Music visualizer video maker

Make a video for a track with custom visualizers based on loudness and frequency

Example video

https://youtu.be/I2Sk8WoMwIU

Environment

Run to create environment

pip install virtualenv

python -m virtualenv <environment-name>

Run to activate environment

environment\Scripts\activate

To deactivate run: environment\Scripts\deactivate.bat

Run to install packages

pip install -r requirements.txt

How to make your own visualizer

Step 1

Imports

import gizeh as gz
import numpy as np
from PIL import Image, ImageDraw
from Visualizer import Visualizer
from Options import Options

Step 2

Create a class for the visualizer in ./visualizers/ directory inheriting Visualizer class

class MyCustomVisualizer(Visualizer)
{
  
}

Step 3

Create a constructor with one mandatory parameter audioFile and your custom parameters, for example: x, y, color, scale...

def __init__(self, audioFile, x, y, color, scalePercentage)
{
  super().__init__(audioFile)
  
  self.X = x
  self.Y = y
  self.Color = color
  self.Scale = int(scalePercentage * Options.HEIGHT)
}

Variable	Description	Type	Required
audioFile	Path to an audio file to use for video generation	str	Yes

Step 4

Create a method Render with one mandatory parameter time and optional keyword arguments **kwargs

def Render(self, time, **kwargs)
{

}

Variable	Description	Type	Required
time	Represents passed time in seconds	float	Yes

Step 5

Inside Render we will be creating a circle which radius changes based on music loudness, but you can create any shape or object as long as it has some parameters that can change based on loudness or frequency

Here we are using gizeh library for vector graphics

Note
PIL usually produces higher quality frames but gizeh is faster to render

radius = self.Loudness[self.GetMoment(time)] * self.Scale

surface = gz.Surface(Options.WIDTH, Options.HEIGHT)
circle = gz.circle(radius, xy=(self.X, self.Y), fill=tuple(x / 255 for x in self.Color))
circle.draw(surface)
return surface.get_npimage(transparent=True)

Example with PIL

radius = self.Loudness[self.GetMoment(time)] * self.Scale

image = Image.new("RGBA", (Options.WIDTH, Options.HEIGHT), (0, 0, 0, 0))
draw = ImageDraw.Draw(image)
draw.ellipse((self.X - radius, self.Y - radius, self.X + radius, self.Y + radius), fill=self.Color)
return np.asarray(image)

1. Loudness is a numpy array which contains music loudness data. Its length is number of frames
2. GetMoment(time) returns current frame number

Step 6

Here we will create a visualizer - a white circle in the middle of the screen that responds to music

from Video import Video
from visualizers.MyCustomVisualizer import MyCustomVisualizer
from Options import Options  

if __name__ == "__main__":
  audioFile = "audio/audio.wav"
  video = Video(audioFile, output="video.mp4")
  
  myCustomVisualizer = MyCustomVisualizer(audioFile,
                                          x=Options.WIDTH / 2,
                                          y=Options.HEIGHT / 2,
                                          color=(255, 255, 255),
                                          scalePercentage=0.9
                                          )
  
  circle = video.ShowTransparent(myCustomVisualizer.Render)
  
  video.Generate([circle])

Video object takes the path to the audio file and output name
Video.ShowTransparent takes a render method and optional keyword arguments, and makes background transparent
Video.Generate takes a list of objects to render

The result is saved in a file named output

Existing visualizers: