Unleashing the Beat: Crafting Epic Soundscapes with Battle Mechs

Greetings everyone, I’m Ilya Viktorov, the Audio Lead at Pixonic (MY.GAMES). I’m responsible for managing the audio design of Battle Robots. Some time ago, I shared our expertise in capturing sound effects for our game. This time, I would like to delve deeper into the audio design process itself, emphasizing the obstacles we’ve encountered and the strategies we implemented. Therefore, whether you are just beginning your sound design journey for your game or seeking motivation – this article will offer valuable insights and practical tips.

*

Interestingly enough, audio design was not prioritized for Battle Robots until the fourth year of development, around 2018. In fact, back then, mobile games did not focus much on sound in general, often relying on pre-existing resources and the capabilities of common game engines.

When we finally decided to take audio design seriously, we inherited a considerable legacy of existing content – making the task of overhauling it daunting.

During the audio design phase, we faced several challenges that arose from the nature of a mobile project:

Related:Revamping Battle Robots’ reward system with style and user-friendliness

The FMOD audio library is an industry standard, but at the start of our project, we opted not to utilize it. (We attempted to implement it later, but it was already too late.)

Consequently, we faced the necessity to reorganize and test a vast number of content assets and countless prefabs. Even with automation, this could consume extensive time and manpower – not to mention the fact that we would have to pause the work of other departments during the content processing. So, what’s the way forward?

Our five strategies for audio production

The first approach involves utilizing ready-made resources and libraries. We prefer subscriptions over purchasing extensive pre-made libraries (the traditional method), where many resources are used while others go unused.

For our primary services, we utilize Soundsnap, Splice, and Soundly (which also serves as a very useful manager for the audio library). Furthermore, when working on weapon sounds, we frequently use the Weaponizer plugin from Krotos along with its extensive library of weapon components.

The second approach is sound recording – often faster and more convenient to create a sound ourselves than to find a new asset. This method becomes crucial when we need a specific sound that isn’t readily available or doesn’t align with the exact atmosphere we are aiming for.

However, recording realistic sounds for large robots and heavy machinery can be challenging. Even if you want a close approximation of real-life sounds, capturing the sound of actual heavy machinery can be very difficult. To achieve this, you might need to negotiate access to an industrial area, and there will surely be additional noise and echo. Keeping this in mind, to obtain the desired results, smaller devices are used, and the sounds are then specially processed.

In fact, the sounds of mechanisms in games and movies often include items like screwdrivers, coffee machines, door locks, printers, or blenders. A well-known example from popular culture is the iconic blaster sound from Star Wars, which was created by actually recording the noise of a Slinky toy using a contact microphone. In our case, the largest object whose sounds we used was a garbage truck unloading bins.

Our third approach is synthesis, which is excellent for generating sounds for energy weapons, magical effects, and user interface elements. We use wavetable synthesizers like Serum and Vital with custom waves, as well as modular systems like Softube Modular and granular samplers based on NI Reaktor.

For simpler sounds like bleeps and sweeps, we tend to rely on external synthesizers. My go-to instruments include the Novation Mininova, Behringer Neutron, and the Minimoog Model D clone.

Often, innovative solutions arise from experimentation: for instance, we discovered that the retro Yamaha RM1X groovebox features outstanding arpeggiator settings that create fascinating glitchy sounds reminiscent of classic interface sounds!

Our fourth…

The ability to convert skills into audio is a significant aspect of neural networks. Neural networks represent a relatively new and exciting domain of audio production. Though still imperfect, they yield intriguing and unexpected outcomes.

We have started employing the ElevenLabs generator to create basic sound elements, which we later combine into more complex sources. While we currently cannot produce final-quality results, we are capable of assembling something engaging from a dozen individual layers. Mechanical and enchanting sounds, in particular, derive much enjoyment from this approach.

We also utilize the Sonic Charge Synplant 2 plugin to generate sound variations based on reference files. This plugin analyzes audio files and synthesizes similar sounds that can be edited later. Magical effects and energy weapon sounds work exceptionally well with this method.

Our fifth technique is randomization; this enables us to produce an unlimited array of sounds from a small set of source materials. For example, we can select a few dozen distinct “mechanical” sounds, load them into the NI Battery sampler, and randomize the starting point, pitch, and volume for each. We can then use a keyboard or arpeggiator to generate more random sounds, from which we choose the most appealing ones.

This method can also be combined with synthesis, where we input random values in synthesizers or add a layer using the Synplant plugin. Additionally, we have developed a custom sampler for Bitwig that loads energy sounds from our libraries into separate layers – low, punch, snap, and tail – and randomly alternates between them.

Ultimately, our approach is truly a blend of the previous five techniques in various configurations. Generally speaking, we apply the primary four methods for several layers, which we then combine through randomization.

Implementing audio assets

As noted earlier, we encountered some challenges with the core functionalities of Harmony Audio, which pushed us to devise innovative solutions on the fly. To maintain control over sound levels, we categorized the audio into different types: Player, Guns, Mechs, Abilities, Ambient, Music, UI, and Sound. These categories were then routed to two main buses — Music and FX — managed through the settings menu.

The sounds produced by the player’s robot are routed through a distinct audio channel, which is prioritized above others and set to be 10% louder. In scenarios where multiple identical guns are equipped on a robot, only the sound from one gun plays while the others are silenced.

To simplify editing and avoid the complexity of managing a large number of prefabs used by various team members, we separated sound sources into individual prefabs. For each scene, ambient sounds, music tracks, audio sources, and reverb zones are organized within their dedicated game objects, placed at coordinates 0,0,0. This arrangement allows for easy editing without affecting the overall scene. A similar approach is taken for all weapon sounds.

This technique not only simplifies the editing process but also enables quick reuse of existing audio sources, eliminating the need to create new sounds when unnecessary. For instance, if a game designer is developing a new weapon similar in function to an existing one, they can easily utilize the corresponding audio prefab.

Considering that Battle Robots has existed for over a decade, the volume of content, including audio resources, has significantly increased. This expansion has resulted in challenges regarding library size. To address this issue, we implemented more robust compression for sounds that are less affected by it (such as clean explosions and hissing effects), reduced dependence on rare sounds for modifications, and prioritized the use of existing audio sources. This choice became challenging; players expect fresh, unique audio experiences.