Important Disclaimer
This project is a non-commercial, fan-made endeavor and is not affiliated with or endorsed by the creators, producers, or distributors of Suzume no Tojimari. The music I've composed is intended solely for creative exploration and does not infringe upon the original series' copyright or intellectual property. I do not claim ownership of Suzume no Tojimari or its content, and this project is unlicensed by any official entity. Cover image used under fair use for commentary. All rights to the original series, its soundtrack and images belong to their respective copyright holders.
As part of my final-year undergraduate composition project, involving the Re-scoring of an episode of Sousou no Frieren I wanted to challenge myself by scoring high-intensity scenes, rather than the slow-paced, primarily peaceful scenes which I felt much more comfortable with writing.
Without any direct industry connections, I chose to re-score existing material as a way to practice composing to picture, while improving my ability to closely allign music with the narrative and the characters’ emotions.
I chose scenes from Suzume no Tojimari because of the strong sense of awe I felt when leaving the cinema after watching the film’s premiere. Kazuma Jinnouchi and Robert Hoffman’s original scoring of the film is incredibly well-constructed and beautiful. Their tight, non-overbearing approach to scoring left me deeply inspired - with the dream that one day I too can reach their level of technical ability.
My goal was never to surpass the original composers’ work, but rather to engage with it as a learning opportunity. Scoring music for fast-paced visuals was something I hadn’t previously practiced much, and so this was both invaluable, and a fun endeavor. When scoring my own soundtrack, I sought to loosely emulate the origina composers’ approach while crafting an original composition that served the same fundamental purpose: to further immerse the audience in the stunning visuals, storyline and voice acting through a closely linked soundtrack.
Technical Process
As with the other material from this project , I made use of AI audio seperation tools to isolate the dialogue and sound effects, using the model Bandit V2 , which is trained on the Divide and Remaster V3 dataset , a dataset specifically using clips of film, and was multilingual. This model proved significantly better results for my use case, and was able to produce much higher resolution results with significantly less artifacting.
Listen to the soundtrack:
View the project files:
To respect copyright, I am not distributing the original audio or video. If you wish to experience Jinnouchi and Hoffmann’s stunning original score (which I highly recommend), please support the official release.
Below you’ll find a link to my music-only project files (no dialogue or video included):
