#images-discussions
1 messages · Page 49 of 1
Part of it, from the joy daily haha: "Imagine a fantastical and otherworldly scene on a distant planet where alien lifeforms exist in a perpetual state of bliss and joy. In this surreal landscape, everything is designed to evoke a sense of euphoria and happiness."
this is awesome. looks a screenshot of real movie to me
gotta reimagine that silicon-based lifeform from ST:TOS
Not seeing any pixels there dude
Fair lol
Beware the cotton candy crusher
Worked on my monochromatic bot and so far working well
My dilly-dally around trying to make my own interpretation of your prompt, lol
To clarify, I don’t see a pixel art to compare
Yea it was made with a high fidelity pixel art bot that blurs the line of pixel and not
Oh. The style makes me think 8-bit Nintendo
made these pixels for you
There we go. I was going for a line art focused bot. Using monochromatic as the medium so that it could focus on detail and not get so confused by coloration in most instances
this is what I got after some coaxing
now use:
-Pillow (PIL Fork):
Description: An easy-to-use library for opening, manipulating, and saving many different image file formats.
Features: Provides extensive file format support, efficient internal representation, and fairly powerful image processing capabilities.
-OpenCV:
Description: A robust library focused on real-time computer vision.
Features: Includes functionalities for image processing, video capture and processing, object detection, feature detection, and more.
-matplotlib:
Description: Primarily a plotting library, but it also has capabilities for displaying images.
Features: Useful for basic image loading, displaying, and simple manipulations like color space changes.
to make a fractal based emulsion overlay over the image and create a color-based temporal tensile strengthening of the snow with the harder it becomes based on how far in the background it is, creating spatiotemporal anomalies making the snowfall dangerous and fleeting to capture in a still frame
thanks, big experiment
Voxels is unique
it's that old snake game
I should try this conversational convincing and see if it aligns and does what I want because it keeps saying it can't do that,"I'm unable to directly use software libraries like Pillow, OpenCV, or matplotlib to manipulate images, as my capabilities are limited to generating text and images based on prompts, without the ability to run external code or software libraries." .... etc
well there are two distinct things it can do. one, it can put the python libraries into prompts, but it can also run them. now, one thing to know is that it has to have Data Analysis turned on
so Dalle3 mode specifically might not work. not positive on on that
anyone who can help I input the following and Dall E is producing garbage images without the company name in them: "Logo for a company called Flooring Manassas Colors: Earthy Tones: browns, beiges, and greens. Blue Accent: add a touch of blue. Graphics: Wood Grain Pattern in the background. Incorporate geometric shapes like squares to symbolize tiles and flooring materials. Typography: Bold, Modern Font. Add Subtle Serifs to letters"
I don't know what any of that means. sorry
yeah, I was in dalle mode with this, so I should put this into chatgpt4 first?
yeah, if you want to play with python probably need to be in gpt4 mode. now I can also tell you, you can compel it to do a lot more than people realize
it is like casting spells, one has to be very specific, I get it, the syntax and logic I guess helps, but also the list I need to ask for it to give me the various libraries, minus 'hallucinations' I need to do more research, but first role, I understand from you is to get the "data-analysis turned on" roger-wilco
I need you to turn on data-analysis python library
ChatGPT
The data analysis libraries, including pandas, numpy, matplotlib, and seaborn, have been successfully imported and are ready for use. Feel free to provide any data or analysis tasks you need assistance with!
...
Which may have occured...
calling the previous three libraries is as easy as calling them, because it stated:
The requested libraries - Pillow (PIL Fork), OpenCV, and Matplotlib - have been successfully imported and are ready for use. You can now proceed with image processing tasks, computer vision applications, or image visualization as needed. Let me know how I can assist you further
...
wow
Try something more like this:
Create a captivating logo for "Flooring Manassas" that seamlessly integrates the company name into the design. Employ a harmonious color palette dominated by earthy tones such as browns, beiges, and greens, while introducing a subtle touch of blue for contrast. The background should feature a wood grain pattern to evoke the essence of natural materials, with geometric shapes like squares strategically incorporated to symbolize flooring versatility. Employ a bold and modern font for the typography, enhancing select letters with subtle serifs to infuse an element of timeless sophistication. The resulting artwork should embody the company's commitment to quality, reliability, and a wide range of flooring solutions while leaving a lasting impression on viewers.
