Tales from the Trenches of Coin-Op Audio

Gama Network Presents:

Tales from the Trenches of Coin-Op Audio
By Chris Granner
Gamasutra
November, 18, 1999
URL: http://www.gamasutra.com/19991118/Granner_01.htm

Editor's note: This paper was originally published in the 1999 Game Developer's Conference proceedings.

Sound for a coin-op game has a great deal in common with sound for a home game or a console game. Regardless of target system or environment, sound takes a flat picture on a video screen and brings it to life in the player's world. No other game element immerses the player like sound does. (Just ask any sound designer!)

The main thing coin-op sound has in common with home game sound is that in either case, the presentation of sound, and the development of that presentation, takes a back seat to the play of the game itself. This is as it should be; we might want to be selling records, but the folks paying the bills are trying to sell games, and we're supposed to be helping them do that.

The world of coin-op game sound, however, is dramatically distinct from the world of game sound as it exists in the home or office. The most fundamental difference is in the way players pay for a coin-op game (50 cents or a dollar per play, versus 50 bucks up front for a home game). The kind of game that succeeds in a coin-op environment has a gratification curve that must be much less steep than for a home game, and the coin-op soundtrack must help to sell and enhance that gratification.

Combined with the need to support "instant" gratification is the fact that coin-op games are played in public places, and in many (most!) of those places the kind and amount of noise tolerated from a coin-op device is extensively restricted. In the case where "anything goes," the arcade or game room, the problem is that all the machines in the room are making a lot of noise, creating a din which dramatically reduces the effective signal-to-noise ratio of a soundtrack.

In this rather hostile environment, the sound designer must "choose wisely" which sound elements to emphasize. Elements that would provide a delicate or "deep" background or ambience behind foreground sound elements will be lost either to having the game turned almost all the way down, or to a public version of stereo wars.

These problems confront the sound designer of any type of coin-op game. Although we're mostly concerned with video games, we face many of the same problems creating soundtracks for pinball & redemption games. However, the nature of these devices is fundamentally different from a video game in ways that affect the soundtrack; we examine these differences briefly.

The most formidable challenge to sound designers of coin-op games is the sentences "They just turn them down anyway." Knowing when and how people started to say such things makes it possible for the sound designer to fight against this extremely demoralizing argument. This battle is fought on the battleground of audibility: by enhancing the effective audibility of our soundtracks, we enhance the importance of sound design to the entire project. So we have to improve at making "indispensable sound."

The Object of the Game

The idea is to enhance gameplay. This is true for any game soundtrack, whether coin-op, PC or console. Here are some thoughts about how to be effective in this context, whichever platform or environment we're designing for.

We're artists. We want to be recognized as such. We leave a piece of ourselves in every sound we make (or should anyway). We understand sound and hear sound on a level that's beyond what we can expect from most listeners. We sweat the details.

We can't allow this concern for the minutiae of audio to get in the way of gameplay. We must simply pick our spots. There are many points in every game where a set piece or little movie is presented that the player is just supposed to watch. During these moments is when it's appropriate to milk the subtlety. But keep in mind that the player can "tap" out of such moments in most games. So our awesome movie soundtrack will be interrupted by some joe with a joystick. The joe pays the bills, however…so if he want to go on, let's let him and not fret his lost opportunity to hear our magnum opus. The moral: we can still do great work in a supporting role; we don't have to be the stars.

Speaking of awesome - since we're working in an interactive medium, the final audio result is somewhat indeterminate, which makes it difficult to compose and shape long passages predictably. But the following rule of thumb is applicable: a graceful large shape is made up of lots of graceful small shapes. So even the smallest foley element needs to have color, profile, and character. This is the best way to insure that the indeterminate result sounds as good as we can make it. This does NOT mean that every element needs to be as loud as possible. There are other reasons we want to make everything as loud as possible which we'll explore presently.

