I've been impressed with Kling AI's text-to-video and then its ability to extend those clips. I started with a 5 second generation, even though you can do 10 second generations on their paid plans. This allows me to experiment with more initial clips, then extend the ones I want. I realize image-to-video is most people's preferred method, but I feel like there's sometimes a "jump" at the beginning of image-to-video that somehow doesn't feel as fluid and natural to me. I didn't do any upscaling or image enhancement to this, it's straight out of Kling, then edited in CapCut to add the water sound effects.
Prompt: cinematic live action, POV riding a polar bear in Amsterdam along the bridge canal area
Water sound effects made in ElevenLabs:
Prompt: the sound of a bear swimming breast stroke in a river
Kling AI
https://klingai.com
ElevenLabs:
https://elevenlabs.io/
#kling #klingai #polarbear #aianimals #texttovideo #aivideo