Player footstep sounds in the FPS genre are crucial, because they can contribute a meaningful purpose to the game’s audio vision! They can be a vital aspect to tell a story and push the overall game experience extensively, as well as help the player/listener to make certain game play decisions during gaming. They can provide disclosure to the listener, revealing either weight, speed, surface material or space the player character happens to be in or walks on. In a multiplayer game play scenario they’re invaluable. They can warn the player when somebody approaches or moves about in distant. Ultimately, which can help the listener to look for cover and estimate its range. However, not only do they provide essential information to the listener, but also arise tricky questions about their production and technical implementation process - leaving it up to the audio artists to decide on how they need to be integrated, and esthetically aligned to the nature of the games needs.
Inside the game audio community there have been current discussions on how to implement player footsteps effectively. Many different opinions from audio professionals exist! My good friend and technical sound designer, Damian Kastbauer has had a recent conversation on this topic over at the Game Audio Podcast, debating with special quests: Julian Kwasneski, David Steinwedel and Kenneth Young who shared their personal views on this. Make sure to check it out if you get a chance! On a personal note, I think that the production and integration of player character footsteps does vastly depend on the game’s nature, beneath genre, design, relevance and style of the game. There is no common rule of whats right or wrong – if they sound good, inside the game, then that’s the way to go :-) It may take hours of experimentation, good team communication and tweaks of your middleware knobs, until you reach satisfaction.
Sound transparency = the dynamic composition of sound effects, music and voice overs, all blended nicely together within the game’s soundtrack, serves to keep the listening experience pleasant and interesting to the auditor. This also may relate to casual and portable games, despite the fact that audio of course is lower in its high-fidelity, and the number of streamed audio files minimal. The more dynamic and balanced a mix is put together, the more likely the listener gets sucked up into the immersive listening experience, to fully relish the game play adventures.
But how does this related to the implementation of footsteps? Well for instance, in some occasions, I would want player footsteps to stand out on purpose – emphasizing on a particular space or location that player is in at a given time. Such as in a tunnel, or a narrow basement. Where room acoustics can be a good servant to paint the tonal sonic picture of the dimension of a room. It is here where footsteps could become prioritized and gain slightly in volume, leaving the remaining sounds low on the totem pole (often archived through bus ducking). Now in contrast, in a battle scenario for example, the games activities (sounds played at once) will be elevated (weapons, explosions, intense VO) and footsteps by default become less notable and relevant to the player. Because the gamer rather wants to pay attention to the sound of close-by bullet impacts or essential voices overs of team-mates, in order to trace their position (multiplayer). Actually, very similar to a real-world situation! Imagine, if you were walking on the street in the middle of the night on a semi-busy street, listening to the echo of your footsteps bouncing off close-by walls. A transporter truck approaches and passes you by… if you’re still moving you wouldn’t hear a single step – only the distinct sound of the truck going by! Right?!
Fortunately, with today’s innovative and advanced audio middleware applications, such as Wwise, make it easy to set up real-time parameters to control the dynamic of an adaptive in-game mixing case. Basically, we’re able to control volume curves, fx and multiple EQ-filter changes adaptively. This enables us to adjust artistic modifications based on the entire game play situation. It relates to ambient, music and single sound effect events! The control schemes are endless, and creativity and a little bit of technical mind-bending (scripting) around on the game engine is key to highly-immersive results. However, there are times when we, as sound designers are limited to on-hand streaming memory – as we constantly have to battle to draw back on sound events. Even the ones we fell in love with.
On the note implementation, there is a great article at DesigningSound.Org with Jeff Wesevich and Justin Drust talking about their Dynamic Wind System which has been applied in Ghost Recon: Warfighter 2. It utilizes visual parameters to control the increase/decrease and intensity level of wind layers to build up tension and draw away from it, based on altitude and location of the player character. Also make sure to check out Damian’s blog and interesting articles on DesigningSound.Org where he shares many informative stories about the technical aspects of sound design.
As with everything, to plan ahead and to write down notes on: ‘what needs to be recored’, will help a lot in order to stay organized throughout production. I even used normal paper with a analog pen! I’ve had several instances where the iPhone made terrible artifacts into the recordings, in search for a near-by Wi-Fi connection. Sometime we just forget to switch-off our phones to Fligh-Mode. Mainly during the day or on my way home I tried to take notes on locations that I’ve passed by, putting down comments on where and what possibly could be recorded there.
It takes good understanding of your environmental surroundings, by that I mean, whether if you life in a big city or a small town, it is good to know your locations to plan for a recording at some point. Right now I life in Gaineville, Florida, a small college town and it sometimes seems difficult to get away from traffic noise to get a clear shot when recording. Remote foley can be a true hustle and requires TONS of time, as well as initial trail and error attempts. It took me a while to really get what I was hearing in my head and there is always the unexpected. Like sick cats or dogs barking, crickets or drunk teenagers etc.! Anyway, a lot of times I found myself headed out for recordings in the middle of the night, when traffic noise was low. It was pretty much a hit-and-run type scenario, with the goal to avoid attracting the police – called up by the local neighbor- watch, informing them that they saw a man holding a big ‘gun’. Oh boy, I twice had the honor to meet them face to face. Plus I have an European accent – it certainly can get complicated Anyway, I could capture two or even four different surfaces per night performing on about two different locations.
Production – Fun Part?
Now – lets plunge a little into the fun part and examine the recording techniques I applied for a first person shooter project I worked on this year. As we’ve learned in school, all footsteps assets (or in general foley props) get recorded dry, within either a foley stage, performed by a foley artist OR in this very case – via remote foley or worldizing! There are good and bad things going about this recording method. I tell you right away, its time consuming, and can be a lot of fun! Here are some of the pros and cons:
- It gives an overall higher sense of natural rhythm
- It adds a unique and exclusive sound texture to the recordings
- Physical work out to get out of the studio (-:
- Continuity is vital (same shoes and intensity level)
- demands extra hours
- To find suitable locations and surfaces
- Very time consuming: Editing and recording!
Here is a video of how I recorded most of the player footsteps – holding a mic in my hand and pointing it towards my feet. Common requirements for variation were: walk, run, shuffel, turn, stop, jump and land- YES all performed by me! For when the player jumps down a higher object or platform, landing sounds were split up into: soft, medium and hard. So did the performance!
Testing & Tweaking
After each completed recording session, the next step was: editing, mastering and integration. We’ve build a footsteps test-map in the UnrealEditor, defining each different surfaces that the player character would encounter to walk on in a specific level. This gave us good feedback time of how they sounded to tell us what needed to be changed. Usually, we would set up random containers (in Wwise), a setting for each of the variations, to randomly play them back to us – focusing on panning, volume & pitch-randomization. I tossed about 10-15 separated footstep samples into each surface defined and named container and successively minimized them down to 4-5, based on performance and overall feel they communicated. I got rid of all unpleasant sounding samples and shaped them so they would work together as set. Here is a video capture of a test map representing different material surfaces that the player performs on.
And here is a video of testing water footsteps in on of the sewer areas. Actually, I recorded these towards the end of the Florida summer period at my apartments swimming pool, when pool maintenance let water out. Phew! I am glad I didn’t slip ending up with a water-soaked microphones. All went well, though!
In the end, it was a great learning experience to apply this recording technique. Even though, it takes up a lot of time and patience (which most of us may not have), I feel that the final quality on when the character moves around in the maps, make in all worth while. Its funny! I sometimes even recognize the sound and recording location, which puts a smile on my face!