5 Tips to Improve in Game Audio

/

Here are some thoughts I had about working in game audio.

Finding another way

Is not unusual for audio to not have as many resources as we would like. The reality is that our discipline is not appreciated in the same way fancy graphics or cool social features are. But this doesn’t mean we can just do mediocre work. Bad audio is noticeable while good audio is often invisible but realy enhances the player experience.

Finding another way means that, as a sound designer, you must work within the contrains you have to make things work and this usually means making compromises with quality, level of detail and performance. So maybe you can’t realy do things the way you initially planned but you must be resiliant enough to go around the obstacles and deliver something great anyway.

Take time to experiment

Audio has personality, it has a spirit. Sounds connects us to nature in an instictive way, they remind us to animals and weather. When creating audio for a machine, a creature, UI or an environment, we are tasked with giving them a personality, a certain flavour. For this, it can be very helpful to think about what you want to convey, what is the function of this thing in the story and in the world.

Sometimes that’s not enough and you just need to try crazy things, random stuff and see what sticks. I have created some great sounds like this but this certainly means you need to be willing to experiment freely which is not always possible when you need to meet deadlines. So remember to take time to stop and smell the roses, even aimlessly. You will get to results that can’t be achieved any other way.

Use limitations to boost creativity

Don’t see limitations as an obstacle, see them as a way to thrive. Less is more, sure, but is deeper than that. When you are limited to, say a single synth or instrument or just a few tracks or voices you are really forced to learn the only resources that you have deeply and get a knowledge ans mastery that you would never get if you have an arsenal of dozens of plusings to choose from.

Keep in mind the big picture

Is easy to over focus on what you need to do each day. You make sounds and implement them following a plan like ticking boxes. This happens usually when you base your work on lists, spreadsheets or jira tickets. It seems like as long as you tick boxes and cross tasks, you are progressing. This is needed, sure, but never forget that that doesn’t mater at all if the overall result is not working.

Always remember, the final user and their experience. At the end of the day, nobody cares about how you made that sound, how that bit of code is brilliant or the fact that you are knocking down tickets. Take a step back and play as a naive player, see what works and what doesn’t.

Be in flux with information

Things are going to be changing and fast. Features come and go, they are transformed and expanded. It can be tough to keep track of all of this, particularly when audio is usually left out of these decisions. Setting good comunications and expectations with the team is important but also remember that game development moves fast and you can’t possibly know every single thing.

You need to find the proper bandwith of information for each phase of development and keep on top of things but never compromising the actual work you need to do. For me, is helpful to remember that things must be flexible, that nothing is set in stone.

Exploring Sound Design Tools: Wormhole

/

Today I’m having a look at Zynaptiq’s Wormhole, a quite versatile warping, pitch bending plugin to use on sound effects, voice over and even musical elements.

Interface & Features

Wormhole has a few sections that operate more or less independently. It features the usual BIG wheel/knob thingy that other Zynaptiq’s pugins usually use, which in this case shifts the main warping effect up or down.

So if we look at the plugin’s signal flow, we see that it mainly consists of a warping section and shifting section. Both the dry and wet signals can be then delayed and finally reverb can be added at the very end or just after the main sections.

If we have a nother look at the UI, we can see where each of these sections live:

Warping

So let’s have a look at the Warp section which is the main feature of this plugin. Zynaptiq describes it as “proprietary time-domain local spectrum inversion“. Somewhat similar to a ring modulator but capable of producing unique sounds. Let’s see its settings:

Warp Depth: This adjust the frequency spectrum affected, low values will only affect high frequencies, while higher values will affect the whole spectrum but would give a milder result. To me, changing this value kind of feels like pitch shifting in a way. You can hear some examples here:

Poles: Increasing this value tightens up the frequencies and harmonics generated around an specific pitch while lower values create a more smeared sound. So use higher values if you want a thinned out type of sound. All the examples below where recorded while depth was at 50%.

Tilt: This shifts the whole warped spectrum up or down. So it kind of works like frequency shifting but with formant distorting qualities. This is what we control with the big circular knob.

Filter: This is just a simple low-pass filter than can be used after the warp and poles parameters (Pre) or afterthe tilt process (Post). The pre setting is generally better to create broken sounds, while post will result in cleaner results. Personally, I generally don’t notice much of a difference between pre and post.

