def create_effect(effect_type):
input_file_path = 'voice_wf.wav'
output_name = 'voice_mod.wav'
BLOCKLEN = 1024 # Number of frames per block
samp_freq, audio_data = read(input_file_path)
if(effect_type == "Darth-Vader"):
output_audio = AudioLib.darth_vader(audio_data)
write(output_name,samp_freq,np.array(output_audio, dtype = np.int16))
if(effect_type == "Echo"):
output_audio = AudioLib.echo(audio_data)
write(output_name,samp_freq,np.array(output_audio, dtype = np.int16))
print('* Finished')
def plot_obs(self):
#Tamano caja
L = self.Lbox.get()
#Posicion de la fuente
X0 = self.xsource.get()
Y0 = self.ysource.get()
#Posicion del observador
XOb = self.xobs.get()
YOb = self.yobs.get()
#Frecuencia
f = self.freq.get()
#Velocidad de onda
vel = self.vel.get()
#Longitud de onda
lamb = vel / f
#Condicional de fronteras reflectivas
reflex = self.frontera.get()
#Intervalo de tiempo
dt = 1 / (44100.)
#Tiempo maximo
tmax = self.tmax.get()
#Arreglo de tiempo
self.tiempo = np.arange(0, tmax, dt)
#Calculo de campo en el punto del observador
self.Z = \
Z_amplitud(XOb, YOb, X0, Y0, self.tiempo, f, lamb )
if reflex == 1:
self.Z -= \
Z_amplitud(XOb, YOb, 0.0, -Y0, self.tiempo, f, lamb )+\
Z_amplitud(XOb, YOb, 0.0, 2*L-Y0, self.tiempo, f, lamb )+\
Z_amplitud(XOb, YOb, -X0, 0.0, self.tiempo, f, lamb )+\
Z_amplitud(XOb, YOb, 2*L-X0, 0.0, self.tiempo, f, lamb )
self.Z *= 1 / (5.)
#Grafica
self.ax.clear()
self.ax.grid()
self.ax.set_title('Amplitud Onda Sonora por Observador')
self.ax.set_xlabel('tiempo [s]')
self.ax.set_ylabel('Amplitud [$A_0$]')
self.ax.plot(self.tiempo, self.Z / (1.0 * ad.Amplitude))
self.ax.set_ylim((-1, 1))
self.ax.set_xlim((0, tmax))
self.canvas.draw()
#Creando objeto de audio
sonido = ad.audio()
#Cargando nota de audio
sonido.load(self.Z)
#Reproduciendo sonido
sonido.play()
def plot_obs(self):
#Tamano caja
L = self.Lbox.get()
#Posicion de la fuente
X0 = self.xsource.get()
Y0 = self.ysource.get()
#Posicion del observador
XOb = self.xobs.get()
YOb = self.yobs.get()
#Frecuencia
f = self.freq.get()
#Velocidad de onda
vel = self.vel.get()
#Longitud de onda
lamb = vel/f
#Condicional de fronteras reflectivas
reflex = self.frontera.get()
#Intervalo de tiempo
dt = 1/(44100.)
#Tiempo maximo
tmax = self.tmax.get()
#Arreglo de tiempo
self.tiempo = np.arange( 0, tmax, dt )
#Calculo de campo en el punto del observador
self.Z = \
Z_amplitud(XOb, YOb, X0, Y0, self.tiempo, f, lamb )
if reflex == 1:
self.Z -= \
Z_amplitud(XOb, YOb, 0.0, -Y0, self.tiempo, f, lamb )+\
Z_amplitud(XOb, YOb, 0.0, 2*L-Y0, self.tiempo, f, lamb )+\
Z_amplitud(XOb, YOb, -X0, 0.0, self.tiempo, f, lamb )+\
Z_amplitud(XOb, YOb, 2*L-X0, 0.0, self.tiempo, f, lamb )
self.Z *= 1/(5.)
