Attempts to AI Generated shoeplay videos

[highheelhumper]

wow...how long did it take to make these nearly 4 minutes? What did you use?

I guess this had been a full working day.
I started with a single pic as a first frame, but enhanced this pic with "Paint Shop" to give her really high arches.
Then I let Grok4 create 6s clips (not all of them are useable. I guess I created "only" 60, most of them had been successful).
I use Camtasia to cut and zoom and to include the audio.
The audio comes from elevenlabs (text to speech) with a quick clone.

[nyllover]

Wow...impressive work and dedication, really. I'm speechless

[highheelhumper]

alternative download: https://k2s.cc/file/5a74a84e7e498/AI_heelpoparches3.mp4

[highheelhumper]

There is a little present for our ADMIN, but also for all of you to hopefully enjoy ;-)

[nyllover]

awwww Thanks!

[highheelhumper]

A HIGH HEEL DREAM

Includes NO nudity.
Visual novel in German language. 10 min talking about High Heels.
Introduction - high heel show - lots of cum on her feet and shoes - and a final surprise





Download link: https://k2s.cc/file/ac2aec0284c52/AI_dream.mp4
Streaming link: https://sendvid.com/xl04dniq

[highheelhumper]

[highheelhumper]

Some more heelpopping fun:

[highheelhumper]

[highheelhumper]

[nikos99]

I've been experimenting various workflows to generate dangling videos, so far I got the best results and most cost effective by writing a prompt to describe the video I want to generate and:
- Using that same prompt to generate an image with Nano Banana Pro
- Use that image as a start frame to generate a video with Grok Imagine using the prompt

Sora 2 is globally better at doing the "candid" look and feel as well as ambient sounds but is kinda hit and miss for the dangling action itself, it's also very costly and has been axed by OpenAI so it won't evolve anymore. Grok Imagine is usually bad when trying to generate a video from a text prompt only however when provided with a starting image it gives surprising results! Here are some examples, some of them have very natural looking dangling action:

https://grok.com/imagine/post/c8fe03a1- ... atform=web

https://grok.com/imagine/post/3c523482- ... atform=web

https://grok.com/imagine/post/70ce0828- ... atform=web

https://grok.com/imagine/post/f3e8c87a- ... atform=web

https://grok.com/imagine/post/dc81d417- ... atform=web

https://grok.com/imagine/post/55b01416- ... atform=web

https://grok.com/imagine/post/e57cc1fd- ... atform=web

https://grok.com/imagine/post/f1b90eaf- ... atform=web

https://grok.com/imagine/post/15f8a7f3- ... atform=web

https://grok.com/imagine/post/64d954bb- ... atform=web

https://grok.com/imagine/post/d815e53c- ... atform=web

https://grok.com/imagine/post/3ae45550- ... atform=web

https://grok.com/imagine/post/e54eb21e- ... atform=web

https://grok.com/imagine/post/5fb41e63- ... atform=web

https://grok.com/imagine/post/93e15eee- ... atform=web

https://grok.com/imagine/post/818f3e7a- ... atform=web

https://grok.com/imagine/post/b1700786- ... atform=web

https://grok.com/imagine/post/d26a8cff- ... atform=web

https://grok.com/imagine/post/fe073106- ... atform=web

[pumplover]

I've been experimenting various workflows to generate dangling videos, so far I got the best results and most cost effective by writing a prompt to describe the video I want to generate and:
- Using that same prompt to generate an image with Nano Banana Pro
- Use that image as a start frame to generate a video with Grok Imagine using the prompt

Sora 2 is globally better at doing the "candid" look and feel as well as ambient sounds but is kinda hit and miss for the dangling action itself, it's also very costly and has been axed by OpenAI so it won't evolve anymore. Grok Imagine is usually bad when trying to generate a video from a text prompt only however when provided with a starting image it gives surprising results! Here are some examples, some of them have very natural looking dangling action:

THESE are excellent! Your prompts are so well written. I'm copying them so I can tweak them for some IRL photos I have. The big thing I think with your success is the in depth description of the actual dangling. The AI just can't imagine it on its own with basic instructs.

Great job and thanks for sharing these!

[rubyred]

I’m having fun creating fake movie scenes. Here’s one with Halle Berry playing a politician. She’s in her office.

[Mr. Mike]

That's really cool. Could you do something with Jennifer Aniston

[rubyred]

That's really cool. Could you do something with Jennifer Aniston

Sounds like fun, yes!

[AntGouv55]

Opinions?