A recent “sponsored feature” on Gamasutra focused on a new online music and sound library for games. An online sound effect library seems like a good thing. I’ve used a couple of those myself for personal projects where I didn’t want to spend the time making my own sounds. But a music library? No, this is not wise for games.
Music libraries certainly have their place. If you need something for a short corporate or personal video, something for your customers to listen to while on hold, or something for various other background type uses you can go grab something that you like and plug it into your project.
But a game is a different beast. How could you possibly hope to capture the character and essence of your unique game with a piece of music created only to fill some abstract “mood”? You built your game from the ground up with a certain style in mind. You hired artists to singlemindedly follow that style. Then you’re going to ruin all that stylistic cohesion by plugging in a random piece of music? Nahhh, you’ve got to bite the bullet and bring an honest-to goodness composer on board to complete your style. Sonically and visually, you’ve got to keep it all going in the same direction.
So the overall visual and audio coherence of your game is crucial. But coherence among all the pieces of music is also just as important. Grabbing random pieces of music to plug into different levels just creates a mismatched mess of an audio experience. Like I said in my GravRally post-mortem post from last month, using different pieces of music from a library within one game is like having anime, line-drawings, and Cubism as your art styles from level to level. Yes, that could be the point of your game but it most likely is not.
You can of course have different styles of music within your game. But wouldn’t it be nice to have bits of melody wander from piece to piece? Players may recognize that orchestral violin solo melody as it is being played by the muted trumpet later on in the jazz combo level. Or they might not but, ahh, they will feel the game pull together into a tight, cohesive, and satisfying experience!
You’re not going to get that from a music library. You will only get disjointed pieces that only tangentially resemble the style of your game and have little to no chance of resembling the other library pieces you cobble together. This isn’t just coming from me as a composer, but as a fan of games. I have not seen one game successfully use music from a music library.
And anyway, wouldn’t it suck to hear your game’s theme tune used in some online ad for Viagra?
To kick things off I thought I’d steal a post from 2004 that I posted on GarageGames. It deals with musical coherence in a game and gets into some nitty gritty details of the now-defunct game GravRally. Too bad that game never came out. It was FUN. Anyway, step into my time machine and enjoy…
(Oct 20, 2004)First off… IGC was great! Next year, gotta get down there earlier. Seeing the 21-6 guys once a year is definitely not enough, though. Some excellent games are coming. What a great time for indies. Now the meat of the post…
I thought it might be interesting to sit down and look at the music I produced over the past 16 months for GravRally, the futuristic racer being made by 21-6 Productions. Specifically, I wanted to analyze my music and see how I attempted to create a coherent whole out of the seven pieces I made for the game. That may sound weird coming from the composer but, while I created the music, a lot of the connections were made intuitively- by “feel”. Analysis certainly goes on while I compose, but it’s really just a part of the compositional process, and I forget a lot of what I was thinking about while I composed the piece. So part of the reason for this analysis is simply to remind myself what the heck I was thinking about at the time, consciously or unconsciously. Structure within pieces and among the pieces is very important to me so it was mostly done consciously, but I thought it would be useful for me (and I hope other composers and game developers) to dig a little into the pieces and see where the connections are.
There are seven pieces that will go into the game if the download size allows it: one menu piece and a piece for each of the 6 racing locations. I will go through the pieces chronologically (earliest to newest produced). Here’s a chart of the connections (download the pieces here):
You’ll notice everything springs from Metro and Menu so I won’t go into those too much. The main link from Metro to the rest of the pieces is its main melody (played on guitar in Metro by 21-6′s own Justin Mette, who I hope you all got to see play live at the last two IGC’s!). It is as follows (starting in Metro at 0:17):
The main link from Menu to the rest of the pieces is its opening motive, played originally in Menu by the guitar. The M3 and m7 intervals are the most recognizable bits of this motive, especially the m7 :
One link between Menu and Metro is Metro’s bass line, which is played by the basses at 0:41 in Menu and then again later at 1:14. I thought it was important to have a link between the two “parent” pieces, if even only a small one. Though that might not have been necessary, it’s probably better that there is one. Now, let’s see how the two seeds from Menu and Metro plant themselves in the other five pieces.
