Music in The Legend of Zelda: Ocarina of Time, part one
April 6, 2008
[Part one | Part two | Part three | Part four | Part five | Part six]
Koji Kondo on his experience writing the music for Ocarina of Time:
I had to create all of those memorable tunes with only five tones of the classic do-re-mi scale. Specifically: re, fa, la, and ti (and the higher-scale re). Since each of those songs, like Zelda’s Lullaby or Epona’s Song, had a particular theme, it was quite challenging, but I think it all felt really natural in the end. Then as soon as I was finished with those Ocarina songs, I had to create even more for Majora’s Mask—I got a lot of milage out of just five tones!
A lot of milage indeed. In this series of posts, I’m going to try to see what he’s up to. If you’re up for a little musical excursion, read on.
There are twelve songs you learn in Ocarina of Time (not counting the improvisatory “Scarecrow’s Song”), and all of them are ostensibly based on the pitches D, F, A, and B, and D an octave above.1
Why ostensibly? Because although Kondo laments the four-pitch limitation, he has more freedom than it might initially seem because of how the game works.
To have Link perform a song, you take out your ocarina and play the beginning of the melody — a motive five to eight notes long, depending on the piece — using the controller. Playing this section, which I’ll call the trigger motive, causes the game to complete the rest of the tune automatically with what I’ll call the answer motive.
The upshot is that Kondo is only limited to four notes for the beginnings of the ocarina pieces. This is a significant distinction.
Take a look at “Sun’s Song,” transcribed below. I’ve put a double bar line between the trigger and answer motives.
Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.
As you can see, the trigger motive is constrained by the four-note palette, but the quick run in the second part has other pitches. After the player triggers the melody, the limit is removed, and Kondo has more compositional flexibility.
Why am I making a big deal about this? Well, “Sun’s Song” is an edge case; it’s the only piece you learn to play that never gets harmonized. Every other song is either filled out with orchestration when you trigger it, or is heard elsewhere in the game as part of the background music.
In other words, these melodies are not only written to be ocarina solos — there are harmonic considerations too. I’d like to suggest that that the the tonal restriction Kondo talks about influences how the pieces are written.
Let’s look at “Song of Storms.” Here’s a transcription of the piece as Link plays it, with a double bar line as before:
Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.
Even without any orchestration there’s not much harmonic wiggle room here. Kondo uses the octave Ds to firmly establish the tonic, and the flatted third (F) in the trigger motive sets up the unambiguously minor answer motive.
Here’s the Kakariko Village Windmill background music, which features the “Song of Storms” melody:
Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.
The harmony indicates that we’re in the Dorian mode, which is the mode most strongly suggested by the available pitches; there’s a complete minor triad (D, F, A) and the characteristic raised sixth (B♮). In layman’s terms, Kondo has found the easiest mode to compose in with these notes and exploited it. “Song of Storms,” then, is one of the most basic ocarina pieces; Kondo has allowed the pitches to dictate the trigger melody, and the consequent phrase in the Kakariko Village Windmill theme doesn’t break any new ground.
Everything can’t be as simple as the Dorian mode, though, and things get more interesting when Kondo breaks out into other harmonic areas. Let’s look at “Epona’s Song.” Here’s the solo ocarina version:
Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.
Out of context, this is a harmonically vague phrase. The tonic is most likely D, but not necessarily; even so, it’s impossible to pin down a key using only scale degrees one, five, and six. Without the third, we can’t even definitively say if it’s major or minor.
More importantly, note how Kondo continues to use the same pitches in the answer motive, even though other options are available. In other words, the piece is composed — intentionally, perhaps — so that you can’t discern the key from the ocarina solo alone.
Now let’s turn to the Lon Lon Ranch background music, which uses the “Epona’s Song” melody:
Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.
The harmony reveals that the tune is in D major, but as we saw above, we couldn’t determine that from the first four bars of the melody. My theory is that this delay is intentional.
Because there’s no obvious key to the “ocarina solo” portion of the song, the harmonic content of the first phrase unfolds relatively slowly. (Indeed, if we were to look at the melody alone, the key doesn’t become completely obvious until the eighth bar.) This protracted development creates a richer, more nuanced melody than “Song of Storms,” which repeats a single phrase ad nauseam. That melody, in turn, provides a basis for the modulation into F major and its new musical idea.
And thus Kondo’s stated difficulty with the limited pitch material ended up helping him out. Restriction breeds creativity.
I hope you all could make some sense of this. I’ll have a few more posts about music from Ocarina of Time over the next couple of weeks; in the meantime, I look forward to your responses.
[Part one | Part two | Part three | Part four | Part five | Part six]
- Looking at those solfège syllables, it seems like Kondo’s a fixed do kind of guy.




13 comments
Very nice.
I agree with your comment that “Restriction breeds creativity”, that has certainly been the case in my personal experience.
I do want to question, however, your characterisation of the melody of “Eponas’ Song” as “richer” and more nuanced. While I think that there is certainly more variation in the melody as compared to the “Song of Storms”, I think that your own point, that restriction breeds creativity, works against you here. I find the “Song of Storms” to be much more interesting (and at the same time much more frustrating when played Ad Nauseum while sitting on that spinning wheel thing I seem to remember).
That said, I think the two are really different genres of song – in the sense that Epona’s is almost like a ballad and the Song of Storms is more like a Jig or Reel, which explains the repeated motif over and over again. One of features of the Jig/Reel (forget which) was that it often did just play through the same musical material, but often adding or changing the instrumentation at certain points, which is exactly what happens in the Song of Storms (it starts to fade out just as the original melody returns on a different instrument).
I hope you don’t feel like I’ve taken you to task or anything, I *really* appreciate the effort you’ve put into this, just wanted to offer another opinion, and I hope you’ll do the same when I get around to analysing the music of Halo for my Thesis and start to put it up on my blog.
Just on a random side note, I somehow managed to get the Song of Storms clip to play BACKWARDS by some strange combination of black voodoo and Ctrl Left Arrow, and it sounded really cool – especially when I was tapping it and getting it to skip and glitch…. but then, I’ve always liked experimental electro so there you go.
by Ben Abraham on April 6, 2008 at 11:49 pm #
Hey Ben, thanks for commenting!
You make a good point about the genre distinction, and you’re absolutely right that “restriction breeds creativity” applies especially to “Song of Storms.” (Actually, if we take a broader view, it applies to Kondo’s work in general; after all, he composed some of his most iconic music with the technical limitations of the NES.)
Looking back at the post, I think I overstated my case by saying that “Epona’s Song” has a richer melody. I didn’t mean to make a value judgment about one song being better than the other; rather, I meant to draw a distinction between the way each melody unfolds. I’ll see if I can clarify a bit.
“Song of Storms” puts all its melodic and harmonic material on the table right away, during what I’ve termed the trigger and answer motives. This has the effect of firmly establishing both the minor tonality and the harmonic quality of the piece; the fact that it doesn’t really develop any further is, I think, a consequence of that.
The melody for “Epona’s Song,” on the other hand, takes a while to establish itself. As I argue in the post, it’s harmonically ambiguous through the trigger and answer motives because of the pitches Kondo chooses. I think this slow development gives the piece a chance to go somewhere else — F major, as it turns out — in a way that “Song of Storms” couldn’t, with its balls-to-the-wall melodic intensity.
At the same time, there are lots of other important factors to consider — such as genre, as you rightly point out — that affected Kondo’s decisions. My point here, which is admittedly labored, is just that composing solo pieces that must begin by using the same five pitches has a sort of butterfly effect on the end result. I’m not sure how valid that argument is anyway, but I hope it’s a little less muddled now. (Also, I want to go on record as saying that I love “Song of Storms” too. I’m especially a fan of that elision.)
Please do feel free to take me to task. Since there are probably only a couple of people reading who have a background in music, it’s great to get feedback. Also, I personally think it would be completely hilarious and awesome to get into a heated discussion about music theory as it relates to video games. :D
by Dan Bruno on April 7, 2008 at 12:23 am #
“balls-to-the-wall melodic intensity” Now THAT’S a saying you’d never get in another context.
Adressing your argument about how composing for 5 notes leads to a chaos theory effect on the end result:
I’m not sure I get the idea – are you saying that you think composing using such a restricted palette (and hence such a restricted / ambiguous tonal centre) results in… what, an unpredictable end composition when fully harmonised?
I get the distinct impression that, especially the ocarina solo’s, when harmonised, have a real ‘call and response’ feel, a common musical device, I’m sure you’ll agree – and I think you picked up on that by calling it the “answer motive”. I’m not really sure however if I’d be willing to say that /because/ Kondo composed only using 5 notes to kick start a musical piece, that it resulted in him going all over the shop harmonically. I suppose it encourages it more than, say, were he able to compose chromatic ocarina solo’s (Shred that Ocarina of Time! WOOOO!) – but then again, maybe he still would (but then at at least, he might not get right that slow development of the material, such as in Epona’s Song).
Does he ever say anywhere if he ever worked the process in reverse order? Composed the “whole song” before the Ocarina melody, because that *would* be convincing, if purely as a COUNTER-argument.
Curses! This is devolving into an Ontological argument! Quick, we need a Descartes quote, stat!
“Cogito, Ergo sum.”
Phew!
It reminds me of the example of trying to prove the existence of a certain object in a room or similar space. It’s generally accepted that its much easier to find the object in the room than it is to prove that it’s not in the room, because the latter means searching *everywhere* whereas the former just means finding it.
