The Making of PatternRecognition.mp3 (or 'Zen and the Art of Songwriting')
A log of the development of a new alternative rock song, "Pattern Recognition". Includes Cakewalk Sonar tips and some notes on the creative process.
Thursday, April 28, 2005
Monday, January 03, 2005
Mining the subconscious for lyrics, more resonances with William Gibson's book "Pattern Recognition"
(Note: all quotes in this Jan 3, 2005 post are from the novel Pattern Recognition Copyright © 2003 by William Gibson)
I've come to realize that, for me, one of the keys to creating is to keep the conscious mind out of the process, and to use it for what it does best: editing and analyzing. The lyrics are a case in point - it's almost like I need to interpret them after I've written them to find out what I meant.
I've had a feeling that the lyrics I've written were somehow echoing some part of William Gibson's book, so I've been re-reading it, looking for influences. Well, I've found the passage that must have been percolating around in my brain. In Chapter 37, Cayce (the protagonist in the story) is brought to an alley in a foreign country: "the only light is from above. A bare bulb, visible up a forbiddingly steep flight of narrow concrete stairs that seems to have no railing" that leads up to a place that was famous in the '80s for hosting an ongoing party that lasted 7 years. In broken english, Cayce's guide tells her that there, people were always "Talking of freedom, art, things of the spirit...People valued friendships, talked endlessly, ate and drank. For many people it was like the life of a student. A life of the spirit." She describes "how it was, here, for artists. Whole universes of blood and imagination, built over lifetimes in rooms like these, never to be seen. To die with their creators, and be swept out."
I remember now how that passage hit me when I first read it. It seems I've somehow subconsciously fused the images and the feeling that this passage evokes with some of my own poem fragments from several years ago, U2's Vertigo, my interest in neural networks and how the brain works, Celebration of the Lizard's journey into the mind, combining aspects of all of these into the lyrics for this song. When I look at the lyrics rationally, from an editorial perspective, I can see so many flaws in them, but for some reason they still somehow work for me - I like the way they sound when sung, and the emotions they haul up out of me. At this point, I'm torn between keeping the song as it stands, extending it with another verse or overhauling it lyrically.
"Whole universes of blood and imagination, built over lifetimes in rooms like these, never to be seen. To die with their creators, and be swept out..." to join the sea.
Thursday, December 30, 2004
PatternRecognition.mp3: adding an intro & bridge, pushing the limits
Here's the latest cut of the song (DreamingInTechnicolor_PatternRecognition2.mp3 2.62MB, 2:51). The name of the song has been changed from AfterImage to PatternRecognition, since the new bridge part flips the song's focus away from the AfterImage and zeroes in on Pattern Recognition. It also honours the William Gibson influence. Since I started the song, I've been re-reading Gibson's book, Pattern Recognition, and there some cool resonances with it. I love Gibson's central theme: "We have no future because our present is too volatile. We have only risk management. The spinning of the given moment's scenarios. Pattern recognition..." And I've come to realize that the process of song writing is a form of pattern recognition itself. Captured beautifully by Gibson's phrase: "He took a duck in the face at two hundred and fifty knots." er, well, maybe not that one, but definitely "She has no way of knowing how she knows." I think one of the reasons I'm writing this at all is to try and figure out how my own creative process works, where the music comes from, with the vague hope of trying to help it along. So, here it goes...
After listening to the previously posted version of the song (1:55, 1.8M), I wanted to focus on adding an intro and bridge / chorus to the song to give it more structure, so I stripped it down to just the drums and the string riff and tried out some ideas with the VL70m. The result (0:16, 265K) was really lacking in any focus or drive, so I went back to the piano voice to try to get a more percussive and chordal feel going (take2 0:17, 274K). Which was totally at odds with the rest of the song, so I kept taking away notes until all that was left was the bass line, which really works for me. So I added it to the bass guitar track and then beefed up the bass sound by layering the VL70M DampBass voice with a fingered bass voice from my old Gravis Ultrasound PnP sample set and a Deep Bass sample, also adapted from some samples for the Gravis. To me, the best part of the vocal at this point was the echoed 'into the rhythm', so I extended the feel of this section to add in a layered background vocal for the night-club sequence. I re-used the new bass line from the intro to lay the foundation for the bridge section, and then shifted it around in the song until it felt right. I really wanted to change the feel of the song at this point, so I took one of the vocal samples I'd recorded, pitch shifted it down an octave and stretched it 2x and then applied some random pitch shifts to get this (0:33 529K). I also added an ascending string line to build up a chordal kind of progression. This was the first time I had a bit of a shiver up my spine with this song. Felt like I was on the right track.
