diff --git a/.gitignore b/.gitignore
index 321fb93..5d30fb2 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,8 +1,2 @@
 fft_calc
-sin01_fft.txt
-sin02_fft.txt
-sin03_fft.txt
-sin04_fft.txt
-sin800_fft.txt
-sin2367_fft.txt
-
+*_fft*
diff --git a/README.md b/README.md
index 7870b88..9455291 100644
--- a/README.md
+++ b/README.md
@@ -19,8 +19,11 @@ except sin03 which is -1.1...1.1.
  * sin2367 - 4s sine wave 2367 Hz
 
 # Build
-g++ -o fft_calc main.c fft.c
+`g++ -o fft_calc main.cpp fft.c`
 
 # Usage
-./fft_calc signal/sin01 signal/sin02
-./fft_calc signal/sin03
+`./fft_calc signal/sin01.txt signal/sin02.txt` \
+`./fft_calc signal/sin03.txt`
+
+To use the additional python script for plotting (requires python 3, matplotlib and numpy): \
+`python3 plots.py sin02`
diff --git a/main.cpp b/main.cpp
index ddac88c..a2bbd7e 100644
--- a/main.cpp
+++ b/main.cpp
@@ -73,12 +73,12 @@ int main(int argc, char *argv[]) {
             
             while (fgets(line, SIZE_OF_LINE, fp) != NULL)
             {
-                signal[i++] = atof(line);
+                signal[i++] = atof(line); 
             }
             
             if (ferror(fp)) perror("Error: ");
             
-            printf("Read %u lines\n", i);
+            printf("Read %u lines\n", i); // entire signal read into memory
             fclose(fp);
 
             /*****************************************************************************
@@ -210,6 +210,52 @@ int main(int argc, char *argv[]) {
 
                 fclose(fp2);
             }
+            
+            strcpy(strstr(res_file_name, ".txt"), "_only_windows.txt\0");
+
+            fp2 = fopen(res_file_name, "w");
+            
+            if (!fp2)
+            {
+                printf("Failed to open %s!\n", res_file_name);
+            }
+            else 
+            {
+                printf("Writing results to file %s\n", res_file_name);
+
+                // Print header row.
+                fprintf(fp2, "xaxis_fq\t");
+                
+                m = calc_power(SIGNAL_BUF_SIZE);
+                for(k = 1; k <= m; k++)
+                {
+                    if (k == m)
+                    {
+                        fprintf(fp2, "win%u\n", k);
+                    }
+                    else
+                    {
+                        fprintf(fp2, "win%u\t", k);
+                    }
+                }
+
+                // Print results to file. Full signal and each window result in separate column.
+                for (k = 0; k < MOVING_WINDOW_SIZE / 2; k++)
+                {
+                    fprintf(fp2, "%lf\t", (fq_step_win * k));
+                    
+                    for ( m = 0; m < num_windows - 1; m++)
+                    {
+                        fprintf(fp2, "%lf\t", fft[k + (MOVING_WINDOW_SIZE / 2) * m]);
+                    }
+                    
+                    fprintf(fp2, "%lf\n", fft[k + (MOVING_WINDOW_SIZE / 2) * m]);
+                }
+            
+                fclose(fp2);
+            }          
+            
+
         }
     }
 	
diff --git a/plots.py b/plots.py
new file mode 100644
index 0000000..1b03728
--- /dev/null
+++ b/plots.py
@@ -0,0 +1,43 @@
+import sys
+
+import matplotlib.pyplot as plt
+import numpy as np
+
+# Signal filename can be passed as: 
+# python3 plots.py sin02
+if len(sys.argv) > 1:
+    file = sys.argv[1]
+else:
+    file = "sin01"
+
+sampling_freq = 10000 # Hz
+sampling_length = 6.5 #s
+
+s = np.loadtxt(f'./signals/{file}.txt')
+t = np.linspace(0, sampling_length, len(s))
+
+# Visualize original signal and fft of full signal 
+fft_res = np.loadtxt(f'./signals/{file}_fft.txt', skiprows=1, usecols=[0,1])
+
+x_lim = -1
+x_axis_freqs = fft_res[0:x_lim, 0]
+energy = fft_res[0:x_lim, 1]
+
+fig, (ax0, ax1, ax2) = plt.subplots(3, 1)
+ax0.plot(t, s)
+ax1.set_title("FFT on full signal", y=1.0, pad=-14)
+ax1.plot(x_axis_freqs, energy)
+
+# Visualize fft of windows
+fft_windows_res = np.loadtxt(f'./signals/{file}_fft_only_windows.txt', skiprows=1)
+
+# take mean of all windows for each frequency in x-axis
+windows_mean = np.mean(fft_windows_res[0:-1, 1:-1], axis=1, dtype=np.float32)
+
+x_axis_freqs = fft_windows_res[0:x_lim, 0]
+energy = np.transpose(windows_mean)
+
+ax2.set_title("FFT on windows (avg)", y=1.0, pad=-14)
+ax2.plot(x_axis_freqs, energy)
+
+plt.show()
\ No newline at end of file