FFTPower returns -Infinity

Question

FFTPower returns -Infinity

asemlucben opened this issue a year ago · 6 comments

Hi, i am playing with your package and at some point, while performing an FFT with a square wave i got -Infinity as output value

IDE Informations:

FftSharp V1.1.6
VS 2022 PRO
.NET6

Code snippet:

double[] signal = GetSignalFromDB();
// calculate the power spectral density using FFT
double[] psd = Transform.FFTpower(signal);

Input data:
Square wave from -10 to +10
InputData.txt

Output result:

Answer 1 · 2023-02-28T15:08:19.000Z

Hi @asemlucben, Thanks for reporting this! I'll take a look later today.

For now, here is the sample data in a more easily accessible format:

with open("InputData.txt") as f:
    lines = f.readlines()

values = [x.strip().split("\t")[1] for x in lines[1:]]
print(", ".join(values))

10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 
-10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, 
-10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, 
-10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, 
-10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, 
-10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10

Answer 2 · 2023-02-28T23:26:02.000Z

Hi @asemlucben, I took a closer look and this actually has a simple answer!

double[] input =
{
    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10
};

double[] fftMag = FftSharp.Transform.FFTmagnitude(input);
Console.WriteLine(fftMag.Last());

double[] fftPower = FftSharp.Transform.FFTpower(input);
Console.WriteLine(fftPower.Last());

0
-∞

If you take the FFT magnitude, the last value is zero.

FFT power is the log10 of magnitude, and according the dotnet math module this is negative infinity

Console.WriteLine(Math.Log10(0)); // displays -∞

Answer 3 · 2023-02-28T23:31:11.000Z

I'm re-opening this because I think there may still be an issue... it surprises me that the final magnitude is zero.

Checking these same values with Python I get different results:

import numpy as np

values = [
    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10
]

fft = np.fft.fft(values)
print(fft)

fftMag = np.absolute(fft)
print(fftMag)

fftPower = 20 * np.log10(fftMag)
print(fftPower)

[40.        +0.j         40.01318228-0.61384268j 40.05278844-1.22918984j
 ... 40.11899712+1.84755755j 40.05278844+1.22918984j
 40.01318228+0.61384268j]

[40.         40.01789049 40.07164545 ... 40.16151639 40.07164545
 40.01789049]

Warning (from warnings module):
RuntimeWarning: divide by zero encountered in log10
[32.04119983 32.04508383 32.05674352 ... 32.07620205 32.05674352
 32.04508383]

Answer 4 · 2023-02-28T23:47:27.000Z

@asemlucben I'll keep working on this, but I can report that using this code to calculate the FFT power manually works and produces values identical to python and Numpy

C#

using System.Diagnostics;

double[] input =
{
    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10
};

// manually load data into complex format
FftSharp.Complex[] buffer = new FftSharp.Complex[input.Length];
for (int i = 0; i < input.Length; i++)
{
    buffer[i].Real = input[i];
}

// manually calculate the FFT
FftSharp.Transform.FFT(buffer);
for (int i = 0; i < buffer.Length; i++)
{
    Debug.WriteLineIf(i % 100 == 0, $"fft[{i}]: ({buffer[i]})");
}

// manually calculate the magnitude
double[] fftMag = new double[input.Length];
for (int i = 0; i < fftMag.Length; i++)
{
    fftMag[i] = buffer[i].Magnitude;
    Debug.WriteLineIf(i % 100 == 0, $"fftMag[{i}] {fftMag[i]}");
}

// manually calculate the power
double[] fftPower = new double[input.Length];
for (int i = 0; i < fftPower.Length; i++)
{
    fftPower[i] = 20 * Math.Log10(fftMag[i]);
    Debug.WriteLineIf(i % 100 == 0, $"fftPower[{i}] {fftPower[i]}");
}

fft[0]: (40+0j)
fft[100]: (-1.150632244976336+31.23986182456488j)
fft[200]: (-11.10781419580801-0.8193615996107755j)
fft[300]: (-0.7297560403960999+6.5804308655978065j)
fft[400]: (-10.767682044079358-1.5972347495074448j)
fft[500]: (0.28805946898834733-1.5471646526486902j)
fft[600]: (-9.87660405558076-2.2178656072326746j)
fft[700]: (2.5892724623023806-9.82386388207797j)
fft[800]: (-6.942558739214263-2.106002169289817j)
fft[900]: (11.099724457196958-32.2645210105908j)
fft[1000]: (49.68716053112529+19.16645281400144j)

fftMag[0] 40
fftMag[100] 31.261044790299085
fftMag[200] 11.137993070545098
fftMag[300] 6.620771424494799
fftMag[400] 10.885501157384676
fftMag[500] 1.5737524329065835
fftMag[600] 10.122560719623266
fftMag[700] 10.159361862717237
fftMag[800] 7.25495465075377
fftMag[900] 34.12042202359976
fftMag[1000] 53.255674205826736

fftPower[0] 32.04119982655924
fftPower[100] 29.900069773777957
fftPower[200] 20.93613886718728
fftPower[300] 16.418171891158536
fftPower[400] 20.736968566258778
fftPower[500] 3.9387282900812077
fftPower[600] 20.10580781079615
fftPower[700] 20.137328591626584
fftPower[800] 17.212694041914926
fftPower[900] 30.66028788331365
fftPower[1000] 34.52731774268314

Python

import numpy as np

values = [
    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10
]

fft = np.fft.fft(values)
for i in range(len(fft)):
    if (i%100==0):
        print(f"fft[{i}]: {fft[i]}")

fftMag = np.absolute(fft)
for i in range(len(fftMag)):
    if (i%100==0):
        print(f"fftMag[{i}]: {fftMag[i]}")

fftPower = 20 * np.log10(fftMag)
for i in range(len(fftPower)):
    if (i%100==0):
        print(f"fftPower[{i}]: {fftPower[i]}")

fft[0]: (40+0j)
fft[100]: (-1.150632244976375+31.239861824564898j)
fft[200]: (-11.107814195807997-0.8193615996107582j)
fft[300]: (-0.7297560403960937+6.580430865597791j)
fft[400]: (-10.767682044079365-1.5972347495074413j)
fft[500]: (0.2880594689883047-1.5471646526487017j)
fft[600]: (-9.876604055580763-2.2178656072326817j)
fft[700]: (2.589272462302379-9.823863882077985j)
fft[800]: (-6.942558739214267-2.1060021692898214j)
fft[900]: (11.099724457196949-32.26452101059081j)
fft[1000]: (49.68716053112527+19.166452814001453j)

fftMag[0]: 40.0
fftMag[100]: 31.261044790299103
fftMag[200]: 11.137993070545086
fftMag[300]: 6.620771424494783
fftMag[400]: 10.885501157384683
fftMag[500]: 1.5737524329065873
fftMag[600]: 10.122560719623271
fftMag[700]: 10.159361862717253
fftMag[800]: 7.254954650753775
fftMag[900]: 34.120422023599765
fftMag[1000]: 53.25567420582672

fftPower[0]: 32.04119982655924
fftPower[100]: 29.90006977377796
fftPower[200]: 20.936138867187275
fftPower[300]: 16.41817189115851
fftPower[400]: 20.73696856625878
fftPower[500]: 3.9387282900812286
fftPower[600]: 20.105807810796158
fftPower[700]: 20.1373285916266
fftPower[800]: 17.212694041914933
fftPower[900]: 30.660287883313654
fftPower[1000]: 34.52731774268314

Answer 5 · 2023-03-01T00:03:00.000Z

What a rabbit hole! I confirmed the identical behavior is reported by Python and numpy, so I don't think there's actually a bug here:

import numpy as np

values = [
    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10,
    -10, -10, -10, -10, -10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, -10, -10, -10, -10, -10, -10, -10, -10, -10, -10, 10, 10, 10, 10
]

fft = np.fft.fft(values)
realFftPoints = int(len(fft)/2+1)
fftMag = np.absolute(fft)[:realFftPoints] / realFftPoints
print(f"last fftMag: {fftMag[-1]}")

fftPower = 20 * np.log10(fftMag)
print(f"last fftPower: {fftPower[-1]}")

last fftMag: 0.0
last fftPower: -inf

Answer 6 · 2023-03-01T07:45:02.000Z

Thsnk you for your investigation, i will add a control loop after the calculation to make sure this value gets erased and replaced with a proper number that does not mess up the autoscale of my graph!