Polymath: Revolutionizing Music Production with AI
Polymath is an innovative AI tool that has been making waves in the music industry. It utilizes machine learning algorithms to transform any music library, whether it's from a hard drive or YouTube, into a highly useful music production sample-library.
Overview
Polymath's capabilities are quite extensive. It can automatically separate songs into various stems like beats, bass, etc. It then quantizes them to the same tempo and beat-grid, for example, 120bpm. Additionally, it analyzes musical structure such as verse, chorus, etc., determines the key like C4, E3, etc., and also processes other information like timbre and loudness. All of this culminates in the conversion of audio to midi, creating a searchable sample library that is a boon for music producers, DJs, and ML audio developers.
When compared to traditional methods of creating sample libraries or working with music elements from different songs, Polymath offers a much more streamlined and efficient process. For instance, in the past, combining elements from different tracks to create a new composition could be a time-consuming and complex task. With Polymath, one can effortlessly pick a beat from a Funkadelic track, a bassline from a Tito Puente piece, and fitting horns from a Fela Kuti song, and seamlessly integrate them into a DAW in record time.
Core Features
The tool employs several neural networks for different tasks. Music Source Separation is carried out by the Demucs neural network. Music Structure Segmentation/Labeling is done with the sf_segmenter neural network. Music Pitch Tracking and Key Detection are performed by the Crepe neural network, while Music to MIDI transcription is handled by the Basic Pitch neural network. Music Quantization and Alignment are taken care of by pyrubberband, and Music Info retrieval and processing is done with librosa.
These features work in harmony to provide a comprehensive music processing solution. For example, the accurate pitch tracking and key detection ensure that when combining different musical elements, they are in harmony with each other. The quantization feature allows for all the elements to be on the same tempo and beat-grid, making the final composition sound cohesive.
Basic Usage
Using Polymath is relatively straightforward. First, you need to have the necessary software installed on your system, including ffmpeg, and a python version between >=3.7 and <=3.10.
To add songs to the Polymath Library, you can add YouTube videos (which will be auto-downloaded) by running the command like python polymath.py -a n6DAqMFe97E. You can also add audio files (wav or mp3) using commands such as python polymath.py -a /path/to/audiolib/song.wav. Multiple files can be added at once as well.
Once the songs are in the library, they are automatically analyzed, which takes some time. After that, you can quantize the songs to a specific tempo, like 120 BPM, using commands such as python polymath.py -q n6DAqMFe97E -t 120. You can also search for similar songs based on a specific song in the library and even convert the audio to MIDI using the appropriate commands.
In conclusion, Polymath is a powerful AI tool that has significantly simplified and enhanced the music production process, offering a wealth of features and an easy-to-use interface for music enthusiasts and professionals alike.