#Grafica
self.ax.clear()
self.ax.grid()
self.ax.set_title('Amplitud Onda Sonora por Observador')
self.ax.set_xlabel('tiempo [s]')
self.ax.set_ylabel('Amplitud [$A_0$]')
self.ax.plot( self.tiempo, self.Z/(1.0*ad.Amplitude) )
self.ax.set_ylim( (-1, 1) )
self.ax.set_xlim( (0, tmax) )
self.canvas.draw()
#Creando objeto de audio
sonido = ad.audio()
#Cargando nota de audio
sonido.load( self.Z )
#Reproduciendo sonido
sonido.play()
tau = 1. #Tiempo maximo [s] tmax = 5 #Intervalos dt = 1/44100. #Arreglo de tiempo tiempo = np.arange( 0, tmax, dt ) #Potencial V = Vfem( tiempo ) #Oscilacion original y_cuerda = y( tiempo ) #Grafica plt.plot( tiempo, V, linewidth = 0.5 ) plt.grid() plt.title( "Potencial inducido por una bobina fonocaptora") plt.xlabel( "t [s]" ) plt.ylabel( "Potential [V_0]" ) plt.show() #Audio producido por la bobina nota_bobina = ad.audio() nota_bobina.load( V*ad.Amplitude/max(V) ) nota_bobina.play() #Audio original de la cuerda nota_cuerda = ad.audio() nota_cuerda.load( y_cuerda*ad.Amplitude/max(y_cuerda) ) nota_cuerda.play()
def fun_effect(stvar):
instrument = stvar
data = pd.read_csv("Notes_Freq_Match.csv")
data = np.asarray(data)
freq_notes = {}
for r in range(data.shape[0]):
for n in range(int(data[r, 2]), data[r, 3] + 1):
freq_notes.update({n: data[r, 0]})
freq_map = {}
for r in range(data.shape[0]):
for n in range(int(data[r, 2]), data[r, 3] + 1):
freq_map.update({n: data[r, 1]})
MAXVALUE = 2**15 - 1 # Maximum allowed output signal value (because WIDTH = 2)
WIDTH = 2 # bytes per sample
CHANNELS = 1 # mono
RATE = 8000 # Sampling rate (samples/second)
BLOCKSIZE = 8000 # length of block (samples)
DURATION = 10 # Duration (seconds)
voice_mod_wf = wave.open('voice_piano_wf.wav',
'w') # Write a mono wave file
voice_mod_wf.setnchannels(CHANNELS) # mono
voice_mod_wf.setsampwidth(WIDTH) # two bytes per sample
voice_mod_wf.setframerate(RATE) # samples per second
voice_wf = wave.open('voice_wf.wav', 'w') # Write a mono wave file
voice_wf.setnchannels(CHANNELS) # mono
voice_wf.setsampwidth(WIDTH) # two bytes per sample
voice_wf.setframerate(RATE) # samples per second
NumBlocks = int(DURATION * RATE / BLOCKSIZE)
output_block = BLOCKSIZE * [0]
Ta = 5
r = 0.01**(1.0 / (Ta * RATE))
# Filter coefficients
ORDER = 2
f = [[]] * data.shape[0]
a = [[]] * data.shape[0]
b = [[]] * data.shape[0]
x = [[]] * data.shape[0]
y = [[]] * data.shape[0]
om = [[]] * data.shape[0]
states = [[]] * data.shape[0]
new_states = [[]] * data.shape[0]
for i in range(data.shape[0]):
f[i] = data[i, 1]
om[i] = 2 * pi * f[i] / RATE
a[i] = [1, -2 * r * cos(om[i]), r**2]
b[i] = [r * sin(om[i])]
states[i] = np.zeros(ORDER)
new_states[i] = np.zeros(ORDER)
x[i] = np.zeros(BLOCKSIZE)
y[i] = np.zeros(BLOCKSIZE)
yout = np.zeros(BLOCKSIZE)
DBscale = True
# Initialize plot window:
plt.ion() # Turn on interactive mode so plot gets updated
fig = plt.figure(1)
[g1] = plt.plot([], [], 'red')
[g2] = plt.plot([], [], 'blue')
g1.set_label('Input')
g2.set_label('Output')
plt.legend()
g1.set_xdata(RATE / BLOCKSIZE * np.arange(0, BLOCKSIZE))
g2.set_xdata(RATE / BLOCKSIZE * np.arange(0, BLOCKSIZE))
plt.xlim(0, 0.5 * RATE)
plt.xlabel('Frequency (Hz)')
plt.ylim(0, 150)
# Open audio device:
p = pyaudio.PyAudio()
PA_FORMAT = p.get_format_from_width(WIDTH)
stream = p.open(format=PA_FORMAT,
channels=CHANNELS,
rate=RATE,
input=True,
output=True,
frames_per_buffer=256)
for i in range(0, NumBlocks):
input_bytes = stream.read(
BLOCKSIZE, exception_on_overflow=False) # Read audio input stream
input_tuple = struct.unpack('h' * BLOCKSIZE, input_bytes) # Convert
X = np.fft.fft(input_tuple)
freqs = np.fft.fftfreq(len(X))
idx = np.argmax(np.abs(X))
freq = freqs[idx]
freq_in_hertz = int(abs(freq * RATE))
f_mapped = freq_map[freq_in_hertz]
idx = np.where(data == f_mapped)[0]
idx = int(idx)
if ((freq_in_hertz >= 200) & (freq_in_hertz <= data[71, 1])):
x[idx][0] = 1000
if (instrument == "Piano"):
for j in range(0, data.shape[0]):
[y[j], new_states[j]] = signal.lfilter(b[j],
a[j],
x[j],
zi=states[j])
if (j < data.shape[0] - 1):
states[j + 1] = new_states[j]
x[j][0] = 0.0
yout += y[j]
elif ((instrument == "Guitar") & (freq_in_hertz >= 200)):
n = freq_in_hertz - 262
yout = AudioLib.set_stretch(guitar_wf, 10, 200, 50)
if len(yout != BLOCKSIZE):
yout = np.pad(yout, (0, BLOCKSIZE - len(yout)), 'constant')
else:
print(
"Frequency not suitable for guitar transformation. Try higher frequncies next time"
)
#for j in range (0,data.shape[0]):
# yout += y[j]
# y[j] = np.zeros(BLOCKSIZE)
yout = AudioLib.bpf(yout, f_mapped - 10, f_mapped + 10)
output_block = np.clip(yout.astype(int), -MAXVALUE, MAXVALUE)
Y = np.fft.fft(output_block)
# Convert values to binary data
output_bytes = struct.pack('h' * BLOCKSIZE, *output_block)
# Write binary data to audio output stream
stream.write(output_bytes)
voice_mod_wf.writeframes(output_bytes)
voice_wf.writeframes(input_bytes)
# Update y-data of plot
g1.set_ydata(20 * np.log10(np.abs(X)))
plt.pause(0.001)
g2.set_ydata(20 * np.log10(np.abs(Y)))
plt.pause(0.001)
yout = np.zeros(BLOCKSIZE)
output_block = BLOCKSIZE * [0]
plt.close()
stream.stop_stream()
stream.close()
p.terminate()
print('* Finished')
# Efecto Doppler #========================================================== #********************************************************** # MODULOS #********************************************************** import numpy as np import os import matplotlib.pylab as plt import AudioLib as ad #Velocidad del sonido [m/s] vs = 331.5 #Creando objeto de audio sonido = ad.audio() #Intervalo de tiempo dt = 1/(44100.) #Frecuencia de nota [Hz] freq0 = 110. #Tiempo maximo [s] tmax = 6. #Arreglo de tiempo tiempo = np.arange( 0, tmax, dt ) #Nota a reproducir nota = ad.Amplitude*np.sin( 2*np.pi*freq0*tiempo ) #Cargando nota de audio sonido.load( nota )
# Efecto Doppler #========================================================== #********************************************************** # MODULOS #********************************************************** import numpy as np import os import matplotlib.pylab as plt import AudioLib as ad #Velocidad del sonido [m/s] vs = 331.5 #Creando objeto de audio sonido = ad.audio() #Intervalo de tiempo dt = 1 / (44100.) #Frecuencia de nota [Hz] freq0 = 110. #Tiempo maximo [s] tmax = 6. #Arreglo de tiempo tiempo = np.arange(0, tmax, dt) #Nota a reproducir nota = ad.Amplitude * np.sin(2 * np.pi * freq0 * tiempo) #Cargando nota de audio sonido.load(nota)
