Machine Learning promises to democratize audio production for everyone, but can it make the right decisions when it counts? We put the very latest to the test.
AI plugins have been making waves in recent months, making the intriguing prospect of automatic processing a reality. While this development raises valid questions surrounding quality, and ultimately the industry as a whole, there are still many tasks that AI driven tech was made for.
To understand their impact it’s perhaps helpful to group potential AI solutions into two categories; creative and utility. We’re all familiar with the familiar creative plugin tools that allow us to sculpt sounds and/or improve their subjective quality. In the other camp, utility tools carry out tasks such as A-B comparison, or metering duties.
Audio limiters tread an interesting line between the two. Of course limiting can be used creatively, but predominantly it is a utility to stop digital overs, and to set levels to meet the loudness specs for any given platform. Given that its main use is an objective one, it should be a pretty good candidate for an AI application.
Sonible smart:limit
This AI powered limiter analyzes your audio in real time and not only sets the amount of limiting, but with the Quality check button engaged, also reports the loudness penalty that will be introduced by any of various services included in its database. It then makes suggestions about limiter settings- the amount of limiting untimately is always set by the user. Features include:
Automatic limiter parametrization in seconds
Extensive loudness monitoring with all relevant information for publishing
Visually guided loudness and dynamics targeting for streaming services, loudness standards, genres and reference tracks
4 sound-shaping tools for fine-tuning and styling
Instant Impact Prediction for a smooth workflow
Quality check for confident settings adjustment
Distortion display and intelligent auto release
Does It Deliver?
Using Sonible smart:limit is totally intuitive, and the content-aware initial limiting settings are always on the mark. Even when you take over and push it into obvious distortion territory, the sound holds out for longer than the average limiter and can hold its own against high-end dedicated limiting tools. That makes it good for anything, even for classical music where some limiting and absolute quality have to work together. Can an AI limiter deliver? This one certainly can.