No worry guys, this is just a shallow test, I'm training my own lora, the realistic version of Fallen Swan is WIP. And even better news, since I'm a huge fan of 'Poor Sakura', I'm using 'Poor Sakura' legacy to train the AI, hopfully you should be able to see some cool AI version 'Poor Sakura' images very soon. Eventually, I'm going to use AI to revive 'Poor Sakura' character, combine with 'Fallen Swan' story. Since 7thDream is gone, and I'm not a Pro of modeling, I guess this is the only way we could see 'Poor Sakura' again. I know a lot of you not happy with AI images, but let's put down the bias, hug the new tech, figure out how to make it satisfy our needs, because eventually it going to replace every tool we're using right now, just matter of time. At least, with my limited skills on AI, it already speed up my work by 100 times and produce much better details than Daz3d. I know we still have a lot issues to fix, but that's something worth to expecting.
Good to hear from you boss~ look forward to whatever the next project brings. It will be good to see Poor Sakura's legacy continue~ combined with Fallen Swan I think we will all really enjoy it. Have you also considered using AI to enhance the DAZ3D images? It could really wonderful using it to enhance these images even if the render time is still an issue. I think using AI with conventional tools is the best way to utilize it.
Yeah, it is a pain to posing characters in daz3d. In the other hand, 3dsmax is much easier to posing the characters but slower in rendering. So I may use a workflow to use 3dsmax to render control net posed images, and let AI do the rendering. But still, I have to train a few LoRa myslef, right now AI does not understand a lot of the key concepts.