Shifting

The shifting section is the other main feature Wormhole offers We find both frequency and pitch shifting. The difference between these two is that with frequency shifting, we are just moving all frequencies the same amount up or down which will results in more dramatic changes for low frequencies, musically speaking but far smaller changes for higher ones. In this sense, we could perfectly shift the fundamental note but the harmonics won’t be in the correct place since they are higher frequency. Pitch shifting, on the other hand, takes this into consideration and moves higher frequencies exponentially further away so the result sounds musical.

Let’s have a look at the settings that it offers so we can understand how it works:

Frequency Shift: It goes from -4000 Hz to +4000 Hz. Yeah, Hz, not semitones since this is frequency shift, not pitch shift. There are two modes here, linear and map. I see that map is the default option as it offers more granular change around 0 while the extremes change pitch in a more dramatic way.

Decay Time: Frequencies that play for longer than the desired time, will be discarded. So at the maximum value, nothing will change but as the value is lower the sound tends to dry out, reverb and noise can be reduced.

Pitch Shift: This is the proper pitch shift section and it uses 4 different modes:

  • Smooth: +/- 48 Semitones, optmized for low transient sounds.

  • Thight: +/- 48 Semitones, optimized for transient heavy sounds

  • Detune A: +/- 48 cents where L and R channels are shifted to produce a widening effect. Tight algorithm.

  • Detune B: +/- 48 cents where L and R channels are shifted to produce a widening effect. Smooth algorithm.

Other Features

We can also find a simple reverb with size and damp controls plus wet/dry controls for each individual section (Warp, Shift and Reverb). These three sections can change in routing order too.

Additionally, there is a blend section that offers something a bit more powerful than a simple master wet/dry mix control. When processing audio, you usually can hear the dry and wet at the same time, instead of a new, unique sound. Wormhole offers you a different way of blending dry and wet as the sound is gradually processed so you can have much smoother transitions and interesting intermediate states. Let’s see the modes it offers:

  • X-Fade: Normal dy/wet mix.

  • Morph A: Uses the properties of the wet signal to create the intermediate states.

  • Morph B: Uses the dry signal as the base, resulting in more subtle blending.

Finally, Wormhole has a delay value so you can blend dry and wet with a timing difference.

Processing Examples

Quick & early iteration for game audio

/

When your job is creative and open ended sometimes is hard to feels things are finished. Even if you design, export and implement a particular SFX on your project, is hard to be sure things are done. Art is not finished, is abandoned. This sometimes produces a kind of “paralysis by analysis” situation where if you try to really make sure everything is perfect you would never finish in time.

That’s why I’m advocating for a diffferent approach on this post: Quick and early iteration. Don’t worry about things being perfect, not even “finished”, just put ideas together and throw them into the real world as fast as possible. There are a few advantages to this approach.

Context

For the most part, audio is never going to be heard in isolation so why bother making a SFX super detailed and interesting if its context is never going to allow it to shine? This is analogous to the mixing engineer trying to make the kick drum sound awesone for hours, only to realize that it doesn’t work in the context of the song.

You don’t always know how much spectral space do you have. You don’t always know how much time you have. If you try to get these things perfect you will probably be too slow on a project with hundreds of sounds.

Instead, make your best guess, put some audio together and get it into the game. As more and more audio gets in, you will start to get a sense of how the sounds work together. Later on, is when you can start thinking about teaking: with context.

Implementation

Getting audio into the game quick also lets you know sooner rather than later how an audio event should work functionally. This is important because it can affect how it should be designed. Should it be a loop? Use a speed parameter? Should you bake many different sound into a single file or have many small sounds played from a scattered instrument?

If you actually implement it, you will be faced with these decisions in a real way that will always be better than just planning. Don’t get me wrong, you should plan your work of course but that is never going to tell you if things are going to actually work. For that, you need to get your hands dirty.

On the code side of things, an early simple implementation will give you insights about your current audio tech. You will know if your audio scripts are sufficient or if you need to add new functionallities or ask a programmer to help you out. In terms of project planning, knowing this soon is a big advantage and producers will appreciate it.

Lastly, somehing nice about early implementation is that once you get feedback on the audio, you just need to swap or tweak the assets and future iteration will be quick and painless.

Focused Priorities

It would be tricky to know what is important if you don’t have a clear idea of what is going to be there. Throwing audio to the game soon gives you a more clear idea of what you need to focus on. Your time is not infinite so you need to pick your battles and add detail and love to the content that most will benefit from it.

