This is very cool! I have a question, I looked at the source code, and it seems to be based on Sliding Windowed Fourier Transform. Is it able to handle any instruments and chords, what about multiple instruments?
I'm unfamiliar with this area, but I'm very interested in it. Based on my previous readings, it seems to be very challenging to recognize chords. But this demo seems to work very well. I wonder if it has any limitations. Thanks.
Essentially, my project implements only the very first step of what's described in the article: turning sound into a picture. Separating instruments is very complicated. Even recognizing chords is a challenge. Nevertheless, my code does visualize multiple instruments; it just does not "know" what it is displaying.
Comments
This is very cool! I have a question, I looked at the source code, and it seems to be based on Sliding Windowed Fourier Transform. Is it able to handle any instruments and chords, what about multiple instruments?
I'm unfamiliar with this area, but I'm very interested in it. Based on my previous readings, it seems to be very challenging to recognize chords. But this demo seems to work very well. I wonder if it has any limitations. Thanks.
Check out this! https://www.lunaverus.com/cnn
Essentially, my project implements only the very first step of what's described in the article: turning sound into a picture. Separating instruments is very complicated. Even recognizing chords is a challenge. Nevertheless, my code does visualize multiple instruments; it just does not "know" what it is displaying.