Sign Up For Email

Subscribe to our mailing list

Wednesday, January 21, 2015

Rap Music Analysis - Why I Do What I Do

Hey there! My name's Martin, it's nice to meet you. If I gave you a link to this post, it's probably because you criticized me on Reddit, or some other site, for something. You may have called me out for supposedly shamelessly promoting this website (in which case, go to bullet point 1); you may have said I was over-intellectualizing rap (please press 2); you may have made fun of a clearly white guy talking about an art form with African-American origins (press 3.) I want to talk about each of these criticisms one by one.

1.) "Martin, stop! You're shamelessly promoting your website!"

I humbly disagree. First, the proffering of this criticism insinuates that I somehow benefit in an untoward way from this site. Well, rest easy, because I don't. I don't make money off this site, even with the ads there; I make a fraction of a penny every month, even. This website doesn't even forward any of my professional goals; I am not using this website as a launching pad to pad my resume or something. The only payoff I get, a rather modest one, is seeing my page views go up, and thereby knowing that someone enjoyed my website. There is no bigger prize for me than checking my site analytics and seeing someone spent half an hour on my site viewing over a dozen articles. I sincerely hope you would not begrudge me this one advantage to my site, when everything else in my life is largely drudgery, haha.

2.) "Martin! You're just over-intellectualizing an art form whose very popularity relies on how accessible it is!"

Okay, fine. I agree to a certain extent. But no one is forcing you to look at rap this way. But I know for a certain fact that there is a non-negligible section of rap fans who do enjoy this kind of stuff. I know this, because I have almost 250,000 page views (Yeah, I'm kind of a big deal.) So if you don't like this stuff, skip it! Just know when you downvote it, you kill my soul. I spend a lot of time and energy on this, and like I said, my only payoff is having people enjoy it. If you downvote it because you don't think it belongs in the subreddit, fine, but if you're downvoting it because you object to my motivations, then I disagree.

Besides, I can't help it; this is the way — almost the only way — I think about rap. To not do this would be deny a part of who I am, which is totally mad bullshit. I try to balance the intellectual stuff with more straightforward articles, like my one on Pharoahe Monch. Neither am I claiming this is the only way to think about rap; I greatly appreciate outlets that are more general in their approach, like HipHopDX. But that's not how I do things.

3.) "Yo dude! You're a white guy talking about an African-American art form!"

Yes, I am. But I am also constantly paying respect to the forebears and predecessors who made my analysis possible. This doesn't make me a card-carrying member of pro-black unions, but after doing my due research, I sincerely think Iggy Azalea is racist. I think parodies of Hip Hop, like Bo Burnham's song here, are racist. I completely agree with Kanye West when he says, "Racism still alive / they just be concealing it," and that George Bush really, really, just didn't like black people, even if he himself didn't know it. I already knew everything that Q-Tip tweeted to Iggy Azalea, because I've read "Can't Stop Won't Stop," by Jeff Chang, which is, as the sub-title says, "A History Of The Hip Hop Generation."

Look man, what I'm saying is that my engagement with Hip Hop didn't stop at these articles, the Internet, or even the music. I read Chinua Achebe's book "Things Fall Apart" after I found the source of The Roots album by the same name. I'm probably one of the few people who can name an anti-apartheid activist besides Mandela because of the Tribe Called Quest song "Steve Biko." I read Malcolm X's autobiography twice after so many rappers mentioned him, like on Run-D.M.C.'s "Proud To Be Black."  I actually know who and what the Zulu nation is, because Mos Def pointed me back to Afrika Bambaataa.

This isn't to establish my cred with rap fans, or any demographics of the population, whatever the color of their skin might be. It's to show that I truly, genuinely love this music, and I'm not just piggybacking off it for some kind of Internet popularity (as groundbreakingly important as that is to some people.)

4.) Do you even know anything about rap? What makes you so qualified to talk about rap music?

Good question! The answer is that I've studied rap music for years, as a job, and have built up my knowledge over that time. If you think money talks, well, people have paid me to teach them how to be better rappers, and colleges have given me freelance work to transcribe rap rhythms.

Think of it this way. You know how you went to school, in high school or college or whatever, and studied math for hours and hours? You went to class multiple times each week for hours at a time, where someone who had studied math for a lot of their life then taught you what you knew. You did homework every day, took quizzes, took tests, and studied a lot. Thus, by the end of it, you knew a lot about math.

Well, that's exactly what I've done with music. It's my job. I don't doubt that you know a lot about rap music, and that you even love rap music. But it most likely isn't your job to know rap music. You didn't take tests on rap music. You didn't transcribe rap music, and then study it for hours at a time. You would never tell a lawyer, "Hey, you're not lawyering well!", if you hadn't been to law school. You wouldn't tell a psychologist, "Hey! You're doing it wrong!" So when someone who's studied music for 15 years of their life, plays piano, plays guitar, tells you something about rap music, don't be surprised if it turns out you might just be able to learn something from it.