The size of a game soundtrack project, in terms of man-hours, is dwarfed by the size of the game development itself. This means that, if we're involved with a game project from the beginning, we'll be placing elements into the game that the developers, and especially the programmers, are going to be listening to from now 'till the end of the project. At the very least, we run the risk that they'll get sick of something. More insidious is the creep in intensity; the new death blow must be bigger, louder, more intense than the others. It is the job of the sound designer to balance these requests in such a way as to keep the developers happy. In a strange sense, the soundtrack represents the mood of the project, and this places the sound designer in the unenviable role of Master of Attitude Adjustment. The best way to serve this purpose is to bring in new stuff regularly. This minimizes the chance of the whole soundtrack becoming stale.

Our reality is that we're making interactive games. We're working as part of a team of designers, programmers and artists to produce a comprehensive interactive experience - emphasis on "interactive." We must find a way, as sound artists, to put heart and soul into a piece of this larger opus, and to make that piece (the soundtrack) do all it can to help the game reach the player, so the player hears, feels and understands the world he's being immersed in. It's not the same as reaching a listener directly with a piece of music or a great solo. But it's tremendously rewarding to read in a review something like "there's a great moment when you pick the lock - the door opens with a grinding snarl, it slams against the doorstop like a kick in your gut, and you feel the door shiver right up & down your spine."

Follow the Money

All the preceding is applicable to sound design in general; let's focus now on some of the challenges characteristic of coin-op sound design. These characteristics can be traced to the way players pay for coin-op games. Players buy PC or console games at a store for anywhere from $20 to $90, making the entire outlay at once. Coin-op players pay by the game, 50 cents or a buck per game. This difference, although obvious, cannot be overstated.

The player with the home game under her arm has made an investment. She's now going to take it home and do everything she can to make that investment pay off. She'll install it and run it right away; she'll watch the now-obligatory introductory movies; she'll dig into the documentation or go online to find the solution to any problems she has with the install, or how to run it; she'll experiment with the controls; she'll forgive overly easy or boring stuff at the beginning if she finds cooler stuff at deeper levels; she'll literally spend hours doing anything she can to make that investment pay off in quality diversion & amusement because she's already made the whole investment. She has a fair amount to lose if the investment doesn't pay off, so she's willing to put in some time to maximize her chances of a good payoff.

The coin-op player makes that investment 2 quarters at a time. In most circumstances, whether he wins or loses, he'll be asked to invest another 2 quarters in about 3 minutes. So at the end of 3 minutes he has to evaluate whether this game is worth an additional investment. This means that he game has 3 minutes to seduce him. Put another way, the game must allow the player to feel sufficiently successful, to "obtain the amusement dividend," within the span of 3 minutes.

This requirement (3 minutes of gratification) eliminates many types of games from contenders to coin-op success. Your strategy game, your mystery adventure - forget it. In fact, anything that wants to move at a slow pace has a pretty poor record in a coin-op context. The games that have succeeded in the last several years, while spanning a pretty broad spectrum of styles, all have in common a large does of in-your-face adrenaline-based feedback. This, as we may have guessed, is where sound design comes in.

Save the dreamy noir music tracks and ballads for the game over screens. Our music will probably all end up running at ¼-note equals 120 or higher (even if we don't start there). As we discover later, if we can make our music feel fast and hot with long notes and lots of space, we'll be doing ourselves a favor; but it has to be hot, hot, hot!

Sound effects must literally explode. Our motto: Bigger than life isn't big enough. We want the player rocked back on his heels going "whoa!" as often as possible. This requirement steers us in the direction of things that have a very short attack time - even for things that don't in real life, like swooshes & yells. We are not so concerned with reality here; like the movies, powerful and compelling wins over accurate every time. In addition, each sound the player hears might be the last one he hears for a few seconds. So make sure each sound has a graceful ending. This is how we create graceful passages in an indeterminate sound context: make sure everything that happens is graceful.

