This graduation project focuses on the following design challenge:
🔮
Design an AI-powered film music discovery tool to enable filmmakers to find suitable music that matches the vision for their film projects in a more efficient and intuitive way.
This digital portfolio contains information about the process and results of the graduation project. Refer to the reading guide to find information about the assignment, reflection and proof of the learning outcomes.
The project was eventually divided into five phases consisting of various steps/deliverables:
Process overview
The process was structured based on the double-diamond design process as each phase started with creating choices and eventually ended with making choices to continue with the next phase. The active application of the DOT framework and the CMD methods with their respective research strategies and methods (library, field, lab, showroom and workshop) helped to define and approach each step of the research process.
Each phase is described in a separate page that explains the research questions, methods, results, conclusion, next steps and sources. Click on the cards below to visit each phase.
The aim of this graduation project was to address the challenge faced by filmmakers in finding suitable music for their film projects. This process often proves time-consuming and exhausting, leading to settling for mediocre music search results, which impedes the creative vision of filmmakers. The following research question guided the project's execution:
"How can filmmakers find suitable music that matches the vision for their film projects in a more efficient and intuitive way?"
To answer this question, an AI-powered film music discovery tool was designed. Such a tool aims to propose suitable music to filmmakers and enhance the efficiency and intuitiveness of the music selection process. This solution incorporates advanced searching techniques powered by Cyanite AI, enabling filmmakers to specify the type of feelings they want their film projects to evoke. Filmmakers can write a prompt describing their desired vision and optionally add an existing song reference as inspiration. Providing the tool with this input allows it to suggest music that is suitable for the described film project. Additionally, the tool offers features such as precise music filters, AI music tips, and song mood analysis to assist filmmakers with choosing suitable music. With this functionality in place, the tool offers a more efficient and intuitive experience for filmmakers.
This solution was realised as a validated interactive wireframe prototype based on user insights, innovative ideas, and usability testing improvements, incorporated throughout the project's phases from problem discovery to realisation. The envisioned user experience of the tool was showcased in a UI design mock-up, while the AI music discovery functionality was demonstrated through a technical prototype. These final deliverables depict the user experience (UX), user interface (UI), and functionality, striving to show how the tool would look as a finalised product.
In conclusion, this project offers a promising solution to the identified problem by providing filmmakers with an efficient and intuitive tool for finding suitable music that aligns with their creative vision. The incorporation of AI-powered searching techniques and unique UX and UI has resulted in a product that has the potential to significantly enhance the music discovery process for filmmakers.
To ensure the success and effectiveness of the tool, the following consideration should be taken into account during the implementation:
It is crucial to have a large and diverse music database to offer a variety of suitable music results to filmmakers. Without such a music collection, the tool may suffer from repetitiveness, limiting its overall appeal and effectiveness. For instance, the royalty-free music platforms Artlist and Musicbed have more than 10,000 songs and new content is uploaded daily. Initially, the client suggested collaborating with starting music artists to acquire music for the tool. This collaboration can be expanded to include music labels in the industry, providing access to large libraries of diverse music. By establishing partnerships, the tool can offer a wide range of musical genres, styles, and moods, ensuring filmmakers find suitable music that aligns with their creative vision.
Furthermore, the following additional features were considered as means to expand the music discovery and suggestion capabilities of the tool:
The feature to generate new music can be a powerful option to allow filmmakers to obtain unique results that perfectly match their vision. The tool can use the provided input to compose music based on existing samples of instruments, vocals, genres, and other styles. When the music needs to adapt to changing situations, for instance, in dramatic film scenes, the tool can be used to adjust an existing song or create an entirely new sound. This feature was inspired by the existing tools Soundraw and Mubert, discovered toward the end of the graduation project.
Adding a community feature to the tool enables filmmakers to publicly share their music discovery journeys and review those of others. Users can contribute to the database by uploading original compositions or recommending existing tracks. This user-generated content can provide a valuable source of diverse and unique music options that may not be found elsewhere. Fostering a community of filmmakers who actively engage with the tool can lead to valuable collaborations, knowledge sharing, and a stronger ecosystem surrounding the film music discovery process.
In addition to the web application, a mobile application could be designed and developed to allow filmmakers to search for music in a more relaxed way. With the mobile application, filmmakers can explore new music for their film projects while, for instance, taking breaks from the creative process or traveling to filming locations. The mobile application will sync with the web application, ensuring access to the results from both platforms.
By considering the above future recommendations, this solution can become the preferred film music discovery tool.
