Link: https://imgur.com/a/fvofeFY

Yay or nay on the idea?

I feel it’s about half way done or so.

I’m not really a narrative writer, but it has been fun both trying to string the pictures together, and to get the generator to make specific images to fill gaps.

  • m_f@midwest.social
    2·
    9 months ago

    I like it! Have you tried making more traditional comic panels? Wondering since DALL-E 3 now can handle text generation in speech bubbles OK-ish

    • setsneedtofeed@lemmy.worldOP
      3·
      9 months ago

      No, I can’t get images with continuity. I will reuse a prompt many times and pick out the best images.

      I still do a decent amount of manual photo editing to clean up what I decide to use. Something like speech bubbles would be much better handled in photo editing than trying to get them done by AI.

      I think a lightbook type format is more plausible than attempting a full on comic.