Generative Art

GasBandit

Staff member
Ş̸̪͍͖̯͔͓͓̦͔͚́̈́͑̇̽̓͐̈̂͛̄̈̕̚͝ͅT̴̻̽Ǫ̷̨͉̹̘̝͂̎̇̌̊̆̈́̽̈́̃̅͝Ņ̴̢̧̤̜͍̺͇͚̫̮̫̳͔̪̊̋̄̄̌̑̕͝K̵̢̛̗̲͍̹̠͉̬̥͈̬̫̭̯̣̓̀̀̌͂̎̎̈́̌̉͘͝Ṣ̷̡̛͎̘̱̲̯̙̰̞͍̼̆̈́̏̐́̓͌͘͜
 
playing around with V5...so, I was trying to make cute girls fighting dragons. MidJourney AI is supposed to not make nudes or porn or the like. But I told it to use Boris Vallejo style because I like his fantasy work. Well, Boris Vallejo puts lots of titties in his art. So guess what I got?
1679409893270.png
 
I am continually and pleasantly surprised by the better hands generated by v5 Midjourney. they're not all perfect, but most of them are much better now. I made this one while showing a friend how I tweak keywords and weighting to get me exactly the image I'm looking for.

1680716934093.png
 
Boy, this guy is really all in on this Balenciaga meme
[Niklass] created the basic static images with Midjourney, evoking the Harry Potter actors and outfits through text prompts such as “male model, grotesque, balenciaga commercial.” Then he used ElevenLabs—a “voice-cloning” tool—to create models of the actors’ voices based on previously recorded audio. Finally, he fed the images into a service called D-ID, which is used to make “avatar videos”—subtly animated portraits, not so far off from those that appear in the newspapers of the Potter world.
--Patrick
 

--Patrick
OMG, I really thought it'd take more effort than that...took a selfie and used those exact keywords...
1681171589040.png


I didn't use the voice AI or anything, I just used D-ID's generated voice, which sucks, but still...


The whole thing took me less than 10 minutes.
 
Last edited:
Midjourney V5 does so much better with reference images than v4 did. Same selife, but the keywords were "steampunk with gun, male warrior"
1681176158566.png
1681176177014.png


And I didn't have to sprinkle a million little 'hinting' keywords into it to get it to make something halfway decent.
 
Top