Welcome to %s forums

BrainModular Users Forum

Login Register

auto classification of sounds

Create your own modules in C++
Post Reply
sm_jamieson
Member
Posts: 551
Contact:

Unread post by sm_jamieson » 26 Feb 2016, 20:25

I have not delved into audio processing yet - still messing with midi, etc.
I am trying to classify my sounds into genre to allow quick selection.
Would it be possible to produce a module that can analyse incoming sounds and classify them into
genre, such as brass, strings, piano, electric piano, etc.?
Simon.

User avatar
oli_lab
Member
Posts: 1261
Location: Brittany, France
Contact:

Unread post by oli_lab » 29 Feb 2016, 00:29

quite a while ago I did some patches to listen to audio and classify it by spectral complexity, from pure to noise.

the way I done it was by doing an FFT, and get the number of different frequencies present in the audio file : the more bands of significant value mean something quite noisy...

but to recognise instruments, my guess would be that you need balaysian networks, have a system that can learn to recognize enveloppe shape in a fuzzy way...well quite a project !
http://oli-lab.org

Win11 Ryzen9/32GB RAM - RME MADIFACE - SSL alpha link 4-16 - OSC capable interfaces

follow OLI_LAB adventures on Mastodon
@olivar_premier@mastodon.social

sm_jamieson
Member
Posts: 551
Contact:

Unread post by sm_jamieson » 29 Feb 2016, 10:49

It might be related to the methods used for speech recognition. I would thinking recognizing vowel sounds might be similar to recognizing the sound of a clarinet, flute or trumpet. A FFT to show frequency content on a graph would be a start.
Some stuff here:
http://hyperphysics.phy-astr.gsu.edu/hb ... imbre.html
And here:
http://amath.colorado.edu/pub/matlab/music/
OK, now this is good:
http://epubs.siam.org/doi/pdf/10.1137/S00361445003822

I won't put any more links in ...
Simon.

User avatar
oli_lab
Member
Posts: 1261
Location: Brittany, France
Contact:

Unread post by oli_lab » 29 Feb 2016, 21:29

that a lot of thick reading indeed !
But could be fun to be able to classified sound in real time !
http://oli-lab.org

Win11 Ryzen9/32GB RAM - RME MADIFACE - SSL alpha link 4-16 - OSC capable interfaces

follow OLI_LAB adventures on Mastodon
@olivar_premier@mastodon.social

Post Reply

Who is online

Users browsing this forum: No registered users and 4 guests