# coding: utf-8

# import Python modules
import os
import sys
import numpy as np
import matplotlib.pyplot as plt
import scipy
from scipy.io import wavfile
from scipy import signal

## 1.
# Edit the script to analyse the audio wav file (short recording of a musical instrument) in time domain and frequency domain using python modules
# you can resuse parts of exercise0.py from your previous exercise
#!! FILL IN PARTS WITH "None"

# Load audio signal
fs, x = wavfile.read('/Users/mac/OneDrive - TUNI.fi/SGN/Audio/Ex1/glockenspiel-a-2.wav')
number_of_samples = x.size
duration = x.size/fs
print('{} {}'.format('Sampling frequency (in Hz) is: ', fs))
print('{} {}'.format('Duration (in s) is: ', duration))

t_scale = np.linspace(0, float(x.size)/fs, x.size) # Hint: use the function np.linspace
amp_scale = x

# Plot time domain visualization
plt.figure(1)
plt.subplot(2, 1, 1)
plt.plot(t_scale, amp_scale)
plt.title('Time domain visualization of the audio signal')
plt.axis('tight')
plt.grid('on')
plt.ylabel('Amplitude')
plt.xlabel('Time (s)')

# Plot frequency domain visualization using FFT
max_freq = fs/2
win_len = 2048
number_frequencies = int(win_len / 2) - 1
t_start = 10000
X = np.fft.fft(x[t_start:t_start+win_len])
frq_scale = np.linspace(0, max_freq, number_frequencies)
mag_scale = 20*np.log10(np.abs(X[0:int(win_len / 2) - 1]))

plt.subplot(2, 1, 2)
plt.plot(frq_scale, mag_scale)
plt.title('Frequency domain visualization of the audio signal')
plt.axis('tight')
plt.grid('on')
plt.ylabel('Magnitude (dB)')
plt.xlabel('Frequency (Hz)')

plt.tight_layout()
plt.show()
# length of signal is 3,6 s
# Fundamental frequency is 1878 Hz
# 2. frequency: 7012
# 3. frequency: 10032

## 2
# Edit the script to generate a Sine wave signal y0,
# with amplitude 3000 (so you will hear something if you play the signal),
# frequency F0, and length of t seconds, as calculated from step 1.

# Generate sine wave y0
F0 = 1878 # To be checked from the previous plot
amplitude = 3000
y0 = amplitude * np.sin(2 * np.pi * F0 * t_scale)

# Plot time domain visualization
short_length = 100
t_scale_short = np.linspace(0, float(short_length)/fs, short_length)
amp_scale_short = y0[0:short_length]

plt.figure(2)
plt.subplot(2, 1, 1)
plt.plot(t_scale_short, amp_scale_short)
plt.title('Time domain visualization of the sine wave y0')
plt.axis('tight')
plt.grid('on')
plt.ylabel('Amplitude')
plt.xlabel('Time (samples)')

# Apply fast Fourier transform (FFT) to the Sine wave. Display the FFT of the waveform.

# Plot frequency domain visualization using FFT
Y0 = np.fft.fft(y0)
mag_scale = 20*np.log10(np.abs(x[0:int(win_len / 2) - 1]))

plt.subplot(2, 1, 2)
plt.plot(frq_scale, mag_scale)
plt.title('Frequency domain visualization of the sine wave y0')
plt.axis('tight')
plt.grid('on')
plt.ylabel('Magnitude (dB)')
plt.xlabel('Frequency (Hz)')

plt.tight_layout()
plt.show()