The Bell
My submission to the Third Edition of @runwayml#Gen48
Synopsis:
In dark medieval Europe, a mysterious Bell tolls across the land, and with each ring, disaster is unleashed before children vanish from their villages. A once-loyal royal guard embarks on a desperate journey to find his missing son, haunted by rumors of hooded figures who appear after the bell tolls.
Just gave away the first Early Accesses to @letz_ai V3 😌
If you receive one of them in the coming days, here are some things you should know.
First and foremost: V3 is a work in progress.
We're learning how to optimize it more every day.
The purpose of Early Access is to give experienced AI users and trusted creators a way to beta test it, provide feedback to us, and help us turn it into its best possible version for the general public.
V3 is an incredible tool, and it’s very different than V2:
Great V2 prompts may work perfectly well on V3, or they might not work at all. So the outcome of your first prompts might be completely unexpected and different when you get started.
This is mainly because V2's token knowledge was different from the new Flux-based V3. V3 interprets things differently than V2.
If you feel that something doesn’t work in V3, don’t give up after the first shot. Instead, embrace it and explore where the new road leads you.
And of course, if you want something very specific: add a Model.
Models are madness on V3. They completely change the game and they can turn a boring base prompt into a masterpiece.
To control your outcomes when using models, the Quality and Creativity Sliders have more effect in V3. Also keep in mind that you can set intensities by typing “:0.7” or another value from 0-1 behind the tagged model. All of this requires some learning by doing.
Finally, at the moment both Image Generation and Training take longer in V3 than they did in V2. @biitzeth , myself and other friends are experimenting a lot to improve speed on all levels. We’re also discussing with our GPU partner @gcore_official to add more GPUs.
Don’t consider V3 a patch. Consider it a DLC with completely new quests. “You’ll need to relearn how to prompt”.
We will keep you posted as things progress. If you want to request Early Access to V3, please DM me or reply below, and I’ll add you to the list! ✌️
CRONUS
AI Short Film
In a galaxy colonized by sentient machines, diverse tribes, and belief systems have emerged. Among them are the Luminates, who grow weary of humanity's perpetual conflicts. They uncover a human enclave constructing a weapon to eradicate synthetic life. The Luminates decide to intervene, fully aware that their actions will spark opposition from other factions and their human creators.
Img: @letz_ai
Vid: @LumaLabsAI
Lip Synch: @runwayml
Music: @udiomusic
Voice:@elevenlabs
The Luxembourgish AI, @letz_ai , just launched its Version 2 and is now officially available in all European countries as of yesterday! Lucky enough to get early access, I've already started experimenting and creating with it. 🌶️
the creating process :
https://t.co/zFCoSuVt6S
A week ago, OpenAI introduced a text-to-video model called Sora.
So, I was wondering how she would look in human form.
Meet my generated version of Sora.
generated in #MidjourneyAI postproduction #Photoshop#OpenAI#openaisora