Speech scripts for a coin-op game tend toward the apoplectic. Even when it's a quit moment, like putting in a golf game, the announcer has to have an expectant, "up-energy" urgency in his or her whisper. Urgency is the watchword, since the situation is urgent: we have 3 minutes!

Noise in a Public Place

There are obviously a lot of different kinds of public places, but for the purpose of this discussion, we can reduce our analysis to two types: the arcade or game room, and the "street location:" i.e. any other place we're likely to encounter a coin-op game. This includes bars, convenience stores, restaurant waiting areas, movie theater lobbies and laundromats; there are probably a few others but that list gives the proper flavor. What all of these locations have in common is that a coin-op game has been placed in a space that wasn't originally designed for games.

Since it is our first priority to animate what the player is doing, and since what the player of coin-op games is doing is largely the same as what he was doing one second ago, a primary characteristic of the coin-op sound experience could be classified as "repetitive." Many peoples' first experience with video game sounds came from Pong, Breakout, Space Invaders or Asteroids. "Beep" and "Boop" does not much understate the sophistication of these early coin-op sound efforts. If we put ourselves in the position of the people employed in such places, listening to repetitive beeps & boops, we can begin to appreciate the difficulties we continue to face today when we place a game in a street location.

The biggest problem with game sound in these locations is that the sounds they make, by and large, don't sound like they belong in this environment. Beep & boop don't sound like they belong in any environment; and although our sonic palette has expanded tremendously, any piece of digital audio or music that's repeated over & over will grate on the nerves of the innocent bystander (which, in these locations, includes everybody but the player). The result: unless the game soundtrack is very carefully designed, the location manager will have the sound turned way down, if not all the way off.

How can we counteract this tendency? By composing our soundtracks from elements we might expect to find in such environments. The great thing here is that these environments, whatever else is going on there, also include television and radio. This implies that, up to a point, the street location public will tolerate the sound of a game if it more or less resembles the sound from any of the other media present at that location. Incredible Technologies places Golden Tee Golf into countless bars all over the world; in general, this soundtrack sounds like a golf telecast (without the commercials - even better), and by and large this only bothers tennis players. So the golf game generally is left turned on at a moderate volume.

Leaving aside places like Houston ground control and Lawrence Livermore, the one place where beep & boop (and their descendants) do belong is the arcade. This is the place where we can make all the racket we want. Most games placed in an arcade are turned up moderately loudly, so that the player can hear what's going on in the game (after all this is why the players are there). The result of this is a public version of stereo wars. Probably none of us has to imagine what a game room sounds like in the aggregate - we've been there enough to remember. But as sound designers we must know what this din implies, namely, that anything lower in volume than our most punchy material will quickly slip below the noise flow.

So the problem we have to solve differs depending on where we expect our game to be installed. In a street location we must accommodate a cultural bias; in an arcade we must contend with a very narrow signal to noise ratio. In either case, we have to make a concession to the environment that steers us away from an ideal mix presented under ideal conditions. In short, we have to make everything in our soundtrack as loud as possible, relative to every other element of the soundtrack. In games that sound good in a street location, such as Golden Tee Golf, this takes the form of leaving sounds loud that in real life the player would expect to be much softer, such as the ball bouncing on the fairway or rolling into the cup. It's not realistic, or rather it's less realistic - but a good balance is struck between the needs of the player and the biases of the environment. In an arcade, the only time things should (relatively) quiet down is when the player stops doing something, like stepping away from his opponent in a fighting game. When the player is punching, kicking, blocking, throwing, catching, or shooting, the sounds all have to fall within about a 10dB signal to noise ratio.

Choosing Wisely

So what should we do? What sounds should we make? It all depends on the game. It always comes back to that. If the game needs to sound like a war zone, it's not going gracefully into many street locations, so don't worry about trying to sound like CNN's version of Desert Storm; Apocalypse Now or Saving Private Ryan are much better models. If we have the opportunity to make our game sound like a TV or radio broadcast, however, we give our game the chance to be audible in a street location.

