Re:Stable Diffusion: Is Video Coming Soon?
by astrange on Thursday September 01, 2022 @11:23PM
(#32686221)
The result is bad though, for the same reason you can't generate video with it. Comic panels need to relate to each other; you can't simply make them out of random images. There aren't sufficient style controls to do that with current technology, even if Midjourney added in "textual inversion".
Last week PhilFTW explained "How To Create a Complete Graphic Novel in ONE Day" with Midjourney in a YouTube video [1]. He uses five tools:
- Midjourney (to generate images)
- InferKit (to generate the story text)
- Word (to rearrange the story text to fit into some narrative)
- Comic Life 3 for iPad (to place the images and text in comic book panels)
- Affinity Designer (to design the cover and export everything to print, Kindle, and Blurb)
[1] https://youtu.be/tjj6KsPSHZc