The Late 1900s

A dad rewinds to 1990s Hackney where three kids face trouble, rivalry, and messy street adventures as they chase a prized catalogue.A funny, tense, nostalgic journey that makes small quests feel epic.

Comment

Inspiration

The project began when my daughter asked what we used to do for fun in the “olden days.” That question brought back a whole world of memories. I have always been interested in the 1990s and have always wanted to make a film or a series that speaks to that era. The music. The clothes. The energy. The way we played. My daughter’s question connected directly to that long standing passion and finally gave me the perfect reason to bring it to life. I also wanted to test a wide range of new AI tools inside a proper narrative. On top of that I wanted to add a seasonal twist. A Christmas setting that felt warm, familiar, and slightly magical without turning into a typical holiday film. The mix of memory, experimentation, and a fresh Christmas angle shaped the heart of the project.

What it does

The film contrasts childhood today with childhood in the late 1990s. It follows the journey of three young characters who set out to get something simple but meaningful. A printed catalog. A small book filled with possibilities. Back then the ritual mattered. Flipping through pages. Circling things you knew you would probably never get. Imagining life with the items long before you could ever own them. That slow, screen-free dreaming carried weight. The film sets that feeling against today’s pace where tablets, apps, and constant notifications compete for attention and imagination. By moving between the past and the present the story shows how the value of play has changed and how something as ordinary as a catalog could once spark a whole world of adventure.

How we built it

I set out to test as many tools as possible and see how far each one could go. The project was built on discovery. As I created key moments the script evolved, and the voiceover shifted to match the emotional beats I found along the way.

I began with sound. I recorded the narration in ElevenLabs and locked the pacing before building any visuals. While shaping the audio I gathered images that captured the feel of the late 1990s and used them to create a storyboard.

For key frames I used text to image models. Google’s NanoBanana shaped most of the visual identity. Reve and Seedance supported specific shots where their strengths were needed. I also built custom GPTs to help me write clean, consistent prompts for both text to image and image to video.

For video I used Veo 3.1, Kling 2.5, Luma, Hailuo, Seedream, LTX Studio, WAN 2.2, Runway, and Mago. Each model brought something different to motion, lighting, and expression.

For aggregator platforms I used Freepik, FAL, Higgsfield, and Krea.

For node based workflows I used Weavy, ComfyUI, and Flora to build flexible pipelines and combine models in a controlled way.

For editing I used Adobe Creative Cloud and CapCut.

For music I sketched ideas in Logic Pro, took them into Suno for variations and textures, then bounced between Logic Pro and Suno until the track felt right. I finished everything in Premiere Pro and used Epidemic Sound and Splice for sound effects.

The whole pipeline blended experimentation with traditional filmmaking structure to keep the film warm, coherent, and emotionally grounded.

Challenges we ran into

Consistency was the biggest challenge. The project began before NanoBanana existed which made the early workflow unstable. When NanoBanana arrived it helped, but consistency was still difficult. Characters aged between shots. Props shifted. Clothing and lighting changed from model to model. Each tool excels in different areas which made blending them into one cohesive world demanding. Updates also landed mid production which added more variation. In the end I chose to compromise certain continuity details because finishing the film mattered more than achieving perfect consistency. I built this alone, and I know that once I assemble a team consistency will be far easier to control without compromise.

Accomplishments that we're proud of

I am proud that the final film feels emotionally honest and visually cohesive. The nostalgia feels real. The world feels lived in. The performances and tone feel warm and truthful. I am also proud that the film resonates with the people I made it for. My friends. My family. The people who grew up in the 90s and immediately recognise the details, the feelings, and the little rituals that shaped that era. I have always wanted to make something that celebrates the 1990s because I am obsessed with that decade. This film carries that love. It is also only a small glimpse of a much larger world I plan to build around this story.

What we learned

I learned to let go of perfection. To choose momentum over hesitation. To move forward rather than stay frozen trying to make every detail flawless. The story and the visuals are far from perfect, but they are a strong first draft, and finishing them taught me far more than polishing them endlessly ever could. I also learned the power of iteration. Each version, each test, each re prompt accelerated my thinking and pushed the film forward. Building, reviewing, adjusting, and continuing on made the process faster, clearer, and far more enjoyable.

What's next for The Late 1900s

I plan to grow the 1900s into a full micro series. More neighbourhood adventures. More imagination. More 90s culture rebuilt through modern tools. I also want to share the workflow and creative system so other filmmakers can benefit from the speed and clarity of this approach. The long term goal is to expand this world and explore more stories inspired by that era.

Built With

Share this project:

Updates