The constraints we face in both types of game environments lead us to question the value of anything that isn't near the top of the volume curve. We can't afford a lack of clarity in the soundtrack, so whatever we want to stand out must be free of competition from anything that lives in the background. The easiest way to accomplish this is to severely limit the amount of background material. Ambience tracks, in general, need to be so quiet relative to foreground sounds that they'll just get lost in a coin-op environment. This is not true of background music since, culturally, music carries a foreground message (i.e. people often listen to music as a foreground pastime), and this allows music to live higher up in an interactive mix than an ambience track. But we can compose hot music that leaves lots of space for foreground events; indeed, how successful we are at this largely determines how graceful our overall soundtrack can be.

A great tool to have at our disposal is a mechanism to "duck" the volume of various elements when something more important comes up, then restore the volume after the foreground event has passed. This allows us to keep the level of the background music fairly high, without threatening to hide foreground elements when they appear. We must be aware, though, that the more things we have to duck, the more complicated the restore operation becomes; further, a subsequent event might end up ducking both the background and any other foreground sounds that might already be running. The moral of the story: the fewer things we have going on at any given time, the cleaner our soundtrack will be and the greater will be our audience's chance of hearing that soundtrack.

Non-video coin-op: Pinball and Redemption Games

Redemption games are games that pay out tickets for successfully performing some simple act, such as rolling a coin down a track at just the right moment to coincide with a particular light lighting up, or throwing a skee-ball into a particular hole, or stopping a light cycle at just the right place (a la roulette wheels). In each case, the game spits out a variable number of tickets depending on how well the player did; the player then redeems the tickets for small prizes such as stuffed animals. These games are wildly popular in kids' arcades and at amusement parks; they traditionally earn two to three times as much as a video game in a similar location. Even after the operator pays for the prizes, the profit margin on these machines is impressive compared with other coin-op devices.

There are three things to mention regarding sound for redemption games. The first is that we almost never see them in street locations; the prize inventory takes up too much room, and in general we frown on sending little kids into bars. So the problems of game sound in a street location don't really apply here: these games are almost all in arcades, so we only need to make sure that our redemption soundtrack is heard by our players clearly.

The second point is that redemption games tend to define simplicity of game play. The player does exactly the same thing, over and over; the result of his actions may produce three, five or, at the most, ten different results. There is no specific "start of game" or detailed "object of the game" beyond the simple act of dropping a coin, throwing a single ball or pressing a button. The game starts when the player performs the action; the game is over when the result is known and the tickets are awarded. That' s it. So our soundtrack, in general, consists of an action sound, a small number of reward sounds, and maybe a ticket spitting sound. We can do the entire job in a few hours, usually, and the beauty of it is we can make every sound the same volume (as loud as possible); the operators will turn us up loud enough to be heard, and our problem of clarity doesn't really exist.

The third point is that redemption games are designed for young kids, say ages 4 through maybe 11 or 12. Whether this is what the kids want, the customers for these machines (the operators) prefer the soundtracks to be upbeat and happy, like the color scheme of the game and cabinet art. Cute arpeggios and fanfares, cartoony swoops & boinks, tend to be where we end up on these projects. We don't use speech much; the games are too repetitive.

Pinball, or its predecessor Pachinko, is probably the original coin-op amusement game. It is a tribute to its spectacular basic game mechanic that the game is still manufactured 66 years after its invention. This basic idea (slanted board, two coil-actuated flippers at the bottom, some sort of plunger to put the ball back in play), which underwent very little outward change for the first 40 years or so, was augmented in recent years in two main ways: 1) the increasingly complex mechanisms into, around and through which the ball could be shot; and 2) the increasingly rich and complex sound system available for a relatively small increase in cost. Both of these developments, both of which were made possible by the digital revolution, led to increasingly complex sets of game rules being grafted onto the same basic play mechanic. Underneath all the new toys, the basic game was the same, and visually the amount of information didn't change much; we were still presented with a more-or-less static backbox painting, with static art drawn on a playfield embedded with blinking lights & flashers. The lights could go on and off to direct the player to make this or that shot, but in order to attach those shots to a theme, sound was increasingly relied on to carry the rules to the player, to reward the player for achieving the object of the game, and most importantly, to bring the theme of the game to life.

