Light Breaks Through

• 

  Light Breaks Through

“Light Breaks Through”

is a cinematic music project that introduces HEDY, a fictional singer whose voice and presence embody strength, emotion, and light. Inspired by Amanda Gorman’s poem “The Hill We Climb,” the project explores hope and resilience in uncertain times — reminding us that true light is something we carry within.

Parts of the song were initially created with the help of AI-generated vocals, which were then reworked, re-arranged, and completed using classic studio production tools. Live-recorded guitars and organic textures were added to round out the sound, blending technology and human expression into one seamless experience.

The music unfolds like a single, flowing composition, designed to evoke an almost meditative, song-like atmosphere. Its emotional intensity is amplified through a distinct visual style — a black-and-white world with color accents that reflect fragility and quiet power.

Together, sound and image merge into an immersive narrative — establishing HEDY as both a musical and visual character who invites the viewer to feel, reflect, and let the light break through.

How it´s built

The music came first, developed through the process previously described — in creative exchange with Suno AI. For the video, the workflow began with generating key visuals in Midjourney, which served as the foundation for building the lipsync performances around HEDY. From there, the entire visual structure of the video took shape — blending metaphorical imagery with abstract dance sequences that embody the interplay of light and shadow.

The animation and compositing were created using Kling, VEO3, Higgsfield, and Seedance, while upscaling was done with Topaz. Editing and final assembly were completed in iMovie, combining all elements into a cohesive cinematic experience where technology, emotion, and storytelling converge.

Challenges I ran into

As the music was created before the visuals, envisioned as a blend of concept and performance video, translating the song’s wide dynamic range — from calm and atmospheric to powerful and intense over its more than four-minute journey — into a coherent visual narrative became one of the project’s greatest creative challenges.

Accomplishments that I´m proud of

I’m proud that I was able to give the character HEDY a believable and coherent voice — and a very distinctive song that doesn’t follow current musical clichés. It’s a piece that stands on its own, emotionally and artistically, while staying true to her unique identity.

What I learned

One year ago, working creatively with AI tools felt far more open and experimental. Now, I’ve noticed that many of today’s systems offer less freedom for abstract or artistic expression. They seem increasingly optimized to replicate reality, rather than explore the full creative potential of generative AI. For example, it has become surprisingly difficult to create convincing lipsync scenes without automatically generating a microphone in the frame — a small but telling sign of how these tools are drifting toward realism instead of imagination.

Built With

• higgsfield
• kling
• magnific
• midjourney
• nanobanana
• openart
• reeve
• seedance
• suno
• topaz
• veo3