Added Python equalizer effect generator

* It uses input from microphone and displays colors based on the frequencies present in the input

host_python/README

@@ -24,3 +24,4 @@ DEPENDENCIES:
 
 demo.py requires PyGame, Python wrapper around SDL
 send_to_serial.py requires PySerial, Python library for work with serial ports
+equalizer.py requires PyAudio (wrapper around PortAudio) and a microphone

host_python/equalizer.py

@@ -0,0 +1,84 @@
+#!/usr/bin/python
+# vim:et:sw=4:ts=4:sts=4
+
+import pyaudio
+import struct
+import math
+import numpy as np
+
+import ledbar
+
+CHUNK_SIZE = 1024
+FORMAT = pyaudio.paInt16
+CHANNELS = 1
+RATE = 44100
+
+PIXELS = 10
+
+HISTORY_SIZE = 4
+SAMPLE_SIZE = CHUNK_SIZE*HISTORY_SIZE
+MIN_FREQ = 20
+MAX_FREQ = 12000
+FREQ_STEP = float(RATE) / (CHUNK_SIZE * HISTORY_SIZE)
+PIXEL_FREQ_RANGE = math.pow(float(MAX_FREQ) / MIN_FREQ, 1.0/PIXELS)
+
+p = pyaudio.PyAudio()
+
+stream = p.open(format = FORMAT,
+                channels = CHANNELS,
+                rate = RATE,
+                input = True,
+                frames_per_buffer = CHUNK_SIZE)
+
+def get_color(volume):
+    p = 1-15/volume
+    if p <= 0: return (0, 0, 0)
+    p *= p
+    if p <= 0.4: return (0, 0, p*2.5)
+    elif p <= 0.7: return (0, (p-0.4)*3.33, 1.0-(p-0.4)*3.33)
+    elif p <= 0.9: return ((p-0.7)*5.0, 1.0-(p-0.7)*5.0, 0.0)
+    else: return (1.0, (p-0.9)*10.0, (p-0.9)*10.0)
+
+l = ledbar.Ledbar(PIXELS)
+history = []
+window = np.array([0.5*(1-math.cos(2*math.pi*i/(SAMPLE_SIZE-1))) for i in xrange(SAMPLE_SIZE)])
+work = True
+
+try:
+    while work:
+        try: data = stream.read(CHUNK_SIZE)
+        except IOError: continue
+        if len(data) == 0: break
+        indata = np.array(struct.unpack('%dh'%CHUNK_SIZE,data))
+        history.append(indata)
+        if len(history) > HISTORY_SIZE: history.pop(0)
+        elif len(history) < HISTORY_SIZE: continue
+        fft = np.fft.rfft(np.concatenate(history)*window)
+        freq_limit = MIN_FREQ
+        freq = 0
+        i = 0
+        while freq < freq_limit:
+            i += 1
+            freq += FREQ_STEP
+        freq_limit *= PIXEL_FREQ_RANGE
+        pixel = 0
+        count = 0
+        volumes = []
+        while pixel < PIXELS:
+            total = 0.0
+            while freq < freq_limit:
+                total += abs(fft[i])**2
+                i += 1; count += 1
+                freq += FREQ_STEP
+            volume = (total/count)**0.5/SAMPLE_SIZE
+            volumes.append(volume)
+            freq_limit *= PIXEL_FREQ_RANGE
+            pixel += 1
+            count = 0
+        for pixel in xrange(PIXELS):
+            c = get_color(volumes[pixel])
+            l.set_pixel(pixel, c[0],  c[1], c[2])
+        work = l.update()
+finally:
+    stream.close()
+    p.terminate()