|
Post by dans on Jan 16, 2021 15:54:16 GMT
Well, they really did it, wiped out Flash. And my two go to sites for creating character portraits no longer work (Hero Machine, X-Girl Dress Up).
Are there any portrait creation sites that do NOT use Flash that you guys know about? Thanks!
|
|
|
Post by DocQuantum on Oct 17, 2023 2:16:46 GMT
Have you found anything, Dan? I've got a ton of characters that I need to design for a certain project, but I haven't been able to find anything rivalling Hero Machine. I've tried some AI sites like Craiyon, but the results are such a crapshoot. Other sites that claim to be able to help you design a superhero tend to make them all look so generic.
|
|
|
Post by dans on Oct 17, 2023 10:55:05 GMT
sorry, Doc, I gave up. I now use some of the quick, free AI sites, and even though they really never produce exactly what I want, sometimes they get close enough. I usually open several tabs and run a prompt multiple times, perhaps making some tweaks to the prompt between iterations, and sometimes I get something that is close enough.
The last time I looked at Hero Machine, it was still possible to download it and run it locally if you have flash on your local computer. But that was at least a year ago.
|
|
|
Post by dans on Oct 17, 2023 12:59:06 GMT
here's some fast, free, but not high quality AI art sites: huggingface.co/spaces/stabilityai/stable-diffusionhuggingface.co/spaces/runwayml/stable-diffusion-v1-5huggingface.co/spaces/huggingface-projects/diffuse-the-restEach site generates 4 images based on your current prompt. The last one allows you to upload a starter image and it will make similar images based on your prompts. Or you can start without an image and then use an image it generates as a base. I usually use the first two a half dozen times and vary the prompts, and save the images I like, and then maybe use the third one to refine one of the images I like. Note - when using the third site, you have to click on double right arrow in the tool bar to see all 4 images (plus your original). You can click through to the best image and run it again if you like and evolve images that way. Not NEARLY as easy as Hero Machine was, but the variety is wider. Supposedly the more detailed you make your prompts, you will get images more like what you want. But with this generation of AI it takes a lot of lather, rinse, repeat. Someday I will get interested enough to examine the other AIs; it is likely that there are others that give you better prompt-based results.
|
|