|
my buddy on mid journey doing some stuff: "realistic artstation style cyborg ninjas showdown in a brutalist architecture environment with dramatic neon lighting and a cinematic style" from elsewhere: "hyperrealistic typhoon thunderstorm serpent rising" I'm over here trying to give credit to people, or something...like it matters with this stuff, LOL.
|
# ? Jul 3, 2022 02:04 |
|
|
# ? May 29, 2024 21:40 |
|
|
# ? Jul 3, 2022 02:38 |
|
I dunno why midjourney is so great at tormenting pikachus but goddam does it have Wayne Barlowe's style down pat https://imgur.com/a/xbwg3Rn
|
# ? Jul 3, 2022 04:33 |
|
On the waiting list for Dall-e 2 but since I'm a nobody in Western Europe I doubt I will get access soon. Is there anybody with more daily prompt uses than creativity who could run mine through the machine? "Formula one cars racing on Rainbow road, photo" "Formula one cars racing on Rainbow road, photo, shutter speed 1/8 seconds" "Formula one cars racing on Rainbow road, pixel graphics atari game" "Formula one cars racing on Rainbow road, grand prix poster"
|
# ? Jul 3, 2022 13:07 |
|
hydroceramics posted:I'm looking for some other terms and artists names to mess around with in dalle mini. I feel like over the last couple weeks it has gotten a lot worse at generating images in the concept art style for some reason. What other terms do you think I should use to generate that same sort of feel? I'd love for it to generate stuff like Simon Stålenhag, but that seems very unlikely given his specific mix of the contemporary and weird scifi.
|
# ? Jul 3, 2022 13:19 |
|
Deltasquid posted:On the waiting list for Dall-e 2 but since I'm a nobody in Western Europe I doubt I will get access soon. Is there anybody with more daily prompt uses than creativity who could run mine through the machine? Just a couple quick ones before I go to work. It's kinda tedious to download all of the individual pictures in a group then re-upload them to imgur. If you want some of them enlarged let me know. Formula one cars racing on Rainbow road, photo, shutter speed 1/8 seconds Formula one cars racing on Rainbow road, grand prix poster "photograph of a pool party with pigs as the guests in the mid 1960s style of slim aarons" sporklift fucked around with this message at 15:45 on Jul 3, 2022 |
# ? Jul 3, 2022 15:08 |
|
Gromit posted:I'd love for it to generate stuff like Simon Stålenhag, but that seems very unlikely given his specific mix of the contemporary and weird scifi. Thats the sort of feel I'm trying to generate - weird, conceptual sci fi like what you used to see in game manuals and book covers. I'm not an art person so I don't really have the language to quite get what I want. Way upthread somebody said the future is going to just be trying to figure out what random words to dump into an AI to manipulate the system and that's definitely what this feels like
|
# ? Jul 3, 2022 16:44 |
|
I want to see what 2 can do after a year of iterating off of mass user inputs and refinements to better teach it what people want to see and mean by certain phrases
|
# ? Jul 3, 2022 16:57 |
|
sporklift posted:Just a couple quick ones before I go to work. It's kinda tedious to download all of the individual pictures in a group then re-upload them to imgur. If you want some of them enlarged let me know. Cheers, thanks. I was expecting / hoping the AI would recognize "Rainbow Road" as being a place from Mario Kart but it took it very literally. The grand prix posters are pretty cool though. Top left one in particular; could you enlarge it? Two more requests /experiments if people don't mind: 1) An elegant woman with heavy makeup looking over the rim of a hand fan, portrait, oil painting 2) The skyline of ancient classical Rome, in the style of Hiroshi Nagai
|
# ? Jul 3, 2022 20:05 |
|
Gromit posted:I'd love for it to generate stuff like Simon Stålenhag, but that seems very unlikely given his specific mix of the contemporary and weird scifi. https://twitter.com/johnnygizmo/status/1543605979853512705
|
# ? Jul 3, 2022 20:23 |
|
Deltasquid posted:Cheers, thanks. I was expecting / hoping the AI would recognize "Rainbow Road" as being a place from Mario Kart but it took it very literally. I don't know why this went b&w. Here's one with "in color" added.
