This repository contains the Jupyter Notebook that was shared as part of the Dolby.io blog post titled Improve Speech by Controlling Sibilance with Python by Rudy Trubitt.
Given the overview of sibilance in speech and how it can negatively impact the quality of audio recordings, the Jupyter Notebook demonstrates how to use the Dolby.io Media API to detect and remove sibilance from audio files.
Knowledge of audio processing and Python may be helpful in understanding the concepts and code presented in the blog. To follow along with the tutorial, you would need to have Python 3, Jupyter Notebook, NumPy, and librosa libraries installed. The instructions on how to install these dependencies are explicit in the code.
As part of performing the detection and removal of sibilance, you would need to sign up for a Dolby.io account You receive 50GBs to start out for free, which should help you get started with this project.
Make sure you have your API key ready and a pre-signed URL before uploading your audio file for enhancing.
In the case any bugs occur, report it using Github issues, and we will see to it.
We welcome your interest in trying to experiment with our repos.
If there are any suggestions or if you would like to deliver any positive notes, feel free to open an issue and let us know!
For a deeper dive, we welcome you to review the following:
- Media Enhance API
- Getting Started with Enhance API
- API Reference
- Sampling Excerpts of Media to Determine Ideal Enhance Parameters
- Predicting Impact of Dialog Enhancement on Large Audiences
- Blog Session - Media API
Using decades of Dolby's research in sight and sound technology, Dolby.io provides APIs to integrate real-time streaming, voice & video communications, and file-based media processing into your applications. Sign up for a free account to get started building the next generation of immersive, interactive, and social apps.
