The Granite Canyon
Division Sound Design
...the evolution of
scale sound for the Athabasca System.
by
Jim Wells
The Symphony
Think of it as an eighty foot wide symphony for a theme park attraction... that's what I did. I took a musical approach to the entire project, I always do. Our theme is The Athabasca System Granite Canyon Division, a mountain division operating across a summit canyon pass in the Canadian Rockies. The Granite Canyon Symphony is comprised of some 12 movements (based on scenes defined along the eighty foot linear layout), each movement may contain several sub-themes. Each complete movement is a separate CD stereo soundtrack. The sequence of these movements is determined entirely by the visitor's interaction with the linear layout. The time signature is, well... daytime. I deliberately avoided any overall beat or rhythm in any of the soundtracks. Indeed, there is no synchronization between the soundtracks necessary, or even possible. The CDs are intentionally designed with different running times, and shuffle themselves automatically as they loop. All of the pieces are written as interconnected, and are arranged to work in concert with each other regardless of when individual CD players may be started. Our Conductor is Clayton Barry (he hits the start buttons!).
Appropriate soundtracks are designed for each scene or scenic element along the eighty feet of the linear layout. It was apparent from my earliest conversations with Clayton that water features were important both visually and aurally. Special Magic Canyon Creek and Magic Waterfall programs were created for these features. Other Environmental (natural) soundtracks are created for the Magic Summit timberline, and Magic Wind at the summit (complete with Canadian goose gaggle fly-bys overhead!).
Additionally, individual movements of The Granite Canyon Symphony are composed for the Magic City, Magic Central Station, Magic Old Town, Magic Industrial Area, Magic Ranch, Magic Farm, Magic Sawmill, Magic Mine, and Magic Summit Way Station (whew!... those are the titles). There are also plans to incorporate "single intermittent cricket" loops into one or two sections, and snoring will be a running gag at various points (usually positions of responsibility) along the several scale miles of the GCD. It only makes sense that the layout should sound like what it looks like. It should also sound like what you can't see... sometimes, hearing is believing! If the prototype made a noise, then it is likely audible on the GCD, as well.
The Orchestra
Each voice (each bird call, or cricket chirp, or sound effect), is first evolved in the computer as an individual dry sound. I have been collecting sounds for more than twenty years (some of them from vinyl), and have a substantial catalog/archive at this point. None-the-less, we did a special research project for the Athabasca soundtracks, collecting birds indigenous to British Columbia and the Canadian Rockies, the setting of the GCD. Once the appropriate voices were collected, they were sampled into the digital audio workstation (DAWS).
In the digital domain we can do anything with a sound we can imagine... so we do! A voice may be processed with EQ, given an exaggerated dynamic range, or run through single ended noise reduction. Sometimes it is necessary to remove unwanted background sounds (other critters, wind, water, or distant man made sounds). It is somewhat ironic that we must do some of the most UNnatural things to these little voices in order to make them appear natural and real, or even audible! But we must compensate for the fact that they will perform for the public at appropriate scale volume levels. Little tiny voices with a lot of character and internal dynamic range stand the best chance of "cutting through".
When the actual sound is not available (or will not work for any reason) a completely new sound will be "scratch built", the true essence of sound design. Existing or scratch built, it is only after being thoroughly tortured in the DAWS that they are arranged with other processed "instrument voices" into a performing section of the orchestra.
The Hall
More important than the individual voice itself, is the ambient sound, or "room" around each voice. This is comprised of early reflections (discreet echoes), and reverb (a smooth decay), within a stereo image. Real sound happens in a full scale ambiance, that's why it sounds real. In order for our scale voices to sound real, we must scale the ambiance back up to the full scale listener's world. It may seem backwards at first, but as we turn the voice itself down to a scale volume, the quieter background ambiance disappears below the threshold of hearing, sometimes vanishing completely. This is the primary reason why real recordings of real environments do not always work as well at scale volumes, the ambiance of the environment around the sound becomes inaudible at very low volumes. We must design an ambiance that reconnects the original voice to the full scale listener, just think of it as a magic concert hall in which we can create any listening environment. The actual method and processes are tedious, geeky technical stuff, and extremely boring! It is also a secret. But I can tell you that it is a direct result of listening at scale volumes, and on scale speakers. In essence, we are simply modeling scale distance.
The Rehearsal
Now it is time to assemble all of the players in the hall... and practice! This is as much fun as it is work. Multi-track arrangements are created by placing the sounds in stereo pairs across a real time framework. Much time is spent in expirimentation, but the final composition is strictly determined by what sounds natural, and by what makes us smile. We are also guided by a fair understanding of the way in which people naturally interact with sound. That multi-track is bounced to disk as a stereo pair, where it recieves a final stereo image tweak and level matching to the rest of the orchestra. With the CD mastering complete, its off to the concert stage.
The Performance
In our little musical analogy, the CD soundtracks AND layout sound systems act in concert. They are the orchestra at all public performance, the various sections of the orchestra are always positioned at their appropriate places across our 80' stage. Together, the soundtracks and the sound systems are the musicians in the orchestra, and the air around the layout is our hall. The musicians will perform it just as we rehearsed it, all the instruments and the hall are actually here... so, this will be a live concert.
Almost any sound is better than no sound at all. A cassette ghetto blaster under your layout, is still better than silence. But we want the Athabasca System to sound as real as possible, with no hint or audible cue to remind you that you are hearing a recording (You know its a recording, we just want to give you every opportunity to forget that!). To do this we developed a reference speaker system that is used not only in the layout, but also in the recording and mix-down processes. Although I used a beautiful JBL 5.1 theater surround system to do the pre-production work to each voice, I literally used the scale speaker components themselves as my monitors throughout the rest of the sound design and mastering processes.
The placement and orientation of the various speaker elements within the scenery, and even the scenic camouflage in front of the speaker devices, can audibly effect the sound image. "Audibly different" does not necessarily mean "sounds bad". Everything we have tried seems to work fine. Some speaker locations will invariably sound "better" than others, so we experiment around with the loose speakers operating before deciding on locations. This is very hard work, you have to actually listen. The use of matched reference "monitors" in the creation and recreation processes makes it almost impossible to position the speakers "wrong", really!
The use of the same scale components
in the studio and on the layout, is at the heart of successfully
marrying the soundtracks to the sound system! When the two act
in concert, viola! We have our orchestra. You can click on the
links below if you wish to see pictures and read brief descriptions
of the various audio system components.
The
Front End
The
"Can"
The
Tweeter
The
Woofer
The Fantasonics Engineering team working on the Granite Canyon Division sound project has been affectionately nicknamed "The GCD Blasting Dept." A themed rope handled explosives crate was built to house the 16 amp 12 vdc power supply and several of the CD player/amplifier front ends. The remaining systems are mounted in the linear layout sections where they are to perform. 12vdc is distributed along the length of the layout to provide power to all systems. If you are a hopeless sound addict (like me) who actually wants to know more about scale sound, or if you are considering modeling scale sound on your pike, you might want to take a look at some of the tutorials at The Model Railroad Magic Website.
Scale sound is fun. I truly believe that the aural image has just as much potential for modeling fun as the visual image does. Modeling in the two images, is always the same... always. There is only one slight difference. Having modeled both images in many scales, I can tell you that building a model of the aural image, although requiring slightly different skills, is much easier! It is always easier and less expensive to build a scale model of a sound, than it is to build the scale model to put it in. The goal is the same for both images. Like any good model in the visual image, our scale symphony should appear to be real, so that you can forget just for a moment, it isn't...
Questions?
Email the...