The most fun Music Technology class I’m taking this semester is Advanced Audio Production with Paul Geluso. A major component of the class is learning how to listen analytically, and to that end, we were assigned to pick a song and do an exhaustive study of its sonic qualities. We used methods from William Moylan’s book The Art of Recording: Understanding and Crafting the Mix. I chose “Tightrope” by Janelle Monáe featuring Big Boi.
Here’s the slideshow version of my analysis. The full spiel follows.
I listened to “Tightrope” about nine thousand times while performing this analysis. It’s a testament to the song’s musical and sonic quality that I wasn’t remotely tired of it after all that scrutiny. It just kept offering up more and more layers, more and more nuances, with each repeated listen. There’s no better definition of a classic song than that. It makes me wonder what might be buried in other songs that I like but haven’t scrutinized this deeply.
“Tightrope” was released on February 11, 2010 on Wondaland Arts Society/Bad Boy Records. Its length is 4:23, which is long for a pop song. The tempo is 83.81 beats per minute, surprisingly laid back given how lively the groove is.
The music and lyrics are by Nathaniel Irvin III, Charles Joseph II, Antwan “Big Boi” Patton and Janelle Monáe. Backing vocal arrangements are by Roman GianArthur. Janelle Monáe sings the lead vocals. Big Boi raps and does other interjections. Backing vocals are by Janelle Monáe, Nate Wonder and Chuck Lightning. Nate Wonder also plays bass and Moog synth. Kellindo Parker plays guitar and ukelele. The strings are performed by The Wondaland ArchOrchestra (Alexander Page on violin and viola and Grace Shim on cello) and arranged by Nate Wonder. Horns are performed by Hornz Unlimited, with arrangement by Nate Wonder, Jason Freeman and Jerry Freeman. Wolfmaster Z plays drums, and DJ Cutmaster Swiff does turntable scratching.
The track was produced by Nate “Rocket” Wonder, Chuck Lightning and Janelle Monáe. It was recorded by Control Z and Roman GianArthur at Wondaland Recording Studios in Atlanta, with additional engineering by Damien Lewis. The mix is by Phil Tan at Soapbox Studios and Christopher Carmouche at Farmhouse Studios, and the mastering is by Larry Anthony at COS Mastering, all also located in Atlanta. Professor Geluso commends Janelle Monáe’s team for keeping it local.
At the most basic level, there are really only two sections, the verse and chorus. The verse is F dorian for eight bars. The chorus is eight bars long too, but is harmonically livelier: two bars of G minor, two bars of Bb minor, two bars of F minor, and then, rather oddly, two bars of D minor. That last chord makes sense setting up the G minor, but in the key of F minor it’s strange since it has that A natural in it. Anyway, all of the other sections use one of these forms or the other. The intro and bridge are each half a verse. The rap section, turntable break and outtro are choruses. The horn break is the second half of the verse left over from the bridge, followed by the chorus. It’s simple stuff, but the arrangement is so lively and varied that it makes the song feel plenty eventful.
The drums are the first thing to jump out at you. They have a tribal sound due to the complete absence of snare and cymbals. The drum kit is just kick, rim, and toms. There’s very little reverb on the kit, as is standard contemporary R&B practice. The kick drum is doubled by an 808 kick, which gives a hugely fat low end. Tambourine fills the role that cymbals normally would. It’s double-trackled and looped, but with a loose feel within the samples. There are also a couple of layers of congas that blend tonally with the drum kit, panned out across the stereo image to create the hyperreal sensation of a space full of percussionists. Handclaps fill the role of a snare; their multiple layering and slight reverb creates a feeling of being closely surrounded.
Janelle Monáe’s lead vocal is recorded beautifully, with very little compression or obvious pitch correction. In keeping with contemporary practice, there’s no reverb on her voice at all. Instead, there’s a single-tap digital delay passing through what sounds like an octave harmonizer. The delay gives a spatial quality without the diffusing effect that reverb has. The backing vocals have a variety of sounds. Some have a classic gospel choir feel. Some are EQed strangely to bring out nasal highs, as in the bridge. Later in the song, backing vocals are built from samples of the verses and rap. Big Boi’s rap is multiply layered — it sounds like he’s doubled himself very precisely, which is no mean feat. His sampled backing interjections are delayed and have an overtly electronic sound, again very much in keeping with current production style.
