Behind the Scenes
From a local experiment on a laptop to a full browser-based music visualizer for creators worldwide — this is how Muvize came to life.
Follow us on @muvizeofficial on Instagram
The very first attempt at turning audio into synchronized visual animations — transforming raw sound into something you can see.
First integration of a large language model into the video creation workflow, automating parts of the creative process with AI assistance.
Created the first timeline interface and a lightweight video editor — laying the groundwork for a proper editing experience.
Successfully recreated the iconic Trap Nation audio spectrum visualization — the style that made music visualizers famous on YouTube.
First implementation of particle systems and depth-based image animation using Immersity AI — bringing static images to life with parallax motion.
Deep dive into depth image animation through the Depthflow project — an open-source alternative to the paid Immersity AI service.
Built a realistic rain overlay effect to add dramatic atmosphere to music visualizations — complete with droplets and splash dynamics.
Everything started with Remotion.dev — a React framework for creating videos through code. This was the foundation that made everything else possible.
Implemented automatic caption generation using OpenAI's Whisper — producing word-level synchronized subtitles similar to TikTok's native captions.
After one month of building a music video generator with Remotion.dev, the entire codebase was open-sourced for the community.
A look at how music videos were being made at this stage — entirely manual, running only on a local laptop with no cloud infrastructure.
Experimented with AI-generated music and learned how to combine it with the visualizer — producing surprisingly good results.
Discovered alternative approaches to subtitle editing and explored how to integrate proper subtitle workflows into the visualizer pipeline.
After a long hiatus from music video experiments, the first product shipped under the hibuno.com domain — Visualizer Studio v1.0 was officially live.
Added a proper timeline interface with support for multiple audio tracks — enabling complex multi-song visualizer projects.
Major release with auto captions, smooth transitions between scenes, animation presets, and synchronized lyrics display — a complete creative toolkit.
Published the first end-to-end demonstration showing how Visualizer Studio v2.1 works from audio import to final video export.
Shipped a variety of new visual effects, creative features, and transformation tools — giving creators more control over their output.
Redesigned the interface with a mobile-first approach — making it possible to create music visualizers from any device, anywhere.
Visualizer Studio reached its first 100 registered users — a meaningful milestone validating the product direction.
Continued refinement of the mobile interface — fixing edge cases and ensuring a smooth experience across different screen sizes.
Began migrating Hibuno Visualizer Studio into a standalone product with its own identity — Muvize (Music Visualizer).
Revisited the rain effect from earlier experiments — now with configurable lightning strikes and more realistic particle behavior.
The official launch of Muvize v1.0 — same core concept as Visualizer Studio but with significant performance improvements and a fresh UI.
Published a walkthrough demonstrating the complete Muvize workflow — from importing audio to exporting a finished music video.
Added a built-in media library and the ability to import songs directly from Spotify and YouTube — streamlining the content creation flow.
Introduced two AI-powered features: automatic vocal removal for creating instrumentals, and a karaoke generator with synchronized lyrics.
Follow along as we build new features and share the story behind every update on Threads.