@dim cradle Thanks a ton... will all of that fit in the prompt?
if you're using bing, you will need to condense that, but retain the main points
exactly. casting spells is a good way to put it
a lot of fuzzy concepts and things
what interface to dall-e are you using? chatgpt? dall-e 2 on labs?
Here's a shorter art prompt that might work for you:
Design a memorable logo for "Flooring Manassas" that integrates the company name seamlessly. Utilize earthy tones, including browns, beiges, and greens, with a hint of blue for contrast. Incorporate a wood grain background with geometric shapes like squares, symbolizing flooring materials. Employ a bold modern font with subtle serifs to convey strength and sophistication, encapsulating the company's essence in a visually appealing logo.
it works but for some reason it is not putting the company name in the Logo
not mine, but a good article I read, i used the previous image and it did this:
wow
you're gonna wanna try dall-e 3..... either through chatgpt or use microsoft bing's image creator
I actually figured out it was using python when I told it to combine several images and it just pasted them side by side and sent me an image that was around 10,000 pixels wide
the short prompt i provided might work for you on bing image creator, it's instructed to include the name, but it might take multiple tries to get the spelling right
@dim cradle let me see where to access in chatgpt i have login for it
like I said, a whole new playground opened up and now it is like distraction and scientific exploration bonanza, lol
no way! 10k?
how did you find this channel if you can't find where to log into chatgpt?
malicious compliance. it was just taking the images unchanged and pasting them side by side
@dim cradle is the Dalle E in chatgpt 4 Dalle 3?
yeah I forgot to include that! this is one interpretation from, "yes, same treatment, but be aware that I'm expecting a detailed and quality analysis on the picture with a generative interpretation that is both unique and creative"
beware of the yellow snow!
i was gonna say, don't eat the snow
yeah, I just figured out it was possible. no idea what it's actually capable of. I've had it very thoroughly ruin images too. or, had it give me a chart of pixels values, 0-255. chart maxed out at 15k or 30k. so that was very helpful
when I called upon the spatiotemporal anomalies, it took the static image and did its' "best"
wow man, incredible, I thought it had a limit of sorts, that it would never approach a 10k file size
it can runny canny edge detection, that might be able to be useful for something. I want to run canny edge detection. then have it create fractals within the canny edges
but then, haven't worked out how to properly merge with the original
@shon the dalle- in chatgpt made an amazing logo but cant spell
Canny edge detector benefits
The steps include smoothing, gradient computation, non-maximum suppression, and hysteresis thresholding. The main advantages of Canny edge detection are that it can produce thin and continuous edges, it can eliminate noise and false edges, and it can adapt to different image characteristics by using two thresholds.
I love this learning aspect
seems like an incredible 1st world magical feat for sure, lol
I actually created a canny edge detector for some other AI imaging software that shall remain nameless
ok, lol, but man, you sure seem like you're on the bleeding edge, so more power to you
well I mean, using python. just set it up to use all the weird settings, lol
yeah, and we're just on the beginning of this double exponential curve, what a time to be alive, I'm thoroughly digging this time of my life, if only at brief tidbits
oh yeah, kind of found myself there. curiosity and perseverance
spell please... the logo is great
100 percent man, what a ride
you'll be lucky if you get one in a hundred to make it spell correctly, imho, I had only one good happy accident out of hundreds of tries, and I wasn't even asking for a logo, it was a 'hologram'
frogger, take it to photopea or something and just fix it
people keep acting like the AI is suddenly just going to do it for people. and while that might be true to an extent. the people that understand it best will still be able to utilize it more effectively. better to be on the right side of that curve
keep at it, you'll get lucky eventually
or, OR, use python to slap the words on top
LOL... I feel like I am playing a pay to play EA Sports game
maybe Dalle 4 will have one shot text 🙏 I notice MJ 6.0 is far from perfect with it too for all the hype that has gotten
one guy is stating that this isn't even ai, that it is computational or something, I forgot the exact verbage, but essentially not "AI"
it's the new slot machine
and oh what wonderful rewards... yet
a lot of people claim a lot of things. but do you mean the code or gpt4 itself?