This discussion just reminded me of that – how do you “find” the smoking gun that proves your idea. I’m all out of directions and obviously don’t know the material as well as you do – your turn, then. =P
Finally, have you thought about submitting a (more complete) version of this series on the music of Koji Kondo to an online journal site?
C Theory comes to mind:
http://www.ctheory.net/static.aspx?id=Submission Guidelines
but perhaps more appropriate is:
http://www.gamestudies.org
They both takes unpublished articles and put them in their online journals. Both are peer reviewed though, so I have no idea how hard/easy it’d be to get through that process.
Cheers, Dan – it’s been fun so far – keep up the excellent work.
by Ben Abraham on April 7, 2008 at 1:17 am #
Yes Dan, Keep up the excellent work.
by Korkie on April 7, 2008 at 8:20 am #
Regarding chaos theory: I don’t mean that each piece individually is harmonically unpredictable, but rather that that the way they unfold is dependent on what Kondo does with the limited palette at the very beginning. Thus when he firmly establishes D minor tonality at the beginning of “Song of Storms,” the piece already has its course charted, so to speak; when the “Epona’s Song” melody begins without a clear harmonic context, the piece meanders more and has “room” to modulate. In chaos theory terms, there’s a high sensitivity to initial conditions. (“Regular” compositions based on simple motives often show the same effect, of course; see, e.g., Beethoven’s 5th. But hey, this is a games blog, so Kondo takes precedence here. Sorry, Ludwig.)
I can’t say I get a call and response sense from the pieces I’ve done so far. I would even go so far as to say that none of the pieces Link learns as a child — “Zelda’s Lullaby,” “Saria’s Song,” “Song of Time,” and the pieces mentioned in the post — are call and response; I have a hard time hearing, say, “Song of Storms” as two distinct phrases. I think the division between the trigger motive and answer motive is motivated by Kondo’s desire to use a larger set of pitches more than by any structural considerations.
However, the “warp” songs you learn as an adult — “Prelude of Light,” “Minuet of Forest,” and so on — most certainly are call and response; they literally feature an ocarina “call” from Link and a harp “response” from Sheik. I’ll probably give over an entire post to those pieces, because there are a few music theory gems in there.
I haven’t found anything from Kondo about his compositional process, but there is at least one piece that he had entirely pre-composed — “Zelda’s Lullaby,” which was already extant as a theme in A Link to the Past years before Ocarina of Time came out. I have a theory about that one too; I’m sure I’ve said this about a million times now, but I’ll be talking about it in a later post. ;)
As for the ontology discussion, I don’t know what I can say except that this is all just idle speculation on my part, so there might be no “proof” to find. Then again, a lot of music theory is ex post facto analysis of what a composer’s intention might have been, so I’m okay with that. :) Naturally, I’d need to have something more concrete than if I wanted to get this published somewhere, but I’m happy enough with the discussion it’s provoking right here. Thanks again for joining in!
by Dan Bruno on April 7, 2008 at 9:24 am #
As Martin Sheen playing President Bartlet in the West Wing said so famously:
“Post hoc ergo propter hoc”
From Wikipedia:
“Latin for “after this, therefore because of this”, is a logical fallacy (of the questionable cause variety) which states, “Since that event followed this one, that event must have been caused by this one.”"
I look forward to these promised posts. =)
by Ben Abraham on April 7, 2008 at 10:57 am #
Hah! Touché. :)
by Dan Bruno on April 7, 2008 at 11:05 am #
Ben keeps us honest, doesn’t he, Dan? :-)
I enjoyed this post even more than the last. Now that you’re in the Zelda/Kondo groove, I look forward to seeing where you go next in your series.
I’m wondering how you feel about the whole MIDI versus recorded orchestrated music question. It’s a no-brainer to me, but I was surprised to receive a couple of responses on my blog recently in defense of the MIDI approach to music in the Zelda series and the hope that it wouldn’t change. Is this just preciousness about the old games, or is there really something preferable about triggered music samples?
by Michael Abbott on April 7, 2008 at 2:52 pm #
I’m glad you’re enjoying the series, Michael. I’m glad, too, that you brought up the sequenced/orchestral music issue — I’m actually going to talk about sound technology in Ocarina in — wait for it — a later post. :)
by Dan Bruno on April 7, 2008 at 8:14 pm #
I think I know you.
by Johnny on April 18, 2008 at 8:02 pm #
Where from, Johnny?
by Dan Bruno on April 18, 2008 at 10:10 pm #
Theses songs are awesome, simple but great! I love the Kakariko Village Windmill background music. Keep up the great work!
by Raquel on July 25, 2008 at 7:23 pm #
Thanks, Raquel! I’m working on another Ocarina post, so stay tuned!
by Dan Bruno on July 28, 2008 at 5:01 pm #