I'd been thinking about ways to extend the meaning(s) of the song, and the one angle that I liked the best was to focus on capturing the way the mind works (perhaps, in retrospect, inspired by Jim Morrison's 'Celebration of the Lizard'). The song itself can be read as some kind of journey into the subconcious and the surfacing of some thought to the conscious 'light of day'. The only reason mammals like us are able to do anything interesting is that we can remember things, recognize patterns in what we've previously experienced and apply that to the present and future. It is so fundamental to us that we take it for granted. But it is also bound up in the creative process; we take all of these different bits and pieces of memory and combine them together in new ways, prodding the process with new stimulii from the outside world, and reassembling them in novel ways that feel somehow 'beautiful' or 'meaningful'. There's a great book called "The Creative Process" by Ghiselin that compiles letters from Einstein, Van Gogh, Mozart and others where the authors try to outline how their creative process works - highly recommended! (BTW, no surprise, my brain doesn't work like Mozart's!) There's also this Zen thing about catching your brain in the act - I have a hunch that the sound of one hand clapping is really an analogy for the left brain / right brain duality - the conscious left side of the brain isn't really in the driver's seat, it just THINKS it is. Is it possible to really get this stuff into a song, though? Should it even be attempted?
Once again, my approach to this was to not try to think too much about it but to get the music going and to free-associate while it was playing. After cleaning it up a bit, here's the result:
Flashbulb, feedback, Pattern Recognition
Seed crystal, flashback, Pattern Recognition
Precog intuition, Pattern Recognition
Visionary insight, distant memory, instant recall, Pattern Recognition
REM sleep, dream memory, Pattern Recognition
Time lapse, holographic, synapse, neural network
[Click here to see the lyrics for the entire song]
What I was trying to do was to set up some cognitive dissonance by using similar words and rhymes and sort of 'pre-create' words like flashback out of flash bulb and feedback and go deeper and deeper into the subconcious and physical processes until a 'eureka moment' of pattern recognition.
To get the right feel in the recording, two takes were recorded, both whispered. The MPX110 pitch shifting + delay was used for the first, the second was treated with Cakewalk FxFlange (Heavy stereo, full mix) and Cakewalk FxDelay (2 delays, Left to right with feedback). A slight delay was unintentionally introduced in the first take after 'distant memory' when I had to catch my breath, which causes a timing mismatch with 2nd take from this point on. This really works for me - it creates a lovely 'spread' between the two vocals, adding to the tension. Here's what the song sounded like at this point (2:36, 2.5M). It felt like the song was building and building and then just ran out of gas.
I got stuck for a while at this point. I tried 4 or 5 different approaches to try to get it to the next level, but none of them worked. After several days, I finally realized what was bugging me: I liked the bridge without the whispered vocals better than the version with the vocals. I finally figured out that I liked the sound of the weird pitch shifted vocal sample and the string progression, but both of these were getting lost in the vocals. So I delayed the whispered vocals until after the pitch shifted vocal sample and really built up the string progression, and then climaxed it with 2 'pattern recognition' sung samples. And that's where it stands right now (2:51, 2.6M).
Right now, the drum line is the thing that I feel needs the most work. Larry Mullen of U2 is my favourite drummer; listening to how he drums demonstrates how important drums are to a song. The song as it stands is very 'linear' right now; the drums just repeat and don't build up or change intensity during the song. So that should be the next thing to tackle.
Sunday, December 19, 2004
AfterImage.mp3 - adding vocals
So, here's the latest cut at AfterImage.mp3 (1.8M, 1:55) Tried out a new approach to vocals - lower pitched with more of an edge to them than I usually go for. Tried to create more of a club atmosphere by adding in some background shouts with echo on them - I've been thinking of trying out this for a while and it does add a lot of energy. Recorded a total of around 20 takes, some as clean lead vocals, some as harmonies, some with echo, some with oohs and ahhs and shouts - had a lot of fun recording this. Some songs require a lot of work to get a clean sound - this time I just wanted to capture some energy and edge, and I'm pretty happy with the results so far. I used a Shure SM-58 and an inexpensive ART tube preamp/compressor and recorded directly into the Sound Blaster Live! analog input. You need to push the input level of the preamp/compressor a bit to warm up the sound, while making sure the signal does not get clipped when recording into the sound card by watching the Sonar input level meters when recording. There are a number of problems with the song right now: there's no real structure to the song (it needs at least an intro and some kind of a bridge / instrumental section), the bass line needs to sound fuller, the sound is very crowded sonically and is poorly mixed - some clipping and balance problems. It definitely needs to be re-mixed. As it stands, the song is less than 2 minutes long, and I'm not sure how to address this without losing the song's energy and groove. I'm thinking of adding a guitar sound that starts off as a simple 3 note chord and then has one note staying constant while the other two get progressively weirder - pitch shifted, gapped, echoed, etc. Stuff that you can't do normally on a guitar, but since I'm getting the sounds out of the VL70m and have to layer them anyways, thought it might sound different enough to stand out. I'll try some ideas out and post the results.
