What happens when an unyielding sailor meets an overly cautious AI robot? Step into the Half-Pipe, a weathered tavern resting on the banks of the Danube, where centuries of harbor tradition collide with the stubborn algorithms of modern technology. When the establishment introduces a "Smart Waitress" to streamline the bar, the local patrons quickly find themselves in a hilarious, high-stakes standoff for the right to order a simple cold beer. Based on the short story by Erdős Sándor, "Artificial Intelligence in the Pub" is a comedic, atmospheric clash between human grit and corporate tech absurdity. The Full Short Movie Coming Soon.

Hey @everyone! We’re launching **Dubbing v2** - our new AI dubbing model built to make localized content sound natural, expressive, and true to the original performance. For the first time, the **emotion, tone, pacing, and delivery** of the original speaker can carry across languages, not just the words. ## ✨ **What’s New** Dubbing v2 conditions directly on the **original audio performance**, not only the transcript. That means translated speech can preserve: • Emotion • Timing • Emphasis • Delivery style • Natural pacing No more flat or disconnected dubs. ## 🌐 **Localization That Sounds Native** Dubbing v2 adapts phrasing for spoken delivery across **90+ languages**, while keeping speech aligned with the original timing. Its sync-aware translation logic helps starts, stops, and pacing match the source content out of the box. ## 🚀 **Built for Global Content** For creators: localize YouTube videos with natural-sounding dubs. For marketers: scale ads and campaigns across markets without rebuilding from scratch. For studios and broadcasters: use ElevenProductions for professional workflows with translators, casting, and mixing. ## 🎁 **Free Trial Access** For the next 7 days: • Free users get **1 minute** of dubbing • Starter plans get **15 minutes** • Creator+ plans get **30 minutes** ## 🔗 **Get Started** Dubbing v2 is available now in **ElevenCreative** and through **ElevenProductions**. Try it here → https://elevenlabs.io/dubbing

Seedance 2.0 by Dreamina AI only lets you upload one reference video. That becomes a problem the moment you also need an empty video carrying audio for lipsync. The reference slot is gone. I tested two workarounds on a downloaded dance clip and I want others to try the same, because one run is not a conclusion. The setup: - Turn the source into a storyboard (extract frames, build a grid image, feed it as a visual reference). - Compress the 10 second source into a 2 second video at 8 fps (Seedance's minimum duration), sampling frames evenly to keep the motion arc intact. Both transformations were done locally with Claude Code driving ffmpeg. No cloud step, no upload round trip. Claude Code wrote the frame extraction, the grid composition, and the resample-and-rewrap into 8 fps in one pass. That part of the pipeline took minutes to set up and runs on any laptop. Same prompt for both runs: two humanoid robots in a junkyard doing the exact dance movement. What I saw: - The compressed 2 second video works. The robots dance. Timing carries over, body mechanics translate, choreography is recognizable. - The storyboard reference is not following the storyboard. The model picks up posture cues but the motion is gone. The dance turns into a slideshow of unrelated poses stitched together. My working theory: motion is not pose. Storyboards freeze geometry and kill the temporal signal. Frame-sampled short videos keep both, even crushed to 2 seconds at 8 fps. If this holds, the practical implication is that you can buy back your second upload slot by shrinking the reference, but only along the video path. Storyboards would not be a substitute. But I have run this twice, with one clip, on one prompt. That is not enough. If you work with Seedance, please try it. Pick any motion-heavy reference, run both versions (storyboard grid vs. compressed micro-video at 8 fps), keep the prompt identical, and post your results. I am especially curious about:

Premiere Topaz Bridge - https://www.youtube.com/watch?v=lafitZNvHz8 It will be free, just to be approved by AEScript site. So, soon, in a couple of days. Need to have Topaz Video Desktop license.

https://filmfreeway.com/bucharestaifilmfestival - open to submit until May 15, 2026