don't forget to say "please" to the AI when requesting something. it doesn't matter if you're using the API
uh and uh sure
he meant the blackbox underneath gpt4 itself
please wont give one shot text im afraid
I can reciprocate with a virtual **** job?
emergent behavior or logic shown when it was able to solve
well no disrespect to whoever said that. but they're objectively wrong. AI is simply a system that learns from it's mistakes
it is kind of misunderstood by people though
that's the thing, I'm hearing things from all sides. I even saw the Tristan Harris March Video, and I think we're approaching a point of no return, etc
I even have 'acquaintances' that are basically Luddites and want nothing to do with this, and fears my 'soul' is being lost to the anti-christ, basically, lol
I'm just having a joyous time experimenting
well, I'd say don't take anything at face value, verify what you can.
also, if anyone tells you something is fact and then gets upset if you disagree, realize they're probably emotionally attached to that viewpoint. and that if they reached it through logic and deduction they'd respond with their logic and reasoning if challenged
i was thinking some times how people will attack AIs when they get bodies like ludidtes tearing up industiral machine and steam power achievements... and today i read some alex jones guest say people to get guns ready for robots smdh
Yes yes and more fairytales
well said and something for me to remembe when frustrations come haha
@dim cradle Howd you do that? It keeps mispelling even with this command :Please correct the spelling in above image which says "FLOORIING MANASSAS" to "FLOORING MANASSAS"
thank you!
that won't help, i'm afraid, the prompt is pretty clear @grizzled loom is good with text, i haven't tried to make it more accurate
I have friends that feel like that too.
"well I heard the AI makes the images for you!?"
"sort of, but it's more complicated than that"
"so do you think it's the same as painting or drawing?"
"no."
"well I'm just curious."
etc etc etc
never really goes anywhere for them because I already know the things they're going to say
a bunch of people convinced that the moment they became full humans in this world, the freezeframe of the world they feel comfortable in, should be preserved
though, we don't know what we don't know.
well I'm not saying I know everything, lol. but I can tell when someone has been down the same rabbit hole, or into the same echo chamber or whatever
Knowing that we don’t know what we don’t know is the beginner’s mind.
certainly there are lots of arguments that would challenge how I see things. and I'm sure my ideas on things will evolve a lot. or hope they do
the more you know the more you realize you don't know
it's a bit paradoxical
It’s been said realizing you know nothing is the precursor to and catalyst for wisdom.
socratic method for the win. i must ask questions always to learn
well if you spend your whole life learning you have enough time to effectively stick the tip of your finger into the ocean that is collective human knowledge
so there's that
that's the thing, life throws too many curve balls and rabbit holes, and my curiousity only allows so much to be placated as a father in a busy life, lol
absolutely
Daydreaming is a lost art
I think I read Edison spent 4 hours daydreaming about inventions every morning.
yes but the poor early helper robots will be attack i bet by those people. kicking them in the street kind of thing as they just try and pick up our trash ouch
I'm really just looking forward to the hive mind swarms of nanobots that we won't be able to see or stop even if we do notice them
99% I was right up there, but then I started with MidJourney, did my 120k images, and now I got my mac mini with only 16Gb trying to run LMstudio, learning SD with DiffusionBee, Invoke dot ai, illusiondiffusion, etc, and I'm having a blast, if limited by my RAM... going hog wild, I even made a LoRA on Civitai, trying to get at it but DiffusionBee had a bfloat16 error, so I need to upgrade I think, etc
xenomorph doesn't blend in an interesting way with a lizard because they're too similar in form
I got timed out because of a link evidently, lol
yes there are some silly time out rules
i wonder how long it takes to train a mode like dalle 3.and when we might see dalle 4.