Friday, December 10, 2004
The making of AfterImage.mp3 – an Alternative Rock song
Started a new song last night and thought I’d try to write a log of the song's development, sketching out the creative process and mechanics involved. This is my first attempt at blogging; no idea where this will go at this point.
I don’t have any ‘formula’ for writing songs; they can start in several different ways. Sometimes it can start with a lyric, other times I’ll get some rhythmic thing going with a drum line and bass line evolving and then layer sounds on top. The new song started from feeling like I wanted to explore some new sonic turf, try to break through to a new sound.
What I’ve been listening to recently: U2 (How to dismantle an atomic bomb), Radiohead (Hail to the thief, the Bends, Pablo honey), Linkin Park (Reanimation). I’m not into sampling or recreating other band’s sounds, but they do open up my mind to new possibilities, help me break out of the past. They get me in the right frame of mind and push me, challenge me to try and go further.
Musically, my background is as a sax player and vocalist. Over the years I've developed some rudimentary keyboard skills and have learned how to use the computer as a musical instrument in its own right. Here’s my setup:
Instruments: Tenor sax, Yamaha WX5 wind controller
Synths: Yamaha VL70M, Yamaha DX7, Korg MS-20 with MPU-101 midi converter
Sampler: Gigastudio 96
Sound processing equipment: Lexicon MPX-110
Mic: Shure SM-58 through an ART Tube Preamplifier and Compressor
Speakers: Tannoy Reveal
Headphones: Audio Technica MH-40fs
Sequencer: Cakewalk Sonar
Soundcard: Soundblaster Live!
PC: 800MHz PIII 512MB RAM
Not high end stuff, but I find it’s everything I need.
With an instrument like the sax, in order to get to a new sound you have to ‘woodshed’ (i.e. practice with no one around) for a while, refining your skills and trying to push your limits. In abstract terms, I think what this involves is getting familiar enough with some new thing that you no longer have to involve the rational left side of the brain in this new activity, you can do it while remaining in the creative / intuitive right-side of the brain and let the limbic / motor system or whatever do the rest. For me, the same holds true for computer-based music. I find I need to get all of the rational stuff – configuring equipment, learning how to use the equipment and software, trouble-shooting – out of the way so that I can get into a right side of the brain zen groove to create music. Once you feel like the equipment is not in the way of the music you are trying to create but instead is a natural extension of you, you can fly! This takes time and planning, however, especially to get the most out of your setup.
Lately, I’ve been trying to push what I can do with the Lexicon MPX110 by creating an instrument map that makes it easy to ‘play’ the effects in the same way that a synth is controlled in Sonar. It’s just a matter of copying all of the config info into a text file in a standard format. You can download the map I created here. I’ve also been playing around with the Cakewalk application language (CAL) to auto-generate midi effects, specifically auto-gapping the midi stream by injecting volume control events. Not sure if I’ll use it in this song yet – we’ll see if it fits the song or not. The CAL file is here.