The link to Metro is very subtle, relying almost completely on the +4 note and two notes surrounding it (P4 and P5) at the A melody. Those notes are hammered away by the melody, never explicity playing Metro’s melody, but instead referencing it by using those same scale notes, with +4 being the most noticeable.
The connection to Menu is almost equally as subtle, with the m7 interval being played at the Intro and then being alternated with the P4 interval in the A sections. The fact that the m7 interval is right at the beginning of the piece makes the connection to Menu even stronger, since it opens with the same interval (interrupted by the m3).
The transitions have a very obvious Metro melody. The link is strong there, though the rhythm has been changed a little bit and a few repeated notes added.
The Menu connection is a bit of a stretch. The chords being played in the A section follow the contour of the Menu motive. Going up, instead of M3 and m7 intervals, they are m3 and m6 intervals. Then it follows the contour back down instead of going back to the m3. To make it even more of a stretch, the m6 is displaced downward an octave. Hey, I said it was a stretch! But that’s ok, because the connection to Metro is so strong.
The Metro melody is fairly recognizable in section A, though it’s obscured a little bit by the different harmonic stuff happening underneath it. So it’s definitely there, but you have to listen a little closely. The “feel” of the melody I think is obvious.
The Menu motive is also there, though again obscured a little bit. The melody at section B brings back the characteristic m7, actually in both the melody and harmony. That M3 is also there, but only in the harmony. From the sketch:
The m6 in the melody is also a little link back (really little, but real nonetheless… you don’t always have to be hit over the head for there to be a connection) to the Refinery piece, adding a little more coherence to the entire collection.
Forest is the least connected of the bunch. Mostly because there is no Metro link. But the Menu link is definitely there. The melody at section B hits the m7 interval hard, while also passing through the M3 interval as illustrated:
So it’s hidden a bit with that A between the D and F#. But this is one of the reasons I love linking pieces together. It’s the same motivic material, but it achieves a completely different effect when played by an orchestra than when played by a guitar. It’s gone from a quiet (expectant?) Menu opening to a brave brass fanfare. Hmm, the D section sure sounds familiar…
Because the Forest piece had the least linkage back to the parent pieces, I really wanted to make some more explicit connections in the Power Plant piece. So you really can’t miss them here. The Metro melody can be heard very easily at section A” (with the great robo-monk backup singing). It can also be heard in section B and B’, though altered rhythmically.
The Menu motive is right out in the open, once again opening a piece and heard throughout (like in the transition to A”). The Power Plant music basically functions as glue for the entire group of pieces, helping bond the various pieces together. When you hear connections between pieces as explicitily as you do here, I think you naturally tend to listen for connections in other pieces.
I’d been wanting to go through these pieces for some time now. I was pretty sure I had tied the pieces together but I wanted to go through and see exactly how I had done it. This wasn’t an incredibly in-depth analysis, but it sufficed in showing how you can make pieces coherent in both obvious and not-so-obvious ways. There are tons of tricks composers use to do this. I used some of the more obvious methods, mostly sticking with rhythmic variations of the same melody. But some other fun ways include inversion, where you flip a melody’s notes upside-down (instead of going up a m7, you go down a m7) and retrograde motion (playing a melody backwards). I didn’t go into orchestration here, either, which is a very effective way to tie pieces together. The orchestral sections of Menu and Forest, for instance, are bonded by their orchestrations.
In closing, let me say I hope this will also dissuade some of you game developers from using music libraries. If you’re creating a unique game, why not create a unique soundtrack for it? For another, you will probably end up with an incoherent mishmash of music. While the music itself may be good, the incoherence between the tunes in the collection can’t help but create a disconnect between the game and the player. It would be like every level of a platformer being made by different art studios, with a different hero in each level. First, you’re Scooby Doo fighting in an anime world… next level, you’re Jackie Chan in an old black and white Mickey Mouse world. It’s much more fun to be consistent and watch how your character, or how your music, develops from level to level. A simple trick done all the time in movies is taking a character’s heroic melody and transforming it into a sad melody. Old trick, but effective. Anyway, that’s the perspective of this game composer. Until next time…