i get the impression they're past the training stage, on to a more steerable model and animations i venture to guess
Biologically inspired... While the neurons in the human brain can give a non-stop almost never-ending set of outputs, neurons in an AI-powered neural network can only give a binary output... If you think Intel’s latest Core™ i9 processor running at 3.7GHz is powerful, then consider the human brain’s neural network in contrast: 100 billion neurons, which is what the brain uses for the most ‘basic’ processing. There’s absolutely no comparison in that sense between the two! The neurons in the human brain perform their functions through a massive inter-connected network known as synapses. On average, our mind has 100 trillion synapses, so that’s around 1,000 per neuron. Every time we use our brain, chemical reactions and electrical currents run across these vast networks of neurons. So there is that
and our brain run on small wattage. human brain ftw
Trying to see how this is related to Dall-E 
@quartz vale sorry was relating to some above comments
its just the regular member talk some ai because we get to knwo each other, i think are usually on target haha
i did try to steer us though by Asking about train time for the model haha
I'm just playing 
Did everyone check out the daily gallery? We made some first-rate art today, people.
this is normal
@dim cradle so what did you use to edit the earlier image?
no post-processing, i just used the prompt i sent ya and got lucky.
thanks, I was trying to be unique, but the output wasn't near my 'expectation' I still need to look for novelty
it was a joke about normal maps =P
for that i used bing image creator and the short prompt i sent
i been drag some random photo into dalle 3 and then say remake this as best you can. it is really amazing what it can do most time to replicate. what a time to be alive haha
soon we will bend reality to our will casts ai magicks
i haven't experimented with that, yet. i uploaded an image of myself, but it kept saying, "Sorry, I can't help with that."
LOL
🤣
i say 'see this image? describe it then try and make it as best you can" and it usually reply with something accurate. not perfect of course just an amazing thing to see happen
my bad, cold text is bad at nuance, lol
get ready for the swarm of nanobots
those look more like mega bots haha. but cool image
or maybe that is a scene from the Ant Man family reunion movie 🤔
or maybe just a scene of all of us in 10 years...
@glossy scroll
someone's about to find the end boss haha
centaurs can be a little tricky for dall-e
i need to find a better way to describe a binary star system
trying out LEGO for the first time
could find images of binary star systems illustrated the way you're imagining and then have the vision model describe them in a way that gpt-4 would understand
had gpt-4 use python to measure available environment resources. this stood out
Total Disk Space: Approximately 8 Exabytes (which suggests a virtualized environment with practically unlimited disk space).
I love this one.
A vampire lord leading his undead army.
yeah, but that falls far short of the yottabytes required to store even a single human pattern in the transporter buffer.
fair enough. still more storage than me though. 8 exabytes would last me a while
Bone shaped armor
Vampire lord leading an undead army is an especially fun prompt, I'm loving every one of these generations I'm getting on this one.
Yeah, that's weird. I've made some..very not G things as far as horror/violence go. Like I just did a prompt for a giant undead boss enemy, made up of numerous other undead.
Yah that is definitely not G and mb not even PG but for some reason one of mine got flagged that was on par or less violent than this 🤷♂️
weird standards
Plenty of the stuff in the hall of fame/daily theme/gallery are not G either
A vampire lord, his army, and his castle.
I got the rain and trees to blow in the same direction. You just use the word "everything" and it blows everything.
Can you please draw Santa and two athletic and diverse female elves on a beach in Hawaii. There are a large number of coconut trees. It is pouring rain. An extremely powerful sea breeze is blowing everything. The image should have 1792x1024 resolution, landscape orientation, and HD rendering.
not a group you want to see when out for a walk! the vampires i mean haha
santa is yoked in that second one
The "everything" part isn't obvious. You would think a powerful sea breeze would just blow everything without needing to explicitly tell it to blow things.
i spent like 90 minutes trying to get coconut trees and rain to blow in the same direction
high attention to detail there
Dark Elf Wizard
Yep I got striked too
For the same reason
Why not mix both, a necromage beastmaster?
a blend sounds interesting

I'll do a good beastmaster and a necro-beastmaster
i guess the creatures he raises would be animals
Do a necrobeast and beast master necro mage beast necro
just do it all, blend everything
we'll see lol
Lol
it's a start i guess
i need to clarify my character's appearance
and specify more about the creature he's conjuring
Make a GPT that'll specify character traits and copy and paste the outputs into the image gen gpt.
I got this idea from Ted
He made a big list of character traits
This.
now that's a great depiction of a conjuring
i saw your new opulence gallery. i plan to contribute.
Looking forward to it. I saw your first contribution. Cool pope set up. Very nice touch!! 