So please, over all, I'm begging you to please think before you speak. Do your research on my articles, just like I've done on Hip Hop, so you get the full picture, not just some title and the first paragraph after browsing through.



Tuesday, January 20, 2015

Rap Analysis - Rap's Rhythms Transcribed

For the past few years, I've been analyzing rap. This has led to an exact transcriptions of rap's rhythms. They have led to rhythms which are, quite simply, amazing. I'd like the transcribe reddit to see this not just because I think it's cool, but also to discuss the philosophy of transcription. See the body of the reddit submission for that part. A note on my transcriptions:

There is no standard way to notate rap. Western music notation fails to account for non-semitonal pitch movements, and it's this failing that has led to some to disparage rap. (In fact, I've tried to rectify this by using a modified MIDI piano roll that can move in smaller amounts than semitones. I combined this notation with an Eminem song in this video here.) For this reason, I've notated the rhythms all on a single staff line, to also emphasize rap's percussive nature. The slurs, far from indicating articulation like legato, instead show where semantic meaning (sentences) stops and ends. Rhymes are capitalized. The clef is included not to indicate melodic range, but to emphasize the melodic nature of rap, which doesn't always happen.

Some of these rhythms, understandably, mean nothing to people who view them. That is, they can't be performed by people who see them, for instance, as by people who see more standard scores, with straightforward 16th notes, for instance. This informs my approach. I have notated the rhythms exactly; other people round them off, such as Kyle Adams in his "On Metric Techniques Of Flow." However, I don't believe rapper's rhythms are swung. They are too intricate for that; to reduce them with a legend would give no information beyond how many notes there are too a bar. They're not swung because they don't relate in a straightforward manner to a bass, as they do in a shuffle rhythm.

I've used noctuplet groupings not to reflect how rappers are counting them, but because it's easiest to notate them in this manner, because it affords so many different types of rhythmic subdivisions. This means I can use noctuplet groupings for every bar, which keeps the music nicer. One drawback is that, just reading this music, it's harder to appreciate that every rhythm is different. Game's sound nothing like DMX's, and DMX's sound nothing like Time Bomb, who doesn't sound like Hittman, etc.

I've used note rhythms not to indicate the length of how long the rapper pronounces the words for, but to make the music easier to read.

These rhythms are really unbelievable. I promise they're true though. Earl Sweatshirt's alone took me 20 hours. I also checked them by overlaying video of the actual recorded music being played over a video of Sibelius with the notation being played.

Without further ado, these are my transcriptions, with links to each song. If people ask, I can put up video of Sibelius playing these rhythms, overlaid with the actual music. The six songs are "Some L.A. Niggaz," by Dr. Dre, "Earl," by Earl Sweatshirt, "Aquemini," by OutKast, "Vomitspit," by MF DOOM, "How We Do," by Game, and "Who We Be," by DMX. I've also offered thoughts on their rhythms sometimes. Here we go!

"Some L.A. Niggaz," - Dr. Dre, from Chronic: 2001; notated below are verses from Hittman, King Tee, and Time Bomb; verses from Defari and Xzibit were omitted.

Notice how they almost without fail avoid the 3rd beat, which is noticeably missing in the backbeat. Notice how Hittman is constantly starting and stopping his sentences in different placs. Notice how the sentences run across the barlines, which rarely happens in rap. Notice King Tee's long pauses.

The Game - "How We Do" - complete first verse, and excerpts afterwards from him and 50 Cent

Notice how Game is constantly changing the final note to fall on the beat, and then to fall syncopated off it. Accent marks with < are on the beat; notes with a tenuto are off the beat. Notice how many times he repeats a 3-note rhythmic figure. Notice how this goes throughout his whole verse, and into other verses as well. Notice that rhymes on the vowel sound "-uh" are all we hear for the entire opening 16 bars. Notice his really short sentences, and lack of full sentences.

DMX - "Who We Be" - complete first verse

Notice how he's constantly repeating his 2-note rhythm that falls on the beat. Notice that he lists just nouns, not verbs. Notice how few notes there are. Notice how small his rhythmic subdivisions get sometimes. Notice how he falls behind the beat at places, and places tons of small rhythms to catch up to it.

Earl Sweatshirt - "Earl" - First Verse

Video Demonstration Of Rhythms Here

Notice how slow his rhythms are. Notice how lazy they are. Notice how long his phrases (sentences, indicated by slurs) are. Notice his crazily long rhymes. Notice how many different types of rhythmic subdivisions he uses. Notice his dancing around the downbeat constantly.

MF DOOM - "Vomitspit"

Video Demonstration Of Rhythms Here

Big Boi, with OutKast - "Aquemini" - His First And Second (Third Overall) Verses

Demonstration Of Rhythms Video Here

Notice how fast his rhythms are, especially in a song with this tempo.

Rap Analysis - Who Is More Versatile, Black Thought Or Waka Flocka Flame?

There are two main challenges in the writing about rap from a strictly musical point of view, and not a poetic one, for instance. One is keeping my argument mostly at the level where anyone, even non-musicians, can understand it. The second is making all of my arguments as quantitative as possible, and thereby most convincing. I have to do that because I’m writing about a strictly sonic phenomenon, and so it’s hard to demonstrate any of my claims, even when I’m just giving general descriptions. If I were talking about poetry or race relations, I could quote or transcribe song lyrics; I can't do that for the strictly musical elements of rap.

All of this is especially challenging when discussing a musician’s versatility, or lack thereof. However, some simple statistical analysis, equally understandable by the layman or laywoman, can bolster what might be the shortcomings of strictly music analysis in dealing with those 2 problems mentioned above.

In lots of my other articles, I've described what bars are. As a reminder, they’re the building blocks of musical time that are always repeated in a song, and always last the same amount of time. They're similar to how minutes are the building blocks of chronological time. However, bars are different from minutes because bars can last different amount of times between different songs. We need this so that some songs can be slow, and other songs can be fast.

I also mentioned that each bar is made up of 4 beats, just as every minute is made up of 60 seconds. As it turns out, the rate at which those beats come, when compared to a minute, can give us a measurement of how fast or slow a song is. The rate of how slowly or quickly beats come in a given song is called BPM, for “Beats Per Minute.” The lower the number of BPM is, the slower a song is. Conversely, the higher the number of BPM is, the quicker a song is. Songs can vary widely in their BPM, anywhere from that of John Cage’s piece “As Slow aS Possible,” which is to last 640 years in a certain performance in Germany, to the 184 beats per minute of Khachaturian’s “Sabre Dance.”

Rap songs, especially recently, have a much narrower range of speeds in which they appear. As we’ll see, this is usually somewhere between 60 and 120 BPM. However, this does not bespeak a closemindedness of rap’s music-makers; instead, it only emphasizes the importance of a unique and personal approach that must come from the rapper on every song.

Using an investigation into the speeds of the songs at which certain rappers perform, we can see who is more versatile in their ability to deliver their lyrics, and who is more narrowly focused.

Using music software, I calculated the speed of songs from different rappers’ discographies. In certain situations, I was able to use every official album from an artist; in others who had a smaller output, I was forced to limit the search to only their studio albums. However, this always resulted in a sizable data set of at least 36 songs.

When I think “versatility” in rap, there’s only one person who comes to mind: Black Thought, emcee for the rap group The Roots.[1]

The Roots were the artistic force behind the second rap album I ever owned, their 2002 album Phrenology.[2] Since then, they’ve dropped The Tipping Point (my personal favorite,) Game Theory, Rising Down, How I Got Over, Wake Up!, Betty Wright: The Movie, Undun, Wise Up Ghost, and …And Then You Shoot Your Cousin. During this time, they collaborated with everyone from Elvis Costello to John Legend to Betty Wright. They’ve also been the backing band for the TV program “Late Night with Jimmy Fallon.” So far, I’ve got a lot of evidence that Black Thought is a jack of all musical trades, and a master of all musical trades. And that’d be good enough for a lot of writers…

But not me, which is where my aforementioned promise of statistical funsies comes in.

As I mentioned before, I went ahead and calculated the speed of Black Thought’s songs for a huge portion of his oeuvre.[3] The results for Black Thought’s 126 songs are below, in ascending order:

Each of those values should be read as, “60.8 beats-per-minute,” "61.8 beats-per-minute,” and so on.

Now, two things are pretty clear from this:

1.) Black Thought has a huge catalogue, and

2.) Black Thought has rapped over songs with tons of different speeds. His lowest song has only 60.8 BPM (on “Boom!”, from The Tipping Point,), while his quickest has 117 BPM (on “Here I Come,” from Game Theory.)

But in that list format, all of that information remains largely intellectual, and doesn’t really hit home. Let’s put it in a graph form that’s much easier to understand, because it’s visual. That same info, in graph form, looks like this:

This is a little more helpful. We can see that at either end of the graph — all the way to the left, or all the way to the right — we start getting some outliers, which are points that aren’t very close to the main portion of the data.

But what’d be really helpful is a graph that described, in detail, where Black Thought’s speeds fell most often, and where they fell least often. That’s exactly what this next graph is: 

The above graph shows where Black Thought’s BPMs fall most often. The horizontal axis along the bottom of the graph shows the BPMs, while the vertical axis along the lefthand side shows how often Black Thought had a song with that BPM along on the bottom. For instance, go to the chart’s highest point in the middle, closest to the very top of the graph. This falls along the vertical “Frequency” axis at exactly 14 times, while it falls on the horizontal “BPM” axis around 93. This means that Black Thought had 14 songs with a BPM of 93. This is his most common BPM, and applies to songs as different as “Stay Cool,” “What They Do,” and “Ain’t Sayin Nothin’ New,” which all come from 3 different albums.

Meanwhile, there are some other BPMs for which Black Thought doesn’t have a single song. If you look at 66 BPM on one end of the graph, or 120 BPM on the other end of the graph, you’ll see that the graph’s line doesn’t rise at all above the horizontal axis, and so it’s value is “0,” which means “0 songs are at this speed.”

As we’ll see soon, Black Thought has a very wide range of BPMs. His slowest song is at 60.8 BPM, and his fastest at 117 BPM.[4] This gives him a BPM range of about 56.2 BPM. His song’s average BPM is 93.2, which I suspect is true for most rap nowadays. That number also fits in very well with the frequency distribution of his BPMs, since it is a number very nearby — 93 BPM — which is the most frequent in Black Thought’s musical speeds.

If I suspected Black Thought to be extremely versatile, then I suspected another rapper, Waka Flocka, to be more limited in his musical approaches. If I applied the same operations to his rap that I just did for Black Thought, could I back this up with empirical proof as well?

Waka Flocka has a much more limited discography than Black Thought, so he has only 36 songs over his 2 official, major record label studio albums, Flockaveli (2010) and Triple F Life: Friends, Fans, and Family (2012.) The BPMs of all of these 36 songs are below, in ascending order:

Even in simple list form, some differences between Black Thought’s musical speeds of Waka’s speeds immediately stands out. For one thing, Waka Flocka’s speeds are more concentrated at the lower end of the BPM speed; 29 of the songs fall between 60 and 70 BPM. (For the musicians: double-time tempos were reduced to a straightforward BPM.) We can also see that Flocka’s slowest song, at 60 BPM, is slightly slower than Black Thought’s lower limit of 60.8. Additionally, Flocka’s fastest song, at 85 BPM, doesn’t come close to Black Thought’s comparatively breakneck speed of 117 BPM. So while Black Thought’s range of speeds is 56.2, as we said before, Waka Flocka’s range is only 25.

This is all represented visually below:

As we can see, Waka Flocka’s BPM speeds are mostly all the way on the left, towards the slower and lower end of the spectrum. Meanwhile, Black Thought’s speeds were more equally spread out.

We can, in fact, combine the frequency distribution graphs of both Black Thought and Waka Flocka to compare them visually:

We see that Black Thought’s output has a wider spread of speeds, as well as more songs at more different speeds than Waka Flocka’s. Waka Flocka is also more consistent in his choice of musical speeds; his most chosen speed was chosen 15 times, while Black Thought’s most chosen speed was chosen 14 times.

These statistics have already yielded some great results in describing the differences between rappers. Simply put, Black Thought raps over quicker songs, while Waka Flocka raps over slower songs. However, we can also calculate how much variation there is in each rapper’s output by talking about each data set’s standard deviation.

As the Encyclopedia Brittanica says:

“Standard deviation, in statistics, [is] a measure of the variability of any set of numerical values about their arithmetic mean (average.)”[5]

If a set of numerical values has a high standard deviation, the values are very spread out; if it has a low standard deviation, the values are grouped more closely to each other.

We want to use standard deviation because it is, in this instance, a measurement of a rapper’s versatility. That’s because the set for which we’re finding the standard deviation is the set we’ve been talking about this entire section so far: the speed of songs. A higher standard deviation means a wider spread of points, which means a wider spread of song tempos, which means a more versatile rapper, because they can rap over a greater variety of musical speeds. Get it?

How do you think the standard deviation for Black Thought’s BPMs and the standard deviation for Waka Flocka’s BPMs compare? Who will have the higher standard deviation, and, thus, the greater amount of versatility?

In fact, the standard deviation for Flocka’s BPM data set is…4.77

And Black Thought’s is…9.16, much greater than Flocka’s 4.77.

This quantitatively confirms the qualitative assertion with which I began this chapter: that Black Thought is a very versatile rapper. So WF might be a great rapper, but it wouldn’t be because of his versatility.


[1] And not just because he too has the good luck to be from Philadelphia.

[2] Will Smith’s Big Willie Style was the first. Obviously.

[3] A rapper who’s been in it since at least 1992 is bound to have a huge amount of recordings, so to gather every song was basically unworkable.

[4] Double-time songs in these charts are represented in straight time, at their “true” speed, without double counting quarters.