Audio-Visualizer (AudiViz)

Takes WAV file and generates visualization as MP4.
You can customize color gradient & number of sound bars, inspect audio file's frequency & volume graphs with pyFFTW & PyQT Graph.

Usage

Change this to your audio file name (works only with WAV) and run the code!
audioFile = "MyAudioFile.wav"

Optional settings

If you want to, you can set resolution of video

w, h = 1920, 1080

You can apply a filter to the input signal and smoothen the audio curve for better visuals (this reduces the distance between maximum and minimum points of frequency magnitudes)

sh *= (1 - (sh/h)**(1./2)) * 3. # 1./2 or 1./8 or 1./312 - n'th root: no limit

You can change number of bars

datas = visualize(100) # 100 bars will be drawn

If you want to see a larger protion of the frequency spectrum, increase f_ratio (min 0.0, max 1.0)

f_ratio = 0.2

You can change update frequency of frequency spectrum to get a smoother video (0.03 is actually perfect. Higher values consume more RAM)

update_periodicity = 0.03

You can change background color of outout video (default: black)

win.fill((0, 0, 0))

Dependencies

• pygame (to visualize bars & frequencies)
• numpy (general math stuff)
• pyfftw (to make fourier analysis of volume graph to get frequency graph, MUCH faster than numpy)
• pyqtgraph (to plot the frequency & volume graphs)
• pillow (to create images)
• moviepy (to create MP4 clip)

Cons (These could be optimized)

• Becomes very RAM intensive, if audio files are longer than 2 minutes. It even gives RAM error sometimes. In that case, it is recommended to split audio file and visualize it in segments.
• Requires some relatively hard-to-install modules

Example

Audio visualizer of these videos was generated with this code: Check this music video on YT

Visualizer_Sesim.mp4