By Kurt Larson
Audio examples available on StarsReach.com
In the first post I outlined our overall approach to audio and why we do it that way. With all that in mind, here we can take a deeper dive.
DETAILS
Within the over-arching role audio plays in a game, different layers of audio have different jobs, goals, and effects. I'll talk about some of them here:
SFX
TOOLS AND WEAPONS
Audio for tools and weapons reinforces and rewards player agency. I believe that when the player is doing something with their (character's) hands, the feedback should be rich and rewarding. For this reason I am putting special attention and much resources into this component. For featured, often-used sounds such as this, it is very important to provide a great deal of variation to avoid the sound becoming tiresome. This is mainly achieved through randomization, and layering, and just having many different versions of a sound available.
For something like the Excavator beam sound, I have several different layers set up. Each time you turn the beam on, all seven layers play. But each layer has twenty-or-so different variations. Each of these is chosen at random by the sound engine each time you turn the beam on. Even with just two layers, that would be 400 slightly-different-sounding variation. With seven layers, it's 1.28 billion variations. They all sound recognizably similar, but never exactly the same, so your brain doesn't stick on the exact sameness and make you feel distracted.
Sounds which are non-continuous, such as gunshots, are easier, because you just fire the group of sounds and you're done. Most of the fine-tuning work comes with continuous, looping sounds, like our beam tools. (Excavator, Harvester, heat and freeze rays, etc.) This is where repetition becomes a problem to be solved. Even though variations are chosen at random, if we just looped those variations, after a while you'd start to notice a recurring, repetitive pattern. So for most of the layers, I took a long evolving varying sound and cut it into many sections of between 3 and 8 seconds each. Then one of those is randomly chosen, played, and cross-faded into the next randomly chosen one, and so on. You may, if you listen carefully, recognize little details which you've heard before, but it will likely never play in exactly the same pattern twice. It's not impossible, but by the time it does, you would likely not remember having heard that exact pattern before. Repetition in audio is only a problem when it grabs your attention.
CREATURES
Audio for creatures, especially creatures which interact with the player, is a huge and important undertaking. As our creature design is still evolving, and since we have been focusing on weapons and interactive objects, enemies in Stars Reach do not currently have any audio. But we are thinking about it.
Most creature vocalization audio in most games and movies is built up using recordings of real-life animals. No matter how alien or supernatural or bizarre the creature is, people respond best when they have at least some sort of recognizable sound as a starting point. (Think Chewbacca) I'd like to go beyond that a bit, but not so far that the voices of the creatures no longer speak to the player. One thing I intend to try is blending real-world animal sounds with completely other types of sounds, using a convolution reverb. (https://www.bhphotovideo.com/find/newsLetter/Convolution-Reverb.jsp)
I also want to try to give each of our creatures its own emotional profile. Like a personality, but more primal. The key here again will be variations. With footsteps you can get away with eight or even five. (I usually use twenty) With vocalizations it's better to have at least 30. The general rule I have observed is: the more personality a sound has, and the more it speaks to the player, the more variations you need.
UI
User Interface sounds are the ones you will hear the most. Counter-intuitively they require the least variation. Why?
My sense of it is that UI sounds are pure abstractions. They represent abstract concepts and actions which are, crucially, not representing something "physically" happening in the game world. It also helps that they are usually very short. As such, hearing the identical same sound every time you open a particular window may not be the best possible sound design, but it is often good enough. Even if you go for a very organic sound set, like, small wooden doors, drawers, and latches being opened and closed, for example, you still can tolerate a fairly high degree of repetition. In fact, the need for close similarity between the variations is quite high for UI audio. Each sound is a symbol for an abstract concept, like opening an inventory window or hitting an 'OK' button, and as such it needs to reliably signal the intended concept. Our UI sounds try to use the same sound for similar things, such that opening a window will always sound similar, but quite different from, say, hitting a 'Confirm' button.
BGA - BACKGROUND AMBIENCE
Background Ambience is all the layers of sound which represent the physical environment in which the character is located. Stars Reach provides a tremendous opportunity to create intricate, intimate background ambience which can continuously change, smoothly and meaningfully. Whereas a more classic game might only have a stereo background track which changes to a different one when you cross the border between the Ice Zone and the Lava Zone, the worlds of Stars Reach are continuously changing and evolving. Since they can be driven by naturally-patterned weather, environmental, and planetary data, and since we can tie all those data into the audio system, we can and will do things like:
Around sunset, we can gradually cross-fade a strong afternoon wind with a gentle evening breeze
As the player moves from a dry, desert environment to a higher-elevation forested biome, we can cross-fade different wind, start up some birds singing, turn on some light reverb-echo from the sound bouncing around in the trees, and you will likely hear some water flowing in some areas near streams.
When rain happens, you'll hear it. Not just the rain sound starting up, but birds will mostly stop singing, wind will change, footsteps will take on a bit of a squelchy tone, etc.
As one moves through different biomes; to a more humid or more dry area, subtle filtering changes can occur which can make all the audio sound slightly muffled or slightly crisper.
Since our worlds will have seasons, our audio will respond according to the weather, temperature, humidity, and other data as the seasons change.
The possibilities are endless, but you get the idea. The important thing is that all this audio behavior can occur naturally, automatically, along with the evolving behaviors of our living worlds, rather than being pre-planned, scripted transitions.
In the third and final post, I’ll talk about voice-over and music.
_____________________________________________
With all this attention to detail in our sound design, we are building an immersive and evolving audio experience that brings Stars Reach to life in a way that feels natural, dynamic, and deeply engaging. Every footstep, every tool activation, and every distant creature call is crafted to enhance the player's connection to the world.
But none of this happens without support. If you are as excited as we are about the depth we are bringing to Stars Reach, we need your help. Our Kickstarter is live, and backing us now means you will be a direct part of making this experience a reality.
By pledging, you are not just funding development. You are shaping a living and breathing universe where sound reacts, adapts, and evolves alongside you. Join us on this journey and help us push Stars Reach to the next level.
[Back us on Kickstarter today!] https://www.kickstarter.com/projects/starsreach/stars-reach
Every pledge counts, and we cannot wait to bring you more updates soon. Thanks for your support!