FFT to dBa to overall dBa

So I have a .wav file and I am trying to determine the overall sound energy from the file. The code I have is pretty basic so far and provides an accurate FFT of the sound being produced from the file.

clc

clear all

close all

[data, fs] = audioread('TEST.wav');

data_fft = fft(data);

dB = 10.*log10((10*abs(data_fft)).^2);

frequency(:,1) = ((1:480000)/10); %480,000 48 kHz (divide 10 second audio file)

Weighting = 2+20.*log10((12200^2.*frequency.^4./((frequency.^2+20.6.^2).*(frequency.^2+12200.^2).*((frequency.^2+107.2.^2).*(frequency.^2+737.9.^2)).^0.5))); %Apply A weighting

dBa = dB + Weighting;

Now that I have the dBa values vs frequency, I want to determine the overal sound value from this. The angle I am working towards is determining the 1/3 octave "averaged" values potentially from a filter and then taking the "rms" of those values but I am not sure how to apply this filter or take into account all of the values.

Answer 1

[zachary gilvey]

Hello,

Simply what you would do is add all of the ratio of pressures. To do this I wrote a while loop. Here 48 is the 48 kHz sample rate. They are summed and then log'd again.

i = 0;

while i < length(frequency)/2

i = i+1;

Lp = dBa(i,1)./10; %Lp/10=log10(Prms^2/Pref^2)

Prms_Pref = 10^Lp; %10^(Lp/10)=Prms^2/Pref^2

P_ratio(i,1) = Prms_Pref;

end

Leq = 10*log10(sum(P_ratio)/48);

Leq = round(Leq,3,'significant')