For non-professionals who want to enjoy music only at the wheel or at home, the difficulty of recording music has always been a deterrent. Notation is not so easy to read, but even more difficult to write fluently, especially in the case of polyphonic sounds. Nevertheless, this is extremely important: many musicians say that knowledge of notation contributes to the creation of better music.
For these musicians, a relatively simple, but at the same time revolutionary and efficient, solution resembling Google Translate appeared, only for musical recording.
The developer of the application was the Swedish company DoReMirMusic Research AB. After all, musical ideas, songs or melodies can occur suddenly, and it is important to quickly record what comes to your mind before it disappears. Writing notes by hand or on a computer is a relatively slow process that takes a lot of time to properly master and requires analyzing musical syntax in terms of size, tempo, types of keys, and so on.
For many musicians, the opportunity to quickly move from directly music to writing notes was just a breakthrough in their work. Now they no longer need to find a solution to this problem: thanks to Swedish development, virtual assistant software for PCs/Macs and iOS applications is now available for free download or by monthly subscription.
ScoreCloud reminiscent of Amazon Alexa and Apple Siri Assistant has the function of instant translation: just sing or play an instrument, and the application will create a note based on which you can further develop the idea and create new arrangements and compositions within the application. In addition, the my.scorecloud web service allows you to share your notes with other musicians from around the world.
ScoreCloud – an application that will help beginners and amateurs, as well as professional musicians whose songs will be performed by other people or who want to develop them further from the music record. The complexity of music is not a hindrance. Cognitive modeling algorithms in ScoreCloud mimic how we perceive music, and aim to create not only accurate, but also easily readable notes. These algorithms can distinguish between voices in polyphonic singing and simplify scores as if a person was writing notes by hand. The need to adapt ScoreCloud to cope with the complexities of music prompted DoReMir to seek funding in the EU. The main problem of automatic note recording was the analysis of polyphonic sounds in which more than one note is simultaneously reproduced. A few years ago, the level of recognition of polymorphic tones was far from ideal and was insufficient to create notes based on the sounds of polymorphic instruments, such as piano or guitar.
Software developers wanted to see if combining modern technologies with cognitive modeling algorithms would help overcome the difficulties associated with automatic transcription of polyphonic music. EU funds gave us such a chance. Now it is safe to say that the polymorphic tone transcription system developed by the DoReMIR project (Automatic musical transcription of polyphonic sound) has exceeded all expectations.
This technology allows users to analyze music content with an unprecedented level of detail, which opens up amazing opportunities for them in the field of applications for searching for music content, music recommendation systems, and so on.