The following small trick was too simple for a post in my eyes, but too helpful in the past for not sharing. From time to time a I have a long running command in my development process, that can be compiling, evaluating data or something else that takes longer than 5 minutes. In the most cases I will focus on something else in that time and check pretty late that the command has already finished. A simple solution for me is to play some audio signal after the command has finished.
I am digitizing all my old vinyl records and I would like to show how I am doing that with Audacity on Linux.
I am a little bit interested in the possibilities of reading audio input from the web browser. So I have written a small audio noise filter based on frame wise speech / non-speech classifcation a few days ago. This filter was based on the research of the paper Quantile based noise estimation for spectral subtraction and Wiener filtering. The paper contains also an other approach, a quantile based noise spectrum estimation.
I have read Visualizations with Web Audio API and just wanted to check on my own if I can build a small working example that can read the audio input from the Web Audio API.