In the introduction of acoustic books and literature, we will introduce books that our R & D representators are always referring to and papers that are attracting attention.
The links to the books and literature introduced are also posted, so please use it if you are interested.
Francis Rumsey
"Audio Processing - Learning from Experience"
(Japanese translation "audio signal processing -learning from experience")
Journal of Aes, feature Article
Volume 69, Number 5, May 2021
Page.361-365
https://www.aes.org/e-lib/browse.cfm?elib=21041
Keywords: Audio signal processing, machine learning, audio technology
"Audio Processing - Learning from Experience"
(Japanese translation "audio signal processing -learning from experience")
Journal of Aes, feature Article
Volume 69, Number 5, May 2021
Page.361-365
https://www.aes.org/e-lib/browse.cfm?elib=21041
Keywords: Audio signal processing, machine learning, audio technology
"I'm excited about the future of audio signal processing!"
That's my thoughts when I read this literature.
This time, there is a headquarters in New York in the United States, with about 20,000 members, from experts, researchers, and university teachers on audio and acoustic technology around the world, and has a branch in Japan. Here is a review article called Feature ArticleS from the journal journal of Audio Engineering Society (JAES) published by AES (Audio Engineering Society), an international association.
This article is originally a professor of Surrey University in the UK and is currently the consultant techNical winner of JAES. The review article he wrote is to take up a theme and introduce the latest AES contribution papers related to it, and the trends and future prospects in the theme field are well organized, and research and develop audio technology. It will be very helpful for going.
The article mentioned this time focuses on audio signal processing, as the title says "Audio Processing". Earphones and headphones use audio signal processing such as equalizer and active noise canceling, so you may be familiar. In recent years, technology has been researched and developed to learn various characteristics contained in audio signals, and automatically switches parameter tables that include gains, delays, and filter setting values. For example, the volume of the audio signal of the multi -channel is analyzed in real time, and the parameter tables are automatically switched to the effect that matches each scene, such as scenes that are important and power -oriented scenes, and the home theater is immersive. Can be improved. The technology of analyzing audio signals is becoming more and more important to maximize the effects. In this article, some papers have been introduced about the technology for analyzing audio signals, and it is a must -read in consideration of the possibility of audio signal processing.
By the way, can you tell the sound source of live recording and studio recording? Perhaps many people can hear and distinguish. So what are you relying on? In response to the question, there are many answers, "from the reverberation of the voice" and "somehow from the feeling of sound". By letting the machine learn this discrimination, you can distinguish live recording and studio recording.
What do you rely on? In other words, what characteristics do people know and what are they? Furthermore, what should I do to use the discrimination result to perform signal processing more effectively? It is also important to think about what kind of state is ideal and set the purpose as the root.
I feel that it is interesting to learn these people's discrimination standards with machines.
I myself have a successful experience in analyzing the movie scene from the audio signal of the multi -channel, automatically generating the sound field effect that matches the scene, improving the power of the entire movie, and increasing the immersion in the home theater. There is a great feeling of the combination of machine learning and audio signal processing.
I think this article, which allows you to know the current situation and future prospects for audio signal processing to make earphones and headphones more attractive, may be very useful.
Shake
