiment" was one she liked, and upon learning that there were two versions—a creative stub, and then what I considered to be a more evolved version—she told me that the stub version was better, and my partner even agreed. So I returned to the stub version and tried to iterate from it in
And here's a screenshot of Bitwig around the time of exporting the audio:
They preferred the bassline in the original deep experiment, so I returned to it. It was a sequence of F - G - C, and with that established, I figured I could try writing in F minor instead of F major. This time around, I did more homework—I wrote out all the notes of the scale and mapped out the scale's triad chords to give me a better map of what things to try. I wanted to evolve the bassline and ended up with more variation, mostly F - G - G# (darker!), with some F - G - C sprinkled in (dramatic!), and a fun F - F - D#- C# in the 8th bar. (I think of it like a little drum fill, but I suspect that's both (a) too soon for a drum fill and (b) not how other musicians think about their parts? Anyway...)
The chord progression and instrumentation in the first version of deep experiment was powerfully boring, and while folks in my family liked deep experiment 1 better, no one said that they liked the 2-5-1 pads, so that's where I started tinkering. The first chord progression I noodled out was a iv - v - VII - VI, and while I liked it, the scale mapping I wrote out made it more clear that this progression wasn't working alongside the bassline notes.
Attempting to work more with the basssline notes brought me to III - iv - v - i instead, which felt more stable. I did a lot of noodling to hear the chords in sequence, but my actual piano playing still sucks—so when the time came to record, I tracked the III chord, and then copied/adjusted the notes of the triad into new chords to make up the remainder of the progression. (Tracking the initial chord has the useful property of lending human unevenness to the played notes, so I still think it's better than simply drawing the notes into the chart.)
It sounded reasonable at this point, but I became much happier with it after playing with some inversions―with the intent of having the chord tones living closer to one another—and adding 7ths. But, I may have overdone it. While it did make for a more cohesive sound, it was also more flat and samey. (And I just noticed in reviewing the Bitwig screenshot that the manipulations I did on the i-chord ended up producing the same notes as the starting III-chord! No wonder it sounded same-y.) But now that I've done what feels like a stabilizing exercise to find out what musical ideas are at the core of this song, I think I'm in a good spot to revisit the composition to add more expression.
Saturday, July 16, 2022
deep-experiment-2-2022-07-16-0011.mp3, an attempted successor to deep experiment 1:
I felt that the Supersoft Pad and the 2-5-1 chord progression didn't work, and so most of the work here involved trying to improve that. I went with a synth instead of a pad and noodled out a new v - IV - i - ii chord progression. Then, feeling a little nostalgic for the Supersoft Pad, I used a layering technique to drive the Supersoft Pad using the notes of the string part.
This felt like a meaningful evolution of the prior ideas, and I was pretty happy with what came together.
Friday, July 15, 2022
It started with the Deep House Kit #1 and some play on the Oxygen Pro 25's really nice drum pads. With "organic house" still top of mind—see creative accountability, organic house from about two weeks prior—I tested out the kit's percussion sounds and was pleased to hear some woodblock-sounding samples, and ended up tapping out a pattern that felt natural. I knew I was making some kind of deep house/organic house track, so I set the typical 4/4 kick pattern.
Next, came the arpeggiated-esque bassline—see creative accountability, aka clouds)—another thematic approach that's been top-of-mind. Constructing the instrument was the most intricate part—it's the "Ganjubass" Polysynth preset with Note Repeater and Quantize devices added to transform played notes into a repeating series of 1/16th notes (if I hold the F key for a bar, it will sound like I pressed F 16 times). The notes were just noodled out; they sounded neat together.
In the past, I've regularly been able to get this far in producing something, but I've rarely been able to go any further. Until recently, I haven't had the time or training to know what to try next. But my instinct was to try to layer in a pad sound; the wide, sweeping tones that create a lot of size and momentum in the deep house/organic house compositions that have been serving as inspiration.
My naive take: this is a tough step because pad sounds are pretty subtle—as far as I can tell, the instrumentation is usually somewhat understated. If I were to dive way in over my head with an orchestra analogy, it's more strings than brass, and it means that it's hard to pick an instrument. Here, I picked something called "Supersoft Pad" knowing that I wanted it to take its time, though I think it's more forward than I really intended. I imagine if this project were to continue, I'd end up putting it through a low-pass filter of some kind to make room for some kind of lead instrument.
