In an interview published today by UnderScores, Arnaud Toulon describes joining Arco before there was a fully formed screenplay. Director Ugo Bienvenu had already given the music a demanding role: it should carry the film’s metaphysical dimension without pushing the story toward conventional magic or fantasy.
Toulon developed themes as the script took shape, then sent mockups against early storyboards and animatics. In some passages the score followed picture. In others, the edit was shaped around the music.
That makes the interview useful beyond the film itself. It documents what early composer involvement can actually change in animation.
The score had a purpose before it had fixed timings
The first direction was dramatic rather than instrumental. Bienvenu wanted spiritual evocation and a sense of timelessness. The human voice emerged early as a way to connect the story’s different periods.
That sequence matters. A composer can explore the film’s musical logic before every cue has a final duration. A dramatic function can survive revisions to story and layout more easily than a narrowly imitated reference.
The interview does not present this as instant inspiration. One broad idea remained stable while many of its musical forms were tested, rejected or refined.
For producers, this is the practical value of an early brief. It gives the score a question to investigate instead of asking for finished music before the film knows what it needs.
The animatic made the conversation measurable
An abstract intention cannot sync a sequence by itself. The animatic supplied duration, cutting, movement and provisional pace. Toulon says it gave him material to react to while the production still had room to change.
For School Ride, the sequence in which Iris and Clifford travel to school, the image was cut to the rhythm of the piece. According to the interview, that relationship then changed very little. The final sequence and its cue, Rainbows, also found a relatively stable exchange between music and edit.
This does not mean score should always lead picture. It shows that animation can let both departments negotiate rhythm before each shot becomes expensive to revise.
A musical acceleration can suggest how long a gesture needs. A long phrase may stop editorial from overcutting an emotional transition. Those are structural choices, not decorative ones.
A long schedule can create endless revision
Toulon calls the animation timeline both a luxury and a curse. It creates time to search, but also makes everything feel available for another pass. He recalls Bienvenu occasionally telling him he was not allowed to revise a cue again.
That is the necessary counterpoint to the romantic version of early collaboration. Starting early does not eliminate revisions. Without approval gates, it may multiply them.
The film’s opening went through several directions. Choir felt too explicitly religious. A more Danny Elfman-influenced approach pushed the film toward a fanciful colour the team did not want. Accordion and whistling only solved part of the problem. The opening unlocked after the wider score language had developed.
By contrast, Toulon describes the idea that became Iris’s theme as an early version that immediately caught the team’s attention. Arco’s theme required many more iterations. The same workflow does not produce a predictable version count.
Temp music was still difficult to replace
The production used relatively little temp music, including for animatic presentations to potential investors. Yet Toulon identifies sequences where those references became hard to replace, such as the run toward the school roof and the cave scene after the fire.
The example is precise. Temp does not need to cover the whole film to become powerful. If one important sequence is repeatedly edited and screened with a track, its particular motion can become the baseline.
Bringing the composer in early can reduce that exposure without removing it. More importantly, it gives the production another option: original mockups can enter the animatic soon enough for picture and music to develop together.
Virtual freedom still meets production limits
Toulon also describes working with orchestrator Moritz Reich to translate the mockups into the available ensemble. A virtual session can imply an unlimited number of players. The recording plan had twenty-four strings.
The team recorded some melody and accompaniment layers separately, then combined them. That approach can build a larger texture, but it requires planning before the session. It is not the same as having twice as many players performing together in the room.
This stage completes the early-development story. Starting from script and animatic opens possibilities. Production must still turn those possibilities into orchestration, session time, takes and mix options that exist in the real budget.
What this case actually establishes
The Académie des César credits Toulon with the original music for Bienvenu’s Arco, which won the 2026 César for original score. The soundtrack album released by Hamburger Records includes the score and Clouds Away, the original song by November Ultra and Toulon.
The award does not prove that early involvement automatically creates better music. The interview supports a narrower and more useful conclusion. Toulon joined at the origin of the project. The animatic became a shared working surface. Some sequences allowed the cut to answer the score rather than treating music as material added after picture lock.
For an animation producer, the takeaway is not to demand final cues before animation begins. It is to invite the composer while picture can still listen, then define approval points so that the freedom to search does not become permission to restart forever.