Audio spectrum visualization normalize loudness countour

Question

I've written an audio spectrum visualizer as you can see on the following screenshot. It works pretty well but the peaks are not really normalized. I think this is caused by this fact: Equal-loudness contour.

Does anybody know how to get balanced peaks (like a horizontal line)?

Here is my code: https://github.com/frankred/openFrameworks-music-visualization/blob/master/ofApp.cpp

enter image description here

score 1 · Answer 1 · answered Mar 29 '15 at 00:02

1

Well, if you already know the equal-loudness contour, why not divide by it?

answered Mar 29 '15 at 00:02

Abhay

768
4
13

Do you know for sure if that is the problem why it is not more like a horinzontal line? – Frank Roth Mar 29 '15 at 00:04
Whatever the cause, the OP wants to get balanced peaks, probably for a good visual effect. – Abhay Mar 29 '15 at 00:07

score 0 · Answer 2 · answered Mar 29 '15 at 11:00

0

Is your horizontal scale linear frequency? Usually, a log frequency scale (i.e., equal dimension per octave-based) is more effective than linear in illustrating spectral distribution of audio.

answered Mar 29 '15 at 11:00

David Reaves

1

I'm not really sure about what you mean. I get this spectrum out of FFT (see function fft::powerSpectrum on file https://github.com/frankred/openFrameworks-music-visualization/blob/master/fft.cpp) – Frank Roth Mar 29 '15 at 12:46

Audio spectrum visualization normalize loudness countour

2 Answers2