Limitations can force you into new approaches, and sometimes that can be a good thing! I love the VL70M sounds for their rich expressiveness, but it’s a mono device so building up e.g. guitar riffs is pretty slow. It involves laying down each note of the riff in individual mono tracks and recording each track to a .wav file before combining them. When you are getting started on a song idea and don’t even know what tempo you want to go with, this can really break that zen thing. So I thought I’d try playing around with the Lexicon’s pitch shifting abilities to see what happened. Picasso once said something along the lines of ‘you are your mistakes’ (?)which I take to mean that true creativity often comes through the way you react to the unintentional. It forces you out of memory and into new learning, and you incorporate that learning into who you are. Well, some pretty cool stuff can happen when you combine VL70M voices with midi controller events and the Lexicon’s pitch / delay program! Just riffing with this for a while was really amazing. Here’s what resulted: AfterImage_a.mp3 (0:13, 212K)
Which is unlike anything I’ve heard before, let alone created before (maybe I lead a sheltered life…there’s probably some electronica / techno out there that has plowed this field already, but hey it’s new to me – ignorance is bliss). Started putting down some drum hits to try to feel out a rhythm – there’s definitely a rhythmic thing in there (maybe due to the echo), even though the midi controller data I injected to control it was intended to be somewhat random, more of a shape than a groove, if you know what I mean. It was originally recorded at 100 bpm, which is quite slow, so I sped it up to 128bpm and added a drum line(: AfterImage_b.mp3) (0:14, 219K). Then started complementing the drum line and VL70M weirdness with a bass line. Not really melodic in feel, more semitone transitions. I’ve been trying to identify why occasionally something I hear or play will send shivers down my spine (which is a good thing!) and I think it has something to do with when the music sets you up to expect something and then twists it slightly so that it harmonizes with what you expected; you end up resonating with the music in this weird way. I don’t have any kind of formula (unfortunately!) for this; it just kinda happens every so often. Hasn't happened yet, but the semi-tone thing feels like its in the right area for me right now to help create this feeling, kind of like sus-2 chords, and hey it just felt right so I kept riffing in that direction. After ripping it apart and reconstructing it a few times, here’s the result: : AfterImage_c.mp3 (0:12, 203K)
Next I tried to re-inforce the 128bpm tempo by adding a Stratocaster midi sample with a 1/4th note echo delay and some pitch bends in for a bit of strangeness.AfterImage_d.mp3 (0:12, 193K)
This started to sound like something, so I ‘jammed’ over top of it and came up with a vocal line:AfterImage_e.mp3 (0:12, 189K)
Adding this got me doing word associations. This is where everything you’ve done in your life starts coalescing. Listening to the groove that was starting to take shape made me think of the phrase Pattern Recognition - it just surfaced from that and fit the music. I really like William Gibson’s books, and read Pattern Recognition several months ago. Didn’t really think of writing a song about it or using it as a lyric until that very moment, but it just fit somehow. A poem fragment that I wrote when I lived in Ottawa (‘bare bulb burning bright like life, swinging to the beat of an unheard rhythm) was the next thing that the music called up. The rhythm of the words and then the image of staring at a bare lightbulb made me visualize the after-image retina burn you’d get. After Image stuck as a great intro line – like some flashback to something that has already happened. And recognizing the pattern in the after image burned into your mind dovetailed so beautifully with it. "After Image... and Pattern Recognition" has a great internal rhythm to it with the alternating stressed / unstressed syllables. Say that a few times over the music and the melody for what must surely be part of a chorus or background vocal coalesced.
When I’m working in this way, I like to build up the sound and song by looping 8-16 bars of the music and just playing it over and over, incrementally tuning the midi sequence and adding parts. Works well for lyrics too – just open up Notepad and start writing down the words that sync up to the music, rapping them over the music to see how they fit, then adding a midi melody line, iterating until music / words / rhythm and melody converge. Here’s what it sounds like so far, with the 'chorus' or whatever it will become repeated 6 times - by around the 4th repetition you start to really dig into the groove: AfterImage_f.mp3 (1:00, 940K)
First cut at capturing the images that lead up to the After Image / Pattern Recognition phrases:
and pattern recognition yeah
burning bright like life
with this unheard rhythm
swinging to the beat
like a light bulb burning
through the night
lighting up the back alley
up the fire escape
into the music
into the rhythm
and from the fire escape
tape to tape
cheek to cheek
dancing to the beat
of this unheard
rhythm and rhyme
driving bass line
deep down drum beat
right off your feet
until the day light breaks
and wakes you up
and its all gone
now all that's left
is the retina burn
and the sonic shock
just an afterimage
and pattern recognition yeah
and pattern recognition yeah
and pattern recognition yeah
The ideas for the song were obviously influenced by U2’s Vertigo (where the singer is calling from a dance club) but are at the same time very different. I started a song a while back called Backstreet Epiphany which geled with the other words to get this image of going up a back alley fire escape into this loud night club, and trying to piece what happened together from the after image of it all. I like the image of entering another world, some kind of parallel universe that you’re not really aware of until you’ve discovered it, and then leaving it and only having the memory of it to say it really existed. The main thing that is shaping the lyrics at this point is kinetics – having the words bounce around aggressively. Obviously, listening to Linkin Park, especially Reanimation, will do that to you. Charlie Parker once said something like ‘You have to live it for it to come out of your horn’. When I was younger I didn’t really realize how true this is. Everything you do combines to create possibilities for you to shape “now” with. One thing I want to try with this song is to write lyrics that work on multiple levels - time for me to really try to push myself here.
That’s it for now – sounds like there might be something worthwhile happening here. I’ll post again when I’ve made some progress on this. Sometimes songs come together quickly (e.g. Secret Agent Girlfriend), sometimes it takes years before they get ‘finished’ (e.g. Crystalize). Feels like there’s some potential in this one – let’s see where this goes!
Dreaming in Technicolor