AudioI2S

This is the base library for the Smart Citizen Kit V2.0 for performing audio analysis.

It's intented to perform audio analysis on a SAMD21 processor of a MEMS I2S microphone output, with the idea of easily expanding it's functionality towards different inputs.

Additionally, the analysis can be performed through a configurable AudioAnalyser, containing a FFT analyser or a FIR (Finite Impulse Response) Filter.

Features:

FFT Analysis

1. Signal windowing through configurable window method (Hann, Blackman, Hamming)
2. FFT analysis with arm's cortex funtion on q31_t data type
3. Spectrum calculation and configurable weigthing (A, C or Z)
4. Custom Spectrum Equalisation (With auto generated parameters in a octave script)
5. RMS calculation
6. dB scale convertion

FIR Filtering

1. Signal downscaling
2. Signal filtering (by chunks) with arm_fir custom filter including equalisation and weighting
3. RMS calculation

A-C Weighting functions

There is an octave script in /OCTAVE/WEIGHTING that generates a custom weighting table as specifies in the actual legislation. It outputs a vector containing the corresponding factors to apply in the spectrum, according to the specifies sampling frequency / fft bins.

---

With kind references to:

• IoT_noise by Andre Filliettaz, 2016
• SpectrumSerialPlotter by Sandeep Mistry, 2016
• SPL meter - Davide Girone