Working on "Don't force it" video clip

Hi everyone,

Well… this was quite a challenge. Obstacles at every step.

It started with a slightly provocative storyline, which the generators kept flagging as violating their rules. I had to switch to a more indirect, almost Aesopian way of telling it.

Then came the lip-sync problem. The voices from ElevenLabs are far more advanced than those built into the video generator—their delivery is more subtle and nuanced. But after a lot of painful back and forth, I had to bite the bullet, drop the nuance, and use the generator’s own voices instead. The character acting in a five-dollar generator—the level I can currently afford—leaves a lot to be desired anyway. Even the simplest things often come out distorted or overacted. No matter how much you push, it keeps producing something awkward. So, the choice felt like an election between bad and horrible.

By the way, ElevenLabs reached out with an offer—they now support Seedance 2.0 video with synchronized voice. That would be a serious jump in production quality. But it’s too early for that. First, I need to squeeze everything I can out of the cheaper tools, gain experience, and only then move up.

Then came the issue of complex prompts that the generator simply couldn’t handle. For example, the scene where the mermaid creates a vortex by circling in the water, eventually pulling the surfer in. That turned into a long series of absurd failures.

In the end, I had to admit the obvious again: with my current level of experience and budget tools, my capabilities are limited. So, where I couldn’t really “show” I went back to using a narrator.

There were also plenty of seemingly small but persistent problems. For example, static water—as if it were the “dead water” from fairy tales. Maintaining character consistency was another ongoing battle.

Anyway, here’s what came out of it.