Unfortunately, I really didn't know what to do with the chord progression itself. I've been watching a lot of introductory piano, basic music theory and composition YouTube videos, and recently stumbled into 2-5-1-land—so an extremely rudimentary version of that is what's here, in the key of F major. To its credit, it sounds "complete," but it's not at all interesting.
The "Leftfield Raver" stabs at the beginning were a short attempt to find a lead instrument. It seemed like a dead end, but my affection for the instrument itself led me to try an early dusting for flavor only. In retrospect, I think it's too noisy and similar to the bass instrument.
Wednesday, July 13, 2022
I've been shying away from samples—they seem hard to work with—but after being inspired by videos about Bitwig's Sampler instrument, I noodled together
This was a good exercise, but I'm not too happy with the final product itself. Finding good source material is an essential part of the process and can be incredibly time consuming. Even in the best of circumstances, I'm short on time, and in this case I was rushing to try something instead of taking the time to find something more ideal or taking the time to shape the sample into something that works better. That said, the process was fun: I went to archive.org and looked through their 78 RPMs and Cylinder Recordings, and eventually found Bing Crosby favorite hawaiian songs Volume One. Track 9 is "Paradise Isle" and I found something likable to try.
I tried to accompany the sample with drums and a melody on the Wurlitzer instrument, but all told it feels quite off kilter. I think I'll have to either find better samples or learn better techniques for noise correction, time stretching, pitch, and slicing them to be played in a more MPC-like way.
Thursday, July 7, 2022
When a drum kit feels punchy, yet minimal, I absolutely cannot keep from playing a pattern like "Drop It Like It's Hot," and so here's
I configured the FM-4 instrument that's providing the beeps.
Wednesday, July 6, 2022
A day or two before working on this, I heard Supernature - Somewhere In Time and took the note "arpeggiated bassline," so that was clearly part of the inspiration for this bit of noodling. This piece has more songlike structure compared to most of the bits I've made prior, primarily in the form of introductory drum bars and more variety in the lead melody.
I've been trying to take inspiration from the dramatic clouds on the island. They sit low in the sky, move quickly, and tell vivid stories with their shapes. I don't think that comes through whatsoever in this composition, but as I tell my kids: trying is important.
Monday, July 4, 2022
I heard a song at Target and pulled out my phone to start up Bandlab's keyboard to identify some of the notes. I caught F-sharp, A-flat and some other notes from the chord progression and planned to try them out later. I didn't feel like taking the time to identify the scale and all of its notes by name (it's bluesy), but I used what I knew to noodle out some chords on Bitwig's Rhodes, and eventually ended up with the following:
To get there, I tracked the chords first, sequenced/tracked the drums, then tracked the melody (improvised a handful of times and used the best takes). The tracking is very loose, but I'm attached to the idea of avoiding quantizing everything (blame J Dilla) until I've improved my fundamental keyboard and drum pad abilities. I panned some of the drum sounds, ran the whole drum machine through a compressor for more punch, and mixed all the tracks into a reverb FX track.
After all this, the track still felt sparse. (I think that sparseness ultimately boils down to amateur musicality—seventh chords would sound fuller and more expressive than triads, for example.) I've been hearing a lot of textural components in songs that I like—samples or stabs or brief, shallow-sounding synth noises—and so I felt inspired to emulate that kind of approach to bolster the mix. After browsing through Bitwig's various FX presets and finding nothing appropriate (but many cool sounds for a different kind of track), I eventually settled on designing a bass noise in Polysynth. This was a really satisfying exercise—I liked the gravity it added, and I got to use Bitwig's automation feature to slowly raise the volume on the Polysynth's mixer track and bring it in relatively late into this short composition.
Sunday, July 3, 2022
The main point of this exercise was to independently configure Bitwig's Polysynth after watching a tutorial video and reading some advice about how to beef up the sound on a synthesizer. I liked the results and hope to use this synth configuration in a more sophisticated composition.
While I'd count this as a sound design victory, not much went into the musicality—the synth drives a bland chord progression (only fifths). The chords I sequenced were little more than a means of testing the synth sound, but I did end up adding more. I sequenced some drums from an 808 kit and toyed with an arpeggiated melody.