-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
DALL-E 3 is live through Bing Image Creator
- Thread starter EviLore
- Start date
Pegasus Actual
Gold Member
There's something about monkeys with donuts, and as an artist I need to continue exploring it.
Radioskugga
Member
Alx
Member
The bing page has been stuck processing my last prompt for days (« Freddy Krueger and Edward scissor hands looking extremely frustrated while playing Mikado » )
Is there a way to cancel a création in progress ? How am I supposed to ask for « Winnie the Pooh and Donald Duck wearing cotton briefs » now ?
Is there a way to cancel a création in progress ? How am I supposed to ask for « Winnie the Pooh and Donald Duck wearing cotton briefs » now ?
Radioskugga
Member
Alx
Member
Alx
Member
Yeah, or some random board games. It wouldn’t recognize “operation” as a game either, but Jenga works.
Radioskugga
Member
Zelda - Link To Hell
-Minsc-
Member
This was supposed to be She-Hulk doing the dishes but instead generated this image.
The Hulk does like to crash the party. His mere presence seemed to make her boobs grow.
Pegasus Actual
Gold Member
Cosmic Smash
Member
Radioskugga
Member
Cosmic Smash
Member
Midn1ght
Member
Very nice! Can you share the prompt you used to get that particular style?
intbal
Member
I wish I could.
These are old images. Long before I started saving my prompts.
However, I tried to recreate them just now with moderate success.
This prompt was:
"A manga drawing of a schoolgirl with white hair standing on a Tokyo street corner, highly detailed, very colorful, Frank Frazetta style"
I would not have used Frazetta in the original images. I was only forced to switch to deceased artists once Bing changed their policy of not allowing styles from any living artists.
I'm not sure who the originals would have been. Arthur Adams, perhaps, considering the intricate detail.
For the other two "non-street" pictures, I originally used "in line at a movie theater" or "standing in line at a movie theater", since that's something I was attempting back then to get full body shots (and usually failing).
-Minsc-
Member
This prompt was simply "Princess Zelda ploughing a field". All four images were quite similar in style. This is the one which had the most coherent "Franken-tool". While I never played Tears of the Kingdom, it ironically does seem in line with the spirit of assembling items in game.
Last edited:
Midn1ght
Member
Thanks a lot, tried a few but didn't come close the specific style you got so I used "in semi realistic manga style, highly detailed" and got these results. Very cool.I wish I could.
These are old images. Long before I started saving my prompts.
However, I tried to recreate them just now with moderate success.
This prompt was:
"A manga drawing of a schoolgirl with white hair standing on a Tokyo street corner, highly detailed, very colorful, Frank Frazetta style"
I would not have used Frazetta in the original images. I was only forced to switch to deceased artists once Bing changed their policy of not allowing styles from any living artists.
I'm not sure who the originals would have been. Arthur Adams, perhaps, considering the intricate detail.
For the other two "non-street" pictures, I originally used "in line at a movie theater" or "standing in line at a movie theater", since that's something I was attempting back then to get full body shots (and usually failing).
intbal
Member
Thanks a lot, tried a few but didn't come close the specific style you got so I used "in semi realistic manga style, highly detailed" and got these results. Very cool.
Looks great. I suspect the "realistic" probably helped out, and I think I might have used that term in my old prompts.
Croatoan
They/Them A-10 Warthog
So i have to confess, this is my favorite artstyle/setting and I have no idea why it hasn't been made into a western open world game. Post Apocalyptic Dieselpunk. BTW, the faces need to be redone with Stable Diffusion for a lot of these. I think Dalle 3 is running out of resolution or something in these very detailed scenes. The ones I like I take through SD to get better faces and details. Might show off a few in the next post.
Radioskugga
Member
Tried with Mario in the Blair Witch Project setting. Not quite what I was going for.
This is more like Mario on shrooms.
Last edited:
-Minsc-
Member
Here is an example of that last image run through various stages of stable diffusion to fix it. Still isn't finished but gives you an idea what you can do.
That is cool. AI enhancing of an existing image is something I was wanting a few weeks ago. Is SD free to use?
I typically stick to established characters from which I assume IC has no shortage of reference content to draw from. Faces usually only become an issue for individuals who are in the background.
I do wonder if it has issues with creating nonspecific large groups.
Here's one of the images I got from my second day making images.
"Toronto Maple Leafs win the Stanley Cup"