Geek asks and answers

Audio Streaming Detection

I want to write / program / develop an algorithm that can recognize many characteristics in the data from a string in / mic audio. The sound stream will be music, and I want to filter out the characteristics to distinguish the songs from each other, different I mean that you can play genres of songs.

One important thing that I absolutely want to discover is which bar / beat in the song. For example, I want to know if a song has 3/4 of the time.

The only useful articles I found were BPM detection, but this is not enough to distinguish a song from another song.

FFT is a good start to getting various features from an audio stream, but I don't know where to start. Is it possible to get bit / bit using FFT? Are there any good code examples / examples?

Is FFT sufficient to get good audio stream performance, or are there any other algorithms that are good for getting audio stream performance?

It is advisable that I do this in C #, because it is the programming language with which I have the most experience. Is this possible in C # or in another language better?

To summarize my question, Im is looking for any information about the detection of characteristics in the audio stream to get a bit / bar and other information to distinguish the songs.

+5
source share
3 answers

I liked reading this blogger's related articles:

http://www.redcode.nl/blog/2010/06/creating-shazam-in-java/

. , , - /, .

, , , .

!

+3

. , , . , .

(, Shazam) . - - . ; . , . (, ), , . , . () . , (, ) , , , "" .

, , , . , ; , , . , 10 .

0

The aubio open source library extracts functions from audio. It is written in C, but can serve as a reference for a managed implementation. Or you can P / Invoke to it.

aubio is a tool designed to extract annotations from audio signals. Its functions include segmenting the sound file before each of its attacks, determining the pitch, listening to the rhythm, and creating MIDI streams from live sound.

0
source

More articles:


All Articles