Once a respectable amount of audio is implemented, it will be easier to decide what is best to focus on and you will get a better sense of what is important.

Earlier Unknowns

Game development is usually about solving problems and these problems are often hard to predict. You can sit all day and plan how you are going to do things but it would be impossible to take some of these future issues into account. The next best thing is to know about them as early as possible.

That way, you can tackle them with more time and have a better chance getting the resources you need.

Optimization

Something to also consider is how efficient you are in terms of memory and CPU. This is hard to determine in the earlier phases of the project and if you go too slow on the first stages of the game development, it may be too late to have enough time to react.

Building the audio earlier would give more information about how things perform and you will be able to make better strategic decisions on how to do things. In turn, this will also influence how audio is built and designed.

80% is enough

Quick iteration has an additional advantage: if anything goes wrong or new time constrains come, you know you will be 80% of the way there and that could be enough to save the day.

This is something that goes against all our instintcs as creative, detail-oriented people. We always strive for more emotive, interesting and inspiring audio but that can only happen if we have the time to do it and sometimes we just don’t. If worst comes to worst, at least we know things are functional and all the main aural information is already conveyed to the player. Maybe is not pretty but it works and sometimes, in an emergency, that could be just enough.

Iteration

So that’s pretty much the idea. As you get more and more audio in place, always trying to be fast rather than perfect, you will get a better idea of what to spend more time on and that’s when iteration comes into focus. As a project reaches maturity, you will find yourself working less on new content and more on iteration.

You will be surprised to see that most things will just need tweaking while only a few will need to be re-done from scratch. On the long run, I think you save time working this way and the result will be better. Another nice advantage is the peace of mind that comes from knowing that most things are in place and now is just a question of iteration and refinement.

Lastly, as you get more experience working this way, you will find that your initial quick, dirty work gets closer and closer to be good enough quality sometimes even high quality as you sharpen your instincts and workflows.

Does this sound good?

/

I’ve been working on audio for a bunch of years now. That means making hundreds of decisions everyday about audio perception. Does this sound too big? Too dark? Too harsh? Too muffled?

You make those decisions based on your sonic experience of the world on the one hand and on your experience of other media on the other hand. So you decide what sounds good based on other people decisions in the past. You stand on the shoulders of giants. How many people do sound design for a living? Ben Burtt must be quite sore.

Audio, good audio at least, is always somewhere between true reality and that distorted perception of reality which comes from years of listening to media. Guns really don’t sound like they do in the movies and probably dinosours didn’t either. We are all subjected to this sonic zeitgeist that we can’t escape. You need to distance yourself from reality and use trickery so the audience feels that what they hear is authentic. Such is the irony of the sound design job.

But I can cope with that. Art is really about encouraging that delusion, isn’t it? About being larger than life, about dreaming, fantasy and imagination. That’s fine. The problem doesn’t really come with being real or authentic, we crossed that bridge when we started doing cave paintings. The issue comes with the simple challenge of evaluatiing quality.

For the most part, of course you know when things sound good. You even know how to improve them if you have the time. So you make choices, tweak things and listen again. But on the back of your brain there is this weird feeling, that tells you, but how do I really know? This is a common illness among creative professionals.

Objective things that you can measure are much more easy on the human neocortex. Maths and physics (maybe not quantum physics) are comfortably predictable. On the same vein, the logic behind programming offers the same warmth. Even though you don’t understand why something doesn’t work, there is a logical explanation and you just need to find it.

But quality appreciation on a creative field is a different beast. Is still processed in the brain, but much more based on instinct, computed using a mix of new and ancient parts of it. This is something that my analytical part dislikes because is a process that I can’t contain and use at my leasure that easily. It is not a lightbulb, clean and convenient, but a fire with uneven heat and light.

So there is not much else to do but to let go and follow your instincts. Sometimes it won’t feel right, sometimes you won’t be able to even tell but that’s the gig.

Is always good to sanity check with the people around you but beware. The perfect recipie to do a bad job is trying to please everybody. People have very particular opinons about audio and they don’t usually know how to express them very well so for sure take the feedback, but not at face value or you will go crazy.

Leaning on your own opinion all the time sometimes feels like that Nietzsche quote "if you stare into the abyss, the abyss stares back at you", but is probably your best bet. Firstly because you will always be improving as you flex your creative muscle. And secondly, and most importantly, because you is the only constant in your career. Leaning on the only tool you will always, always have seems like a good idea to me.

Or if everything else fails, you can always take a break.