@nickfloats Last week at work I needed an specific action and expression. MJ was really stubborn about it. I had to use the Dalle3 inital image and then using Img2Img on Leonardo. Final touches with Photoshop. That day, MJ let me down. But Im not Nick, so, what do I know.
@nickfloats Today I had a hard time simply trying 3 AIs to generate a Ground Level Shot of two feet wearing orange flip flops at the center of the image while surrounded by other feet wearing other shoes, and yesa at the end I had to take the damn "best" image to photoshop for a final touch.
It's Real! ๐ณ @Scenario_gg leading the way with LoRAs and Character Consistency! And I will keep studying this lovely platform, thanks a lot @emmanuel_2m@araminta_k ๐๐ป (And I only needed 8 images for training ๐ณ)
@emmanuel_2m@Scenario_gg Beautiful! Im trying to learn it and soon I'll finally share my experiments with @Scenario_gg when you learn new stuff it sometimes can be scary, thats how I feel at the momento but I know in the future I'll be using this tool a lot. Congratulations to the Devs, amazing work โค๏ธ
@amli_art Im also share this one only because Im proud of my self for keep trying to learn new stuff such as ComfyUI, that actually makes me feel so dumb every time
@thepaulmontreal@midjourney@Scenario_gg@runwayml@Artlist_io@elevenlabs Great work, its just something with the pace, the rhythm... from 00:39 it gets better. Maybe a few more cuts, shorter takes, and the sound effects matching the actions on time. But the most important: Character Consistency! ๐๐ป๐๐ป