If I take the audio data on the iPhone (i.e. real) data, I perform the FFT, and then I take the values (Re ^ 2 + Im ^ 2).
They vary from> 0 to some large numbers, so I do 10log(n)to get it in dB.
10log(n)
This gives me outputs that are negative (for inputs that were <1) positive.
But the examples I saw for this (and also for drawing the spectrum in Sonic Visualiser) always have a positive spectrum when measured in dB.
So what did I miss ?!
In a broader sense, since I understand that decibels are a ratio, so in this context, when you turn the FFT values into dB, what are their relationships to?
, , , . , iPhone, , , , .
, , , dB (SPL), , .
:, , , . , , 0,1 , , 100 , 100, 0,1 ( 2, -1). , . , dB SPL, , , log (P/P ref), .
FFT , , . FFT iPhone , , , AGC . ( , , , ..), , - .