|
# ? Jul 3, 2022 20:51 |
|
Weird that it would do black and white. I wonder if it got a lot of old B&W textbook articles as training? https://imgur.com/a/l0bzYxY I saw these birds in cybernetic battle armor posted a while ago, apparently the prompt was [bird type] wearing futuristic cybernetic battle armour, dramatic lighting, portrait, realistic reflections It seems to work well enough, so maybe a couple other variations on that theme? Any of these appeal to you? Party Parrot wearing futuristic cybernetic battle armor, dramatic lighting, portrait, realistic reflections Fat plush frog wearing futuristic cybernetic battle armor, dramatic lighting, portrait, realistic reflections Frog-themed Hot Dog wearing futuristic cybernetic battle armor, dramatic lighting, portrait, realistic reflections Certified Forklift Operator wearing futuristic cybernetic battle armor, OSHA-compliant, dramatic lighting, full body, realistic reflections Also, let's see how well it can extrapolate. The forgotten Mario brother, Σario, Official Nintendo Reference Art
|
# ? Jul 3, 2022 21:11 |
|
wa27 posted:
Wonderful, thanks. It's a shame the AI doesn't seem to know who Hiroshi Nagai is, though. Odd that it went for black & white!
|
# ? Jul 3, 2022 21:26 |
|
Okay, I did a search for rome skyline picture on google image search, and there's a LOT of black and white vector art. That's gotta be the explanation
|
# ? Jul 3, 2022 22:41 |
|
This last one really speaks to me, I kind of want to build an online identity around it Chainclaw fucked around with this message at 23:09 on Jul 3, 2022 |
# ? Jul 3, 2022 22:54 |
|
Black August posted:I want to see what 2 can do after a year of iterating off of mass user inputs and refinements to better teach it what people want to see and mean by certain phrases Do you rate the pictures after they're presented to say if it matched what you were going for or not? Or is it sophisticated enough to just somehow understand based on if you tweet it out that it was probably correct?
|
# ? Jul 3, 2022 23:45 |
|
Deltasquid posted:Cheers, thanks. I was expecting / hoping the AI would recognize "Rainbow Road" as being a place from Mario Kart but it took it very literally. Haha. I was thinking you were going for LGBTQ or the rainbow road from Thor? You have to be pretty specific. A Strange Aeon posted:Do you rate the pictures after they're presented to say if it matched what you were going for or not? Or is it sophisticated enough to just somehow understand based on if you tweet it out that it was probably correct? You can choose to download the picture. You can also save pictures to a private gallery for yourself or publish a link to a picture. I'm kinda surprised there isn't more of a watermark or EXIF data. A funny vintage photograph of a cow riding a bicycle macro photograph of a wolverine action figure with his claws popped in the art style of bill sienkiewicz They have an instagram prompt today "Dinosaur Birthday" that you can either suggest a prompt for them to do or make your own. "A child's crayon drawing of a dinosaur with a birthday cake and jet airplanes in the background"
|
# ? Jul 4, 2022 00:08 |
|
A Strange Aeon posted:Do you rate the pictures after they're presented to say if it matched what you were going for or not? Or is it sophisticated enough to just somehow understand based on if you tweet it out that it was probably correct? you can on midjourney, you can rate other people's stuff too
|
# ? Jul 4, 2022 00:12 |
|
The Vermeerpocolypse
|
# ? Jul 4, 2022 02:52 |
|
LifeSunDeath fucked around with this message at 03:27 on Jul 4, 2022 |
# ? Jul 4, 2022 03:19 |
|
Playing with DALL-E's edit function. I took a watercolor painting I'd done of a raccoon's head, made the paper transparent, cropped off the bottom, and resized it to give DALL-E some space to work with. There's some neat potential here!
