Word Embeddings for Automatic Equalization in Audio Mixing

Venkatesh, Satvik, Moffat, David and Miranda, Eduardo Reck 2022 Word Embeddings for Automatic Equalization in Audio Mixing. Journal of the Audio Engineering Society. (In Press)

Submitted Paper
Text (Submitted Paper)
2202.08898.pdf - Submitted Version
Available under License All Rights Reserved.

Download (375kB) | Preview


In recent years, machine learning has been widely adopted to automate the audio mixing process. Automatic mixing systems have been applied to var- ious audio effects such as gain-adjustment, stereo panning, equalization, and reverberation. These systems can be controlled through visual interfaces, pro- viding audio examples, using knobs, and semantic descriptors. Using semantic descriptors or textual information to control these systems is an effective way for artists to communicate their creative goals. Furthermore, sometimes artists use non-technical words that may not be understood by the mixing system, or even a mixing engineer. In this paper, we explore the novel idea of using word embeddings to represent semantic descriptors. Word embeddings are generally obtained by training neural networks on large corpora of written text. These embeddings serve as the input layer of the neural network to create a trans- lation from words to EQ settings. Using this technique, the machine learning model can also generate EQ settings for semantic descriptors that it has not seen before. We perform experiments to demonstrate the feasibility of this idea. In addition, we compare the EQ settings of humans with the predictions of the neural network to evaluate the quality of predictions. The results showed that the embedding layer enables the neural network to understand semantic descrip- tors. We observed that the models with embedding layers perform better those without embedding layers, but not as good as human labels.

Item Type: Publication - Article
Subjects: Computer Science
Divisions: Plymouth Marine Laboratory > Other (PML)
Depositing User: Dr David Moffat
Date made live: 26 Oct 2022 13:56
Last Modified: 26 Oct 2022 13:56
URI: https://plymsea.ac.uk/id/eprint/9747

Actions (login required)

View Item View Item