5. the word lists ; index ; 7. conclusion
(Because of the nature of listening, I think it's better
to hear the songs and get a chance to digest them before you read a
detailed analysis about how they were written. If you wouldn't mind
listening to the songs before you read the rest of this page,
here are the mp3s:
Kissing God.mp3
I'm Already Dead.mp3)
(Bold indigo words get a positive score in the word lists in the preceding chapter. They are highlighted here to illustrate the links between the songwriting and the data.)
I started the writing process by taking samples from piano pieces and string quartet pieces. My first goal was to get some simple soundscapes over which I could start writing melodies; I thought that the best place to start would be classical music since many of the classical instruments (and the word “classical” itself) got very high wordscores.
Based on the tempo and feel of the loops I was working on, I started experimenting with various drum sounds and drum patterns. I programmed the beats instead of playing them on the drumkit, although some of the drum sounds were sampled from my own drumkit. I wanted to established a strong groove, so I introduced a lot of fairly complex rhythmic content into the drums while sticking to 4/4 time to keep the songs danceable.
Sad and dark moods are preferred in the word lists, so I had a couple of brainstorming sessions and made a list of depressing past experiences. I matched them to loops based on feel and wrote lyrics using those depressing past experiences as guidelines. Whenever I came across a lyric and a melody that seemed to work well together, I recorded it to keep from forgetting it.
I let the songs structure themselves based on how many lyrics I wrote for each section, tweaking transitions only when things didn't flow smoothly on their own. I went back and wrote extra musical material for a chorus or bridge when I needed one. After I had basic song structures down, I went back over the songs and made a special effort to add some unexpected transitions. I also paid close attention to dynamics, since fast and slow crescendos both score highly.
When arranging the songs, I layered more parts than I normally would have. I took care to represent the entire frequency spectrum in hopes of getting a massive and expansive sound, and I used subtle processing (digital effects) in places where I could make the instruments sound distorted, echoing, lush, ethereal, or swirling and still fit the context of the songs.
I mixed the music with the knowledge that sudden loudness of certain instruments can make them appear heavy, and juxtaposed those sounds with sections of softer sounds. I used performances of acoustic instruments as a way to provide a bit of contrast to the sampled loops and programmed electronic drums. I also spent a lot of time creating prominent and unusual guitar sounds. The vocals are mixed to sound distant at times.
The first song: “Kissing God”
an mp3 of “Kissing God”, final mix
This song is written about a relationship that I had with somebody who frequently used heroin. I tried to approach it in a different way by writing it from the other person's perspective; the lyrics touch on a few of our experiences together and hint at our eventual breakup. The title (which I sing in the last line of the chorus) is a reference to a quote by the comedian Lenny Bruce where he addresses his own heroin addiction: “I'll die young, but it's like kissing God.” (lyrics)
The basis of this song is a droning piano loop. I constructed the loop using a very short sample from the middle of a piano piece, cutting it in many small pieces and pasting them together in different orders until I got something that I thought was rhythmically pleasing. (The clicks between snippets are left in on purpose.) The original sample is played in its entirety at the very beginning of the song, putting piano in the listener's mind right away. The same sample appears drenched in reverb after the first chorus, providing ambience and again reinforcing the piano-based feel of the song.
The loop is accompanied by layered and panned acoustic guitar strums and a complex, heavily distorted breakbeat. To make the breakbeat, I cut each individual drum sound out of a live drum performance and resequenced those sounds to make a new set of drum patterns. (The drums were originally played by Brian Sulpizio as part of a recording session for my pop band Starlister in March 2004.)
To provide dynamic shifts, I decided to make the choruses very loud in this song. The first chorus is quiet in order to establish the chorus melody and lyrics, but all other choruses include multiple layers of distorted electric guitar and bass guitar, and the transitions into those choruses are designed to make the chorus sound like a blast of sound. The electric guitars are tuned and played in order to make them sound slightly dissonant.
The vocals are recorded in multiple layers in the choruses, and they are run through a tube preamp circuit in order to provide warmth and slight distortion. I tried to sound delicate and avoid screaming; even though I thought that screaming might fit at times, the word "screaming" gets a very low score.
The bridge (which is placed in an unexpected position directly after the middle verse) has a surprisingly different arrangement from the rest of the song, thanks to performances by Deirdre Kelly (cello) and Sarah Powers (violin). This is probably the most unpredictable part of the song, both with regards to placement and arrangement.
Finally, near the end of the song, there are ethereal filtering and echoing effects on an extra snare drum, used to add a sense of ambience to the third verse and outro. This same effect is used on the repetitions of the piano loop after the first chorus. I also add another distorted guitar playing higher-register chords and a higher-pitched vocal harmony to the last repetition of the chorus, which makes it the loudest part of the song and makes it sound a bit more frantic.
The second song: “I'm Already Dead”
an mp3 of “I'm Already Dead”, final mix
In this song, I attempt to convey a feeling that I had once near the end of a particularly bad relationship. Things had gotten so negative and stagnant between the two of us that I wasn't able to function anymore, hence “already dead.” I felt completely unhappy, but unable to leave. I doubted myself as a person (“I'm not worth anything to anyone”) and couldn't see our relationship for what it was because I was too lost in thinking about it (“I'm trying to see the waves I'm drowning in cause I hear they're beautiful”). (lyrics)
The music is based around a set of samples from a couple of string quartets by one of my favorite composers. In the verse, the material is heavily effected and almost resembles horns at times. The string samples in the chorus (16th note rhythmic chords) are not processed, but they sound like echoes because of the way they're played. Those who are familiar with the original piece might be a little shocked at the new context of these samples—they sound surprisingly natural.
The string samples are joined by programmed two-step beats and very low-pitched synthesized bass tones. The style of the drums and bass reference a genre of uptempo, danceable music called “UK garage,” which I'm very fond of. (The bass is loud, but the notes are so low that they're inaudible on some stereo systems—a common trait of UK garage tracks.)
The vocals are layered onto several tracks at once. I used falsetto and head voice to double the main vocal melody, trying to make my voice sound higher than it actually is (high voices get higher scores), and I panned vocal tracks hard left and right to create a sense of space. I recorded extra tracks of whispered vocals on the second verse and the breakdown. I put the vocals through a tube preamp circuit to create a sense of warmth and add distortion, and ran the main vocal line through a digital delay effect to create echoes and ambience.
Guitar tracks appear at the beginning of the second verse, and I process multiple layers of guitars in several different ways in an attempt to make them sound lush. During the second verse, the guitar is played backwards and forwards at the same time, panned to opposite speakers. The second chorus guitar is processed with a flanger/phaser effect, which makes a swirling sound. Also, the coda features reverse reverb on the guitars, which is a way of creating a blurred, shimmering sound by reversing a waveform, applying reverb, and then reversing it again.
There are three unexpected dynamic shifts in this song: the first one is a short, quiet pause between the first verse and the first chorus, the second one is the suddenly quiet music in the beginning of the second verse, and the third one is the very loud distorted coda which happens after the breakdown section near the end of the song and features a new, distorted drum loop as well.
Two sections introduce new vocal melodies in unexpected places: one is in the first half of the second verse, and the other is during the breakdown section. The breakdown section includes lyrical and melodic content quoted from a catchy song by an 80s British pop band.
Lastly, the alternate hi-hat pattern that I use in the second chorus is provided to make the song ever so slightly louder. This is a bit predictable, but it makes the song more dynamic and helps distinguish the second chorus from the first.
5. the word lists ; index ; 7. conclusion