|
# ? Jul 4, 2022 07:58 |
|
photo of rappper Riff Raff riding a shark while drinking purple drink photo of a teenaged werewolf slam dunking a basketball Teenwolf reboot not looking so good.
|
# ? Jul 4, 2022 09:08 |
|
Last Visible Dog posted:Playing with DALL-E's edit function. I took a watercolor painting I'd done of a raccoon's head, made the paper transparent, cropped off the bottom, and resized it to give DALL-E some space to work with. now crop the top and have it recreate the head
|
# ? Jul 4, 2022 09:29 |
|
Dogs playing poker in the style of Ralph Steadman.
|
# ? Jul 4, 2022 09:31 |
|
|
# ? Jul 4, 2022 09:31 |
|
If you ask DALL E2 to generate prompts in the style of DALL E Mini does it kinda do a diss post?
|
# ? Jul 4, 2022 09:52 |
|
How do knees work
|
# ? Jul 4, 2022 10:03 |
|
Fintilgin posted:The Vermeerpocolypse All of these are absolutely off the chain! I would like to request some prompts of paintings in the style of Kay Sage to see what it spits out. The ability of this thing to resurrect long dead artists is bonkers!
|
# ? Jul 4, 2022 10:09 |
|
major xtro vibes
|
# ? Jul 4, 2022 10:10 |
|
Humbly requesting any Dalle-2 access holders to do variations of Francisco Goya's Saturn Devouring His Son but with Saturn eating funny things
|
# ? Jul 4, 2022 10:33 |
|
Scholtz posted:Humbly requesting any Dalle-2 access holders to do variations of Francisco Goya's Saturn Devouring His Son but with Saturn eating funny things Saturn devouring my homework Saturn devouring a bowl of ramen in ukiyo-e style
|
# ? Jul 4, 2022 10:49 |
|
Saturn devouring the sun
|
# ? Jul 4, 2022 11:39 |
|
That's a 90s video game ad
|
# ? Jul 4, 2022 12:29 |
|
Bro this is amazing, how the hell did it know that I also wanted the ocean to be codeine colored? AI is literally alive and throwed and trill.
|
# ? Jul 4, 2022 13:51 |
At some point it's effectively the AI determining the "Cool S" of all components of human language translated to image. Which feels a bit like some sort of weird backwards human dopamine training exchange.
|
|
# ? Jul 4, 2022 14:55 |
|
Last Visible Dog posted:Playing with DALL-E's edit function. I took a watercolor painting I'd done of a raccoon's head, made the paper transparent, cropped off the bottom, and resized it to give DALL-E some space to work with. This is incredible! Theoretically, you could also give a certain person a bunch of different outfits (or bodies) by just cropping their head into this and giving it a prompt, right? Also, most of the simple generators I've played around with are pretty good with horror type images, but that might just be because the things they produce are always a little disturbing. Would someone with DALL--E-2 or Midjourney mind trying the prompt "a press photo of an evil nightmare horror magician"? I'm curious as to how it'll compare to one of the replicate generators or dall-e mini's stuff. Here's replicate's attempt at it:
|
# ? Jul 4, 2022 21:06 |
|
Once in a blue moon Dall-E mini produces something that actually looks awesome. Someone please try this prompt in Dall-E-2: frank zappa sci-fi paperback cover colorful illustration
|
# ? Jul 4, 2022 21:14 |
|
Pitdragon posted:This is incredible! Theoretically, you could also give a certain person a bunch of different outfits (or bodies) by just cropping their head into this and giving it a prompt, right?
|
# ? Jul 4, 2022 21:15 |
|
axolotl farmer posted:Once in a blue moon Dall-E mini produces something that actually looks awesome.
|
# ? Jul 4, 2022 21:17 |
|
|
# ? May 29, 2024 21:40 |
|
huh, so it IS a completely different kind of result! very interesting! and these are great, thank you!
|
# ? Jul 4, 2022 21:20 |