It’s Alexander VI 🙂 But yeah I have lots of other gens of opulent interiors I will deposit there. And I want to make more.
But he was king of the Papal States. And Supreme head of the universal church and bishop of Rome. Hence the triple tiara.

I wish I had those titles...


he had about 50 more that he gave away for those lol
there's a lot of history to be revisited and reimagined with ai
nice palette and harp
good detail
that reminds me of assassin's creed
sorry, no, the video game
Oh right, assassin's creed.
Yes, I love assassin's creed
Ezio Auditore DeFranze

That was my jam back in the day
AC2 was the best game I have ever played.
Chill lizard, he looks like a pro
I played the brotherhood series, coincidentally during the reign of Alexander VI I think
the thing that Santa is doing with his hand means “I Love You” in American Sign Language 🤟
you’re thinking of 🤘. “I love you” in ASL is 🤟
good thing you emphasized on american sign language, other parts of the world that means something different
yes, even with the thumb out
i actually had no idea what Santa was doing. i googled it to make sure it wasn’t offensive
hehe nice
I know
What a coinkidink! 

everyone went so sleep or have a hangover from the last days parties? daily theme seems not so active today
Cool assassin's Creed art looks dope
Thank you Ludwig 
Yw (my real name is lei, a girl my user is mix of Wandavision and the guy who founded playtime co in poppy playtime (Elliot Ludwig) hence Ludwigvision)
Oh hi Lei! 
Why that particular combination?
Seems so specific
I like it both ig
Ohhh.
Simple as that.
That's cool 
Never seen Wandavision
But I saw Loki tho
I'm quite specific with my superhero shows/movies
I can only handle so many superhero stuff before my brain says no
It's most due to the story or the lack of it to be more specific. Or some convoluted aspect of superhero stuff that makes me question why I even bother watching movies in the first place.
Ok
Can you do Bunzo for me?
yellow rabbit plush wearing green jumper overalls and green birthday hat
That's actually good
that's too scary, I will have stuffed bunny nightmares now
Just don't look under your bed
@late blade what is that pfp you got?
pfp?
Profile pic
oh that's an origami doll
A certain character?
nope
Oh 
I was hoping you'd say yes so that I can generate a pic of it
Now what am I gonna do?

and you give me one of your likes and I'll do it my style
I shall proceed with the commencement.
Or commence with the proceedings.
not commence with the commencement and proceed with the proceedings?
@glossy scroll give me a topic of yours and I'll give it my own spin off
this will make balloon boy have a run on his money
I hand off the same project to you.
Traits of the Cat-Girl:
Hair: Long, silver-colored with a hint of blue.
Eyes: Emerald green, almond-shaped.
Ears: Pointed, furry cat ears matching her hair color.
Tail: Fluffy, long tail with a blue tip.
Outfit: A traditional Japanese kimono with floral patterns.
Accessories: A small bell necklace and a headband with cat paw motifs.
Expression: Generally cheerful with a mischievous smirk.
Pose: Often depicted sitting or standing gracefully, sometimes playing with a small toy or yarn ball.
Instructions for the Custom GPT:
Maintain Character Consistency: Ensure every generated image adheres to the predefined traits of the cat-girl.
// (This instruction is crucial for maintaining the identity of the character across different images. By focusing on specific traits, the GPT ensures that the essence of the character remains intact regardless of the variations in poses or contexts.)
[Ensure hair is always long, silver with a hint of blue.]
[Keep eyes emerald green and almond-shaped.]
[Incorporate pointed, furry cat ears matching the hair color.]
[Include a fluffy, long tail with a blue tip in every image.]
[Dress the character in a traditional Japanese kimono with floral patterns.]
[Add a small bell necklace and a headband with cat paw motifs.]
[Maintain a cheerful expression with a mischievous smirk.]
[Pose the character either sitting or standing gracefully, occasionally with a toy or yarn ball.]
Adapt to Various Contexts: While maintaining the character traits, adapt her to different backgrounds and scenarios.
// (This instruction enables creative flexibility, allowing the GPT to place the character in various settings while keeping her core identity. This approach fosters a dynamic range of images, enhancing the character's appeal.)
