Inspiration
Part of the team is formed by self-taught musicians, who know how hard it can to develop a practice regimen for more technical and less instantly rewarding skills, such as consistently working on your hearing and studying music theory. Having a circle of friends to learn together or just an online community of strangers to engage and compete with is a boon to any student, as Duolingo proved for learning languages. Bragging rights on Facebook also seem to factor in.
Why not make a similar platform that does what many other apps already do offline, but with all the hooks for our attention that are used by gaming and social networks? Where your online buddies can track and compete with your progress in interval and harmony recognition, rhytmic hearing and music theory? _ We'll be rich! _
What it does
It offers music lessons that develop music skills of aspiring musicians, followed by musical exercises for pitch recognition and playing/singing accuracy. Music. The app analyzes (monophonic) audio input and compares how well the played (or sung) pitches match to the pitches of the lesson loaded from a server. The players' score across exercises is tracked on the server and gamified. The app also offers the option to challenge a friend to an accuracy showdown, remotely or locally, providing shame or validation, both a welcome escape from the grey dullness of existence.
How we built it
We painstakingly cobbled the front end together in Java on the Android platform and the backend in Python on AWS. Git held the project and our minds in one (not necessarily wholesome) piece. The input pitch recognition was implemented with the TarsosDSP library for digital signal processing, and the lessons are implemented as MIDI files, from which its symbolic note and rhytm representations are converted to absolute time and frequency values. The front end and the server use regular post requests to communicate, since no real time processing is required. We haven't quite gotten around to implementing video streaming YET.
Challenges we ran into
We ran out of things to do much too quickly, serenely waiting for the hour of the deadline. Meditating. Softly humming. It was not as we anticipated, or at all as stated in this paragraph.
Unfamiliarity with the Android environment was certainly an obstacle, Jure pulling the bulk of the weight in making the app actually run, after bullfighting a Linux kernel - he had to reinstall it so mobile devices would even be recognized (we kid you not; this happened).
Accomplishments that we're proud of
Actually making a presentable app, in spite of falling short of our goals.
What we learned
Along with one half of the team becoming familiar with a new platform and one half with a new specific domai. In retrospect, we all also learned to appreciate the benefits of rigorous time planning and immediate bug testing.
What's next for Tunegether
Actual content in the form of lessons and exercises and the user leveling with points won for scoring high, like in a videogame. This is expected and the bare minimum for viability. Added value can be found in offering a platform where paying users can upload their performance as a private video to music teachers for asynchronous expert feedback - many self taught musicians can't find the time to schedule regular visits or even video conferences with an instructor, and teacher feedback is critical for .
Also, further developing challenges: a user will be able to play a melodic line and send it as a challenge for others to repeat. Also planned, a jam session mode, where every participant has a role in the band to fill, and the software mixes their contributions to build an online garage band demo. We should probably build a web app in HTML5, too.
After that, probably venture capital investment or a buyout.
Log in or sign up for Devpost to join the conversation.