Kellindo Parker is a virtuoso guitarist, but here his presence is minimal, playing short percussive chord stabs on a few sections. His sound is bone-dry. The bass is rich and natural-sounding, probably plugged straight into the board and doubled. The bassline is the hookiest little meme in the song. The horns are mixed oddly, with extreme reverb that makes them sound as if they’re in a distant and enormous room. I guess they were going for a classic Motown vibe, but I think they overdid it on the diffusion. The strings, on the other hand, are mixed beautifully. They’re so compressed as to be artificial-sounding, with an effective large reverb. The ukelele is percussive and bright, and it’s doubled, with one copy panned hard left, the other hard right. The little Moog wobble is a pure sine tone with no particular spatial location; it seems to just exist. Finally, the turntable is high-pitched, clean, and tonal. It wanders rapidly back and forth across the stereo image, implying a dreamlike atmosphere more than a physical space.
Dynamics and intensity
This graph shows the loudness of each part of the track. RDL stands for “reference dynamic level,” which in this case just means “a moderately loud volume.” You can see the abrupt breaks punctuating the ends of some sections. Hip-hop doesn’t use drum fills to punctuate sections and phrases, it uses silence. Janelle Monáe isn’t a hip-hop artist per se, but her sonic sensibility is clearly informed by her work with OutKast. The dynamics get really lively during the horn and scratch breaks, where both the instruments and vocals jump from quiet to loud and back dramatically.
The graph below compares the musical balance (loudness) to the performance intensity of each instrument or voice. The two don’t have to be related; a sound can be mixed to any level regardless of how hard the musician was playing or singing. For example, the horn players blast away pretty intensely for the most part but they’re mostly placed far back in the mix.
It’s interesting to note that while Janelle Monáe’s performance intensity varies quite a bit, the actual level of her voice stays constant. She belts most of the song in full voice, with a hair-raising Michael Jackson scream at one point, but at the end is singing at a conversationally quiet level. Either she has amazing microphone technique, or they mixed her very carefully, or (most likely) both. If you look at the waveform of the acapella track, you can see how steady its amplitude is throughout the song.
Imaging and environment
This graph shows the stereo image. The vocals, bass and most of the drums run dead center. The congas fan out widely to the sides, as do the horns and strings.
Here’s the perceived distance from the listener of each sound source. The reverb on the horns, strings and backing vocals places them very far away. The bone-dry bass puts it right in the listener’s face. The distance of the lead vocals is hard to resolve; they have no reverb, indicating closeness, but the delay on Janelle Monáe and Big Boi’s doubling give an otherworldly quality.
The verse melody is extremely simple. Janelle Monáe sits on Eb, scale degree b7, for almost the entire time, jumping up to the tonic F occasionally. The melody borrows liberally from James Brown’s “(Get Up I Fell Like Bein’ A) Sex Machine.” When Janelle Monáe sings “gettin’ funky on the scene” you can neatly substitute “get on the scene like a lovin’ machine.” The song apparently started out with an even stronger James Brown influence before they dialed it back a bit.
The chorus melody suddenly expands out to span an octave and a half, with dramatic swoops and dives.
Customarily, we don’t think of rap as having melody, but Melodyne detected a very clear tune to Big Boi’s verse. His singsong delivery actually covers an entire octave. Not every emcee has the lilt that Big Boi has, but this makes me want to run some more rap vocals through Melodyne to see what it finds.
By the way, Big Boi deserves some kind of prize for rhyming “NASDAQ” with “ass crack.”
The turntable break
The most interesting part of “Tightrope” comes after the three and a half minute mark, when a conventional pop song would be over. The turntable break plunges you into a whole new sonic landscape — Professor Geluso described it as a whole new song starting. This is where the eccentric sounds take over — the Moog, the ukelele, the sampled backing vocals. I chose this section to analyze in more depth.
Here are the pitch ranges and densities of the sound sources in this section. As you can see, the drums span a very wide range of pitches, but don’t hit very many pitches within that range. By contrast, the turntable covers a narrower range but hits just about every pitch within that range.
The lead vocal melody in this section contains the emotional peak of the entire tune, Janelle Monáe’s Michael Jackson scream in measure 115. She hits F5, the highest pitch in the entire melody, and then falls off an octave and a half. People in class who were hearing the song for the first time thought it was actually a sample of Michael Jackson himself. (Janelle Monáe does sample MJ elsewhere; the song “Locked Inside” interpolates the opening drum fill from “Rock With You” about fifteen seconds in.)
The graph below shows the frequency spectrum of Janelle Monáe’s and Big Boi’s lead vocals in the turntable break. The lowest frequencies are the fundamentals of their melody notes, but the harmonics of both their voices extend all the way up to the limit of our hearing. The voice is an incredibly rich blend of harmonics, especially when it’s a strong and distinctive one like these two have.
The final two graphs dive deeper into the sonic qualities of the Michael Jackson scream. It really has two components, the scream itself and the delayed copy, which has a slightly different spectral envelope.
About the graphics
The charts were made using Omnigraffle and Excel. The waveforms are screencaps of Ableton Live, and the melodies are screencaps of Melodyne.