In the last ten years or so pinball soundtracks have come to resemble nothing so much as interactive radio plays. The challenge of designing such soundtracks in the arcade context is formidable; the game requires a compelling background track to carry the basic style or theme of the game, and that track has to be louder than we would like relative to the sound effects and speech cues. All the techniques we discussed above regarding arcade sound must be brought to bear on this problem. The result, if done well, is tremendously rewarding; the fact that we're not tied to a particular picture allows us unparalleled freedom in designing set pieces and reward sequences, and pinball sound designers over the year have responded to this freedom by producing some of the most wonderful (if not always wacky) soundtracks ever heard in any game.

The Indispensable Element?

Sound doesn't get the respect it deserves from coin-op management. Unlike home game developers, coin-op developers can with some legitimacy claim that "they just turn it down anyway." Incidentally, they continue to claim this today even in the face of a 1991 market survey of coin-op operators, whose number one request for improvement was "more and better sound!" How can we gain the respect we need to continue to innovate?

First of all, we must be able to refute the argument that "they just turn it down anyway." We know that a street location employee will get a game turned down that he or she doesn't want to listen to. We must do everything we can to create soundtracks for street games that those employees won't want to turn down. We do this, as we discussed above, by incorporating elements that sound as if they belong in the environment, and by providing as wide a variety of sound as possible for each repetitive thing the player does.

We should only have to jump through this hoop if we're working on street pieces. If a game is targeted for the arcade, we should be able to point out to recalcitrant management that we'll be laughed out of an arcade if our sound is turned down; that arcade operators rely on sound to attract players to a game; and that if they really believe that no one turns the sound up, they should simply take the sound hardware off the bill of materials and eliminate the sound from the development schedule entirely. This last sentence usually is sufficient to call management's bluff.

The ultimate way to refute this argument lies in remembering that the distaste for coin-op game sound was born in the 70's, with games that could only make beeps and boops. The fact that we can now incorporate recorded sound into soundtracks, and that we can vary the playback of those recordings in meaningful ways, means that we have the capability to sound much less out of place than these early efforts sounded. The short retort to the argument is, "They don't turn it as far down, as often, as you once thought." And we must follow through on this position and provide soundtracks that don't get turned down.

Two final thoughts: First, the single most dramatic improvement in the sound of a coin-op game would come from doubling the amount of money spent on speakers and amplifiers. A more powerful amp that avoids clipping distortion at higher volumes, coupled with more efficient transducers, will do more to warm up an electronic playback device than all the compression technology, all the interactive soundtrack technology, all the multiple voices, that we could (and would like to) throw at a coin-op game. If we play back a natural recorded sound cleanly, it will sound pretty much like what it is and be recognized as such. If that playback is clipped, the sound will be perceived as "electronic."

Finally: Our best friend in the industry is the game programmer responsible for putting sounds in the game. If that programmer likes sound and music, has a good sense of rhythm and timing, is given ample time to implement sound, and if we join with that programmer in giving our undivided attention to the implementation process, our game will sound better than even we thought it could.

Chris Granner studied composition and computer/electronic music at the University of Illinois, and computer programming at several jobs before joining Williams Electronics (now Midway Manufacturing) in 1986 as composer/sound programmer. He has composed music & designed soundtracks for about forty arcade titles, including Terminator II, Fish Tales, WWF Wrestlemania, Trog, and The Addams Family. He is currently the director of the audio group at Incredible Technologies. He can be reached at cgranner@itsgames.com.