Our audio lead, Andrew, stops by with a soundscape created to conceive what it feels like being on the player's home planet, Timber Hearth! Thanks, Andrew!
Hello everyone - this is Andrew Prahlow, audio lead for Outer Wilds. As some of you may know, I have been the composer for this game from the very beginning, but over the past months, I have also been redesigning the sound for OW to create a more refined atmosphere for the player.
There will also be a new music update very soon as I have been re-recording and rearranging the soundtrack and the old Outer Wilds themes to fit the new art and audio style. Below is a 4 minute audio walkthrough of Timber Hearth. Thank you all for your support!
Our tech artist/programmer/wunerkind Logan discusses how we went about creating the ocean feel for Giant's Deep!
Since the start of the project, I knew the gas giant of Giant’s Deep would be a significant tech art challenge. Hidden under a thick layer of clouds lies an ocean planet, and I knew I wouldn’t be happy throwing some scrolling normal maps on a sphere and calling it a day. As such, I decided I wanted an ocean with real waves.
The first big stumbling block to achieving this was having the polygons to turn into waves. We’d need to have it be a pretty high-resolution mesh, as you can jump down into the ocean in person, which makes the small waves just as important as the big ones. But we couldn’t have that detail everywhere, as that would choke things to a crawl pretty quickly. So we needed a giant sphere that gave us detail where we needed it, and faded it out where we didn’t.
Thinking about the problem, I broke it down into an idea of having patches of different resolution stitched together into a sphere. A quad-sphere seemed the best solution, as that would be the easiest topology to work with. Using Philip Nowell’s excellent cube-to-sphere mapping equations meant that I could just treat it like a cube, and then have the GPU do the work of warping it into a nice rounded shape, like so:
We talked with Outer Wilds creative director, Alex Beachum, about changes to one of the main player tools, the signalscope!
We recently went through a revision to the probe design and now we've gone through a similar revision to the signalscope. Can you explain why you wanted to revise the tool's design?
ALEX: The original concept behind the signalscope was that you would use it to find a signal and then you would then go investigate what it was. That's pretty straightforward. So the big structure of the game is that you have these clues that point you towards these major curiosities and a few of them are tied together heavily via using the signalscope.
For example, if you find a quantum area through the signalscope you learn the quantum signal you can then subsequently use your signalscope to find other quantum locations. So you could use the signalscope to hunt down the other threads to a specific mystery. That's the goal.
But it never 100% worked because players didn’t understand how to use it properly and even if they understood it, the tool wasn’t easy to use.
What specifically wasn't working about the tool?
ALEX: Originally all the sounds you could pick up were all on the same frequency so it was hard to isolate a signal. Additionally the algorithm we were using to pinpoint the signal’s location wasn’t optimized and the UI didn’t tell you enough.
So for the UI we added a direct indicator when you were on a signal that let you know when you were exactly pointed at it, where before it was just the signal’s wave getting bigger or smaller: that didn’t tell you much.
The big thing we added recently that’s been kind of a gamechanger for the tool is a distance indicator. The tool’s only useful if you can pick up signals on the planet you’re on AND on planets far away and know where those signals are. Players were pointing the signalscope at, say, the rock in front of them, and think the signal was coming from that, when it’s actually from the other side of the planet. And it was kind of impossible to tell what distance the player was attempting to aim at, but we know spit out the distance of a signal that’s hit. That seems like an obvious solution, but it took us a while to get to that
But you've made alterations in addition to those UI changes, yes?
ALEX: Yes. So instead of having all the signals in one channel we’ve been increasingly splitting them into their own channels, that happened a while ago, but we’ve been delving further into it.
We still have frequencies (and we’ve added to that list) for all the specific signals but we’ve added to this list a natural channel that you can you just pick up sounds from around the solar system like the tornadoes on Giant’s Deep or the waterfall on Timber Hearth. We wanted a channel where you could just hear everything and were rewarded for exploring.
The final big thing we’ll be doing it making sure people use the signalscope better is improve the village tutorial. So in addition to the planets you can see in the sky to find signals we’re adding something that you can isolate and find in the village.
Because the one thing we currently never teach players--and we need to--is the idea that the signal is being made by a particular object and they can hunt down that object and find that sound. The other thing we need to teach them is that if an object is making that sound in the solar system, there might be other objects in the solar system making the same sound. And if they’re getting the same signal from different locations, we want the player to put together that those signals are probably related somehow.
You mentioned previously that you wished you knew of the changes to probe design when doing original level design. Do you feel similarly about these signalscope updates?
ALEX: The signalscope is a little more planet-design agnostic and a large part of the original narrative design of the game was created with the signalscope in mind. Some of the main clues and curiosities were designed to be investigated with the signalscope. So I think the signalscope is more clearly integrated in the design of the game more than the probe was originally.
Wanting to hunt down a sound works no matter where you are, while the probe is really good at situations that aren’t as required to play the game, in a weird way, while the signalscope is really good at being required in an overt way.
It’s easier to explain to players what signalscopes are good for: we haven’t quite synced everything together, but we’re close. The probe is trickier because it’s a more nebulous concept to invite players to think, “Hey, maybe I should throw a probe over the horizon to see what’s there.”
Trying to teach them without forcing them to learn: it's one of the hardest design challenges/imperatives of Outer Wilds, isn't it?
ALEX: On a certain level you can’t just throw everything at the player in the beginning stage of the game. Outer Wilds has a lot of mechanics that you have to get good at that you can’t just throw at players at the beginning of the game and expect them to retain it, which means you have to distribute the tutorialization over the rest of the game which is really hard when the rest of your game is completely non-linear.
That’s why we’re putting a lot of tutorial things on the moon, because everyone in playtests goes to the moon early since it’s safe in comparison to other planets. Essentially our “second” tutorial is going to be on the moon and I think that’s going to be pretty successful once we finetune what they are.
Updates on our games, our process, and the joys of being Mobius Digital.