GAZE

The Idea: From a Box to a Window

This project started from an idea I had after my film "BOX." I wanted to keep using the frame concept, but this time as a window that could look into different worlds and eras. I realized that a window doesn't have to be a literal window. If you look at things from a certain perspective, any moment can be a frame, and we are the "viewers" watching the story happen.

Unlike "BOX," which was more of an experiment, "GAZE" adds a new layer. I wanted to add a story and a narration that could help connect the visuals, the music, and the overall meaning of the film. For me, "GAZE" is about the "process" of the world and how we watch it all unfold. It’s a chance to reflect on the past, the present, and the future. It tells a non-linear story to show how what happened before can affect what happens next.

How It Was Made: Music, AI, and a Bit of Strategy

My workflow always starts with a feeling, and for me, that feeling comes from music. I found a track, edited it down to under 60 seconds, and then brainstormed ideas that fit the mood.

Once I had the concept, I started planning the visuals. I used Gemini 2.5 Pro to build a Master Prompt. This process much faster than my last project. For all the visuals (both images and video), I used OpenArt. I wanted to experiment with their new model, Seedream 3.0, for the images, and then used the Hailuo 02 model to turn them into videos. The storyboard feature was very helpful for keeping track of all my prompts. After that, I did all the editing in Premiere Pro, adding the music, sound effects, and getting the pacing right.

The biggest challenge was the concept itself. I wanted every "window" to feel natural and not forced. But you can't know if a prompt will work until you generate it. To save credits, I decided to work with the results I got instead of trying for perfection every time. I always focus more on visual storytelling; the image quality can follow the technology and budget later. I generated more shots than I needed so I had more choices during editing. This was my strategy to avoid regenerating video, which costs a lot of credits and has no guarantee of success.

What I'm Proud of and What I Learned

I am very proud of the final shot, with the dandelion and the bird. It was a "first try" generation, and it perfectly captures the film's message of hope. Most of the shots were actually first tries, which was amazing. The scene with the queen in the throne room was the one that gave me the most trouble!

With this film, I learned so much about how fast I can edit and how important sound effects are. Choosing the right SFX and mixing them properly was a big effort for me, as I'm still learning.

I know there is still a lot to improve. I hope in the future I can use higher-resolution video. And for the story, I wish the dialogue could be more poetic. English is not my native language, so I chose simpler words. Maybe one day I will use Indonesian, but for now, the text-to-speech tools don't have a voice that I like.

But what this project taught me is that I want to focus on storytelling. Visual generation tools will always get better, but the ideas, the concepts, and the final edit, that comes from the human. For now, I think I will try something new, but if I return to the "window" concept, I want to tell an even deeper story.

Built With

• openart