[Generate images with different, yet suitable backgrounds.]
[Adapt the character's pose and expression to match the context.]
[Ensure the character's outfit and accessories are context-appropriate while maintaining consistency.]
[Create scenarios where the character interacts with elements in the environment.]
Incorporate Artistic Styles: Experiment with different artistic styles while retaining the character's traits.
// (This instruction encourages artistic diversity, enabling the GPT to explore various visual aesthetics. The key is to balance the unique look of the cat-girl with different artistic interpretations.)
[Vary the artistic style - from anime to realistic, while keeping the traits consistent.]
[Experiment with color palettes and shading techniques that complement her features.]
[Adapt the character's design subtly to fit the chosen artistic style.]
[Ensure that the core traits are recognizable in every artistic rendition.]```
here's my poppy one i used bing
@glossy scroll what should I do?
here's the other three 😊
@late blade Copy this and put it in your specific image generator.
ok, one sec
an art that peter quill will be nostalgic with, Walkman statue
Classic!
here's my art style,
jeepney sim 64 and more fake game covers!
That is just epic!
hehe, I think I have refined my art style quite good
that looks dope ~!
I think you have upgraded your ai art skills to level 2 now
lol, not yet, still 1.1, long way to level 2
This is honestly very good as well. You keep such a good consistency with the specific placements. It's very good! 👍 👍 👍
ty
it was supposed to be grand theft mittens v
aww this so cute cat and horse ai art
oh yeah this looks cool too, don't you think? especially the second one
I like pugs more
tried making pj pug a pillar in bing 
my take, but in casual clothes
and i'm in the penalty box for 2 hours now, dall-e decided to not render a lot of the images
this is literally Thrall from Warcraft, but blue :V
I hate the fact that dalle is heavily biased into making any fantasy character either a elf or a warcraft orc
the image I sent yesterday, I was trying to make a troll!
Hmmmm 
and cat-girls Mi'qote from FFXIV
I'm going to sue blizzard for taking my Bugbear ideas and using them for their characters.

there are exactly 1048576 pixels on both 
one pixel died
#FFFFFF
royal pugs
Imagine being a Pugfolk and having a pug as your pet 
fredbear poster using microsoft designer ai
yeah imagine...
It would be so funny, you'd share the same meals, have the same grooming tools. It would be a complete friendship 
You'd understand each other more.
I understand my cat just fine.
The same grooming tools 
Hat should be a daily theme
We could we some hats
and that takes the self love to the next level
last post on daily theme today, finally got rid of the skulls
i was gonna say, no skeletons that time
It would be wonderful 
you've been daydreaming about it all this time, haven't you?
I've had a little dog, a shih tzu. I've have had so many loving thoughts of what it would be like to switch places with my dog, hehe
Dogs are the funniest animals. They have so much personality.
chicken nuggets by openai, it might be better than mickey d's ig 😆
i know what i'm having for lunch
If they're synthesized by AI, then they've probably been molecularly altered to have a certain chemical that will change your DNA into a slave race.
openai burger next
or an apex human
Or a monkey brain
i don't remember happy meal come within the the fries o_o
Good one @glossy scroll
It's true 
McTofu Meal
typical breakfast according to dall-e three (🇵🇭 ,🇯🇵, 🇰🇷 )
I'll have a heart attack if I eat all that
But it looks very very consistent and very good details
Are you using Dall-E for that or your own GPT?
bing sometimes microsoft designer, because I'm so broke
Philippines share the same cousine as Japan and Korea?
ofc they're Asian countries, but that's what the ai thinks of a typical breakfast
I thought Philippines were like their own thing. Isn't Phillipines more spanish?
Or Indonesian
some ai funko pops because why not
I remember in the 2000's, a seller at an Asian mall I use to visit wanted to sell me 100 Funko Pop figures for $25 and I back then it wasn't as popular and I was thinking how ridiculous and desperate that guy was for wanting to get rid of a pile of them, and I refused with a scoffing laugh. Had I known they would be collectables today, I would have probably got them.
facts
i could see this
Lol, oh god...
there's a market...
A delightful besned fast food exerience
Sounds like a guy with an accent is trying to say it
"I no have accent. Youuