In a previous post on this blog, I investigated an oddly-timed riff from Metallica’s “Master of Puppets” that is allegedly in 5/8 meter. This certainly isn’t the only song in which Metallica does really interesting things with their timing—meter changes and tempo changes are practically a basic element of the band’s style in their early albums, and their music would sound really different if they followed a metronome nailed to a single tempo for each song. But there are also several riffs like the one I analyzed from “Master of Puppets” that don’t easily fit into a recognizable meter. In this post, I’ll be looking at a riff that accompanies the guitar solo in their first original song “Hit the Lights.” This riff is usually transcribed with a 7/8 measure at the end, but I’ll be arguing that it actually never changes out of 4/4.

This riff begins at 2:35 in the album version of “Hit the Lights,” right after the third verse. After a few repetitions of the riff, it is transposed up a whole step and continues underneath the guitar solo. This riff is then repeated kind of like a vamp through to the end of the song.

0:00 Intro
0:36 First riff in rhythm guitar, other instruments slowly join
0:54 Verse 1
1:12 Chorus "Hit the Lights!" (which has it's own 7/8 meter change)
1:23 Lead Guitar Interlude I
1:29 Verse 2
1:47 Chorus
1:58 Lead Guitar Interlude II
2:05 Verse 3
2:22 Chorus
2:35 Single guitar introduces a new riff (see the transcription below)
2:58 Guitar Solo
3:55 Outro: Repeating only the last bar of the riff 3x, then holding a chord with lots of drum fills, until the last note.

In most transcriptions, especially the officially approved transcriptions of the guitar and drum parts to this album published by Cherry Lane Music Company, ((These two official transcription books are Metallica: Kill ‘Em All (Play It Like It Is) for guitar and Metallica: Kill ‘Em All (Drum Edition).)) this riff appears as three bars of 4/4 time followed by a single bar of 7/8 time. The 7/8 bar doesn’t have any instrument marking the eighth note, but just has a series of six notes in rhythmic unison between the two parts. The only departure from this unison in any repetition of this riff is that occasionally, the drummer Lars Ulrich plays what is called a flam or grace note, in which he hits the snare drum twice nearly simultaneously. Below, I’ve uploaded just one statement of this riff from the Cherry Lane guitar transcription, and two excerpts from their drum transcription to show the difference. I’ve put labels on the notes in this alleged 7/8 bar that I will be referring to later. I say “alleged” because, as I’ll explain, I think this rhythm is actually in 4/4.

In the guitar version of the riff, I’ve marked in purple “1,” “2,” and “3” for the beginning of the three 4/4 bars. Each of the notes in the last bar gets a letter, a through g, where g is the downbeat of the next repetition of the riff. This means that g and 1 are the same note, after the first statement of the riff. I used the same process to study the timing of this riff as I did the previous riff: I opened the .mp3 file in Audacity and placed labels any time I could identify one of the locations in the riff marked in purple in the guitar transcription above. There’s always some degree of imprecision with this method of studying microtiming, because frequently you can’t tell exactly when the beginning of a sound is. I tried to limit my markers to places where there was a snare drum or bass drum attack that had a clearly defined beginning point.

The next step was to review each of the markers I made in Audacity and copy the timing data into a spreadsheet. The empty boxes are places where there wasn’t a precise enough feature of the sound data to place a marker. If I were a real scientist I would have error margins for all of these timings, and I would use that information to limit the significant figures of my numbers. I’m not, but I believe the imprecisions that do exist aren’t strong enough to affect the conclusions I reach later.

The next step was to calculate the durations between each of these timings, which you can see in the chart below. You will notice that a few of these durations in the “1 span,” “2 span,” and “3 span” columns look obviously wrong because of the empty boxes in the chart above. I removed this data before moving on to the next stage of analysis.

There is already a striking result in this duration table that casts doubt on the accuracy of the Cherry Lane transcriptions I excerpted at the beginning of the article. Those transcriptions show the rhythm as three dotted eighth notes followed by three normal eighth notes. Looking at this notation, I expected the a, b, and c spans to be approximately 1.5 times longer than the d, e, and f spans. ((I say “approximately” because nobody is perfect at rhythm, and nobody should be. As many metal critics and music theory researchers have noticed, perfectly quantized rhythm tends to sound mechanical and inhuman.)) But if you look at the chart above, there is no clear categorical difference between the a, b, c, and the d, e, and f spans! Instead, each time the riff is played, there is a gradient of slowly changing durations, each one slightly smaller than the previous. A is the longest, then b is a little shorter, and c is a little shorter still, and so on, so that there’s no clear point at which anyone can tell the rhythm changes between eighth notes and dotted eighth notes. The 7/8 transcription of this rhythm in eighth notes and dotted eighth notes (with a duration pattern of 3+3+3+2+2+2) is at best a reductively simple approximation. ((One could make a more accurate transcription by adding or subtracting thirty-second notes, sixty-fourth notes, or even smaller values, but that would make these transcriptions difficult to read, and I think it’s unlikely that very many human performers have the precision to accurately perform such miniscule distinctions, especially the amateur or student audiences that these transcription books are marketed to. Highly-skilled professionals would probably just figure the rhythms out by ear instead of paying hard-earned money for a book.))

In the next stage of the analysis I wanted to compare the length of the supposed 7/8 bar with the length of the three 4/4 bars in each riff. In the chart below, the “7/8 span” is the difference between the g column and the a column in the same row of the timing chart. I couldn’t measure the length of the 4/4 bars directly because I was missing so many markers in the 2 and 3 columns, so I calculated the “avg 4/4 bar span” column by dividing the distance between the 1 marker and the a marker in each row by 3. The column “avg bar length” is just the length of the entire four measure riff divided by 4, but I don’t find this statistic particularly informative.

The crucial column to observe is “diff. btw. 7/8 and 4/4,” which is just the number in the second column minus the number in the first column. It is clear to see that in the first four statements of this riff, the 7/8 and 4/4 bars are almost exactly the same length—and in two cases, the 7/8 bar is actually longer! Beginning with the fifth bar, the duration of the supposed 7/8 bar actually becomes about an eighth note shorter than the 4/4 bars. For comparison, I calculated what the duration of an eighth note theoretically should be relative to the tempo at each moment by dividing the length of the 4/4 bar by 8. One reason I can think of for the sudden change in the duration of the alleged 7/8 bar in the fifth row is that this is when the drums begin playing a backbeat for the three 4/4 measures at the beginning of the riff. Whatever the reason, the length of this last bar of the riff is not uniformly consistent, and choosing either 7/8 or 4/4 only matches one part of the

I think the image below is a better transcription of this riff. This isn’t because it matches the timing of the durations a-g any better, or because it matches the overall duration of the “7/8 bar.” The reason I chose this transcription is that the rhythm it has (with a duration pattern of 3+3+3+3+2+2) is a common rhythm in heavy metal, rock music more generally, and even other kinds of pop music. ((Nicole Biamonte has written about this rhythm in her article “Formal Functions of Metrical Dissonance in Rock Music” in Music Theory Online (2014). Mark Butler talks in greater depth about the same type of rhythms, but in electronic dance music, in his first book Unlocking the Groove (2006).)) Metallica uses this kind of rhythm in a number of other songs (like “Master of Puppets”), but I don’t they’ve ever used the pattern 3+3+3+2+2+2 again.

But more significantly, even though there is no clear distinction between the duration categories of eighth and dotted eighth notes, there is one musical detail that matches my transcription better. In all of the repetitions of this riff beginning with the fifth one, the first four notes are played with simultaneous hits on the bass drum and crash cymbal, while the last two notes of the measure (e and f) are played on the snare drum. Also these last two notes, e and f, are closer in duration on average than any other pair of notes a through f. This evidence seems to indicate that these two notes e and f could belong to a different category than the preceding four notes.

Under my interpretation, the last measure of this riff is still in 4/4, and the gradual change of durations from a through f creates a feeling of rushing that is part of what makes this song so exciting. This feeling of speed amplifies the image the band was trying to create: that they were relentlessly fast, faster than any other metal band on the planet. This feeling of frenetic speed and ferocity is really what separates Metallica’s first album from the work of New Wave of British Heavy Metal bands like Diamond Head or Angel Witch that inspired Metallica in the first place. Metallica’s playing on this album may seem untutored and messy, but this feeling “rushed” timing contributes to the exuberant, wild character of their commercial debut.