https://reddit.com/link/1tjdo99/video/ac9k3fra6g2h1/player
Curious how everyone here is handling prompts for motion-heavy scenes, are you getting better results with detailed camera instructions or simpler cinematic direction?
Made this short sequence as an experiment with Seedance 2.0 and I was honestly more impressed by the motion consistency than the visual quality itself. The transitions between shots felt unusually smooth compared to most AI video models I’ve tested recently.
What surprised me most was how “directable” the output felt once the prompt pacing was right. Small wording changes completely changed the cinematic feel.
Curious what everyone else here thinks Seedance currently does best:
And where do you think it still breaks the most?
Made this scene with a vintage countryside setting mixed with a heavily customized hot rod, and somehow it feels like a frame from a movie that never existed.
Curious what people enjoy more in AI art lately:
perfect realism, or slightly chaotic “wrong timeline” aesthetics like this?
Instead of writing one long descriptive prompt, I broke the scene into timed cinematic shots with movement, reactions, and pacing cues.
What surprised me most was how much smoother the animation felt once the sequence had actual structure behind it. The comedic timing between the raccoon and squirrel ended up feeling oddly intentional.
Feels like timing and camera direction are starting to matter more than raw visual detail in AI video generation.
I’ve been testing different approaches to AI-generated videos and tried focusing less on “viral visuals” and more on scene composition, movement, and storytelling rhythm.
This concept was basically an attempt to recreate the feeling of a futuristic cinematic trailer from a dream-like idea in my head. The evolution of these tools is honestly wild.
Would love feedback on what could make it feel even more immersive.
Experimented with a cinematic cyberpunk scene featuring two futuristic samurai clashing in the middle of a neon-lit street during heavy rain.
Focused on:
I wanted it to feel like a frame from an unreleased sci-fi anime movie or AAA game cinematic. The reflections and rain physics were probably the hardest parts to balance naturally.
Curious what stands out more to you here:
Been testing different AI workflows lately and tried making a short gaming-style cinematic inspired by cyberpunk arenas, fast-paced PvP energy, and futuristic game trailers.
This one was generated using Seedance 2.0 Fast while experimenting with some of the free credits available for new users, so I mainly focused on testing how well it handles:
Honestly didn’t expect the scenes to come out this cinematic from quick generations. Still learning prompt structure and editing flow, but this experiment turned out pretty fun.
Would love feedback from fellow AI video creators — what would you improve or add to make it feel even more like a real game trailer? 👀
https://reddit.com/link/1t66rea/video/g7af7eea2pzg1/player
I've been playing around with AI video tools for a while now, but last night something clicked differently.
I made this short clip - a lone cyber-samurai standing in a rainy neon city, glowing blade, full cinematic vibe and when I watched it back I genuinely got chills. Not because it's perfect. But because six months ago I couldn't have made anything close to this.
I'm not a filmmaker. I don't have a studio or a team or any real budget. I'm just someone who has always had these visual worlds in my head with no way to get them out. And now, kind of suddenly, I can.
It's exciting and a little overwhelming at the same time. I keep thinking about all the people with incredible stories to tell who never had access to the tools to tell them. That feels like it's changing really fast.
Anyway, I'd love to hear from others who are experimenting with this stuff. What moment made you realize this technology was something genuinely different? Are you using it for personal creative projects or more for work? And what still frustrates you about where it's at right now?
No right answers. Just genuinely curious what people are experiencing out there.
I’ve been testing a new kind of AI platform where instead of paying upfront, you get credits and unlock more as you use it.
Here’s something I created with it 👇
I kind of like the idea, but I’m curious how others see it:
