Music Generation from RGB Images

RGBeats is an Android application that generates musical tones based on a series of images.

This project was completed during TartanHacks 2013, which is a 24 hour hackathon at Carnegie Mellon University each Spring semester.

Project Description

The musical tones are generated by layering instruments/sounds based on a series of images.

Once a photo is taken using the camera or chosen from the user's gallery, the average R, G, and B values for the image are calculated and the color with the largest average value determines the musical instrument/sound that is played.

  1. The first photo determines the underlying rhythm and bass style.

    • Red: rock rhythm and bass

    • Green: electric rhythm and bass

    • Blue: jazz rhythm and bass

  2. The second photo determines the accompaniment style.

    • Red: rock accompaniment

    • Green: electric accompaniment

    • Blue: jazz accompaniment

  3. The third photo determines the lead style.

    • Red: rock lead

    • Green: electric lead

    • Blue: jazz lead

By layering these musical instruments and sounds together, we can generate a total of 27 unique musical tones, which were prerecorded and stored in .mp3 files.

Examples of musical tones:

More Information