I have noticed that the audio tech industry at large comprises of 3 worlds that rarely collaborates with each other and I really wish they blend into one, soon:
- The DAWs/Plugins world — C++, JUCE, VST3, CLAP, web apps etc.
- The ML world — Music Information Retrieval (MIR), live coding & all other research efforts
- The Game Audio world — Wwise, FMOD, MetaSounds, game engines etc
Where do things stand now?
- Game audio — this cohort does not talk to ML folks (almost hates them). They also rarely speak to the DAWs/plugins world in terms of making features, even when they creatively use them all the time. At max, they’d talk about REAPER.
- ML world knows nothing about game audio. I know startup founders sitting on $M funding having no clue what this means and I’m afraid when the hype grows, they’ll just jump to it for the sake of it. Some devs have started to finally explore audio plugins coz they realised just doing research and making APIs will never make them reach the creatives. But they still are far from DAWs.
- DAWs/plugins is too constrained within itself and does not see possibilities in game audio. The tools that do make a bridge are bare minimum. And they are now exploring ML coz a few plugins have set an example. But no one is still seeing the potential of ML models in workflow improvements. Nuendo & DaVinci Resolve did recently!
What these 3 worlds need to learn from each other (give me credits if you’re using any of these ideas):
- Version Control: The Game Audio world solved this well. I’ve pitched this to almost ALL DAW makers, only to get sceptical replies. Wwise’s WorkUnit system enables people to work on the SAME project file and merge the data easily. It’s 2025. Why don’t DAWs have this yet? Some allow remote realtime collaboration on same file but still it’s not very efficient. Why do we still need to export STEMs? Even dawProject cannot solve this coz it’s all about “entire project” or nothing at all.
Please, can we have at least one pioneer in this, so that everyone else can copy?!
- Common Plugin Format: The DAW world solved this well. But for the audio industry at large, are we all adopting CLAP or not? Enough with all the suspense already! We’re all in the same industry. If someone wrote a kickass sounding algorithm, how is it that we still cannot just take it to games engines, video editors and other apps with ease?
I’m not a fan of the process of deploying plugins on gaming hardware — feels like gatekeeping! I’d love to see a much easy process that’s as intuitive as it is to export VST3 for DAWs.
- Cloud Processing: The ML world has solved this well. Don’t have a powerful system? No worries, just make a Google Collab notebook and you’ll have access to world class GPUs in no time. I don’t see any DAW utilising cloud (WavTool did, but it got acquired)
Worried about the “wait time”? This is a curse word in the world of DAWs, but it doesn’t have to be. Explore non-realtime processing features. Use the cloud. Save project files. Render audio in the cloud!
One place where these worlds can start a conversation is the Audio Developer Conference (ADC), and I'm keen on continuing my efforts in that direction.
What do you think about the collaboration between these subdomains of the audio industry? Let me know and I'd be happy to discuss more on this.