Day: February 27, 2018

$ ~/anaconda3/bin/python test4.py ~/40wpm.wav 0 -1
 
<BT> 
NOW 
40 
WPM 
<BT> 
TEXT 
IS 
FROM 
JANUARY 
2014 
QST 
PAGE 
46
<BT>
SETTING 
THE 
VFO. 
WITH 
THE 
TRANSMITTER 
OUTPUT 
CONNECTED 
TO 
A 
DUMMY 
LOAD, 
SET 
THE 
VFO 
TO 
THE 
CORRECT 
FREQUENCY 
FOR 
THE 
BAND 
OF 
INTEREST.

import sys
import numpy as np
import matplotlib.pyplot as plt
import soundfile as sf
from scipy.signal import hilbert

morse ={
	"" : "",
	"A" : ".-",
	"B" : "-...",
	"C" : "-.-.",
	"D" : "-..",
	"E" : ".",
	"F" : "..-.",
	"G" : "--.",
	"H" : "....",
	"I" : "..",
	"J" : ".---",
	"K" : "-.-",
	"L" : ".-..",
	"M" : "--",
	"N" : "-.",
	"O" : "---",
	"P" : ".--.",
	"Q" : "--.-",
	"R" : ".-.",
	"S" : "...",
	"T" : "-",
	"U" : "..-",
	"V" : "...-",
	"W" : ".--",
	"X" : "-..-",
	"Y" : "-.--",
	"Z" : "--..",
	"1" : ".----",
	"2" : "..---",
	"3" : "...--",
	"4" : "....-",
	"5" : ".....",
	"6" : "-....",
	"7" : "--...",
	"8" : "---..",
	"9" : "----.",
	"0" : "-----",
	"." : ".-.-.-",
	"," : "--..--",
	":" : "---...",
	"?" : "..--..",
	"'" : ".----.",
	"-" : "-....-",
	"/" : "-..-.",
	"@" : ".--.-.",
	"<BT>" : "-...-"
}

morse_inv = dict((v,k) for (k,v) in morse.items())

filename = sys.argv[1]
nstart   = int(sys.argv[2])
nframes  = int(sys.argv[3])

wav, fs  = sf.read(filename,start=nstart,frames=nframes)
analytic_signal = hilbert(wav)
env = np.abs(analytic_signal)

th = 0.5*max(env)
env[env<th] = 0
env[env!=0] = 1

t_up = 0
t_down = 0
up = []
down = []
string = ''
t_ws = 5000
t_ls = 2500
t_dd = 2500

for t in range(1, len(env-1)):
	if env[t-1] == 0 and env[t] == 1:
		t_up = t
		t_down_duration = t - t_down
		down.append(t_down_duration)
		if t_down_duration > t_ls:
			print(morse_inv[string], end='')
			string = ''
		if t_down_duration > t_ws:
			print(' ')
	if env[t-1] == 1 and env[t] == 0:
		t_down = t
		t_up_duration = t - t_up
		up.append(t_up_duration)
		if t_up_duration < t_dd:
			string = string + '.'
		else:
			string = string + '-'

print('')

$ ~/anaconda3/bin/python test4.py ~/40wpm.wav 0 500000

-...-
-. --- .--
....- -----
.-- .--. --
-...-
- . -..- -
.. ...
..-. .-. --- --
.--- .- -. ..- .- .-. -.--

$

import sys
import numpy as np
import matplotlib.pyplot as plt
import soundfile as sf
from scipy.signal import hilbert

filename = sys.argv[1]
nstart   = int(sys.argv[2])
nframes  = int(sys.argv[3])

wav, fs  = sf.read(filename,start=nstart,frames=nframes)
analytic_signal = hilbert(wav)
env = np.abs(analytic_signal)

th = 0.5*max(env)
env[env<th] = 0
env[env!=0] = 1

t_up = 0
t_down = 0
up = []
down = []
string = ''
t_ws = 5000
t_ls = 2500
t_dd = 2500

for t in range(1, len(env-1)):
	if env[t-1] == 0 and env[t] == 1:
		t_up = t
		t_down_duration = t - t_down
		down.append(t_down_duration)
		if t_down_duration > t_ws:
			print(string)
			string = ''
		elif t_down_duration > t_ls:
			string = string + ' '
	if env[t-1] == 1 and env[t] == 0:
		t_down = t
		t_up_duration = t - t_up
		up.append(t_up_duration)
		if t_up_duration < t_dd:
			string = string + '.'
		else:
			string = string + '-'

print('')

Once you get the envelope, it is easy to draw a histogram of the Dot/Dash durations.

import sys
import numpy as np
import matplotlib.pyplot as plt
import soundfile as sf
from scipy.signal import hilbert

filename = sys.argv[1]
nstart   = int(sys.argv[2])
nframes  = int(sys.argv[3])

wav, fs  = sf.read(filename,start=nstart,frames=nframes)
analytic_signal = hilbert(wav)
env = np.abs(analytic_signal)

th = 0.5*max(env)
env[env<th] = 0
env[env!=0] = 1

t_up = 0
t_down = 0
up = []
down = []

for t in range(1, len(env-1)):
	if env[t-1] == 0 and env[t] == 1:
		t_up = t
		t_down_duration = t - t_down
		down.append(t_down_duration)
	if env[t-1] == 1 and env[t] == 0:
		t_down = t
		t_up_duration = t - t_up
		up.append(t_up_duration)

plt.figure(1)

plt.subplot(221)
plt.plot(wav, color='green')
plt.title("input signal")

plt.subplot(222)
plt.plot(env, color='red')
plt.title("binary signal")


plt.subplot(223)
plt.hist(up, bins=50, color="orange")
plt.title("Dot/Dash Histogram")
plt.grid(True)

plt.subplot(224)
plt.hist(down, bins=50, color="blue")
plt.title("Spaces Histogram")
plt.grid(True)

plt.subplots_adjust(hspace=0.5, wspace=0.35)
plt.show()

The envelope of the signal is obtained by computing the magnitude of the analytic signal.

import sys
import numpy as np
import matplotlib.pyplot as plt
import soundfile as sf
from scipy.signal import hilbert

filename = sys.argv[1]
nstart   = int(sys.argv[2])
nframes  = int(sys.argv[3])

wav, fs  = sf.read(filename,start=nstart,frames=nframes)
analytic_signal = hilbert(wav)
env = np.abs(analytic_signal)

plt.plot(wav, color='green')
plt.plot(env, color='red')
plt.title("input signal")
plt.xlabel("time")
plt.ylabel("amplitude")

plt.show()

In my previous aritcle, Analytic Signals (2), the same thing is done using the C language.

A short Python program to show the waveform from a sound file.

import sys
import numpy as np
import matplotlib.pyplot as plt
import soundfile as sf

filename = sys.argv[1]
nsbegin  = int(sys.argv[2])
nsend    = int(sys.argv[3])

wav, fs  = sf.read(filename)

plt.plot(wav[nsbegin:nsend], color='green')
plt.title("input signal")
plt.xlabel("time")
plt.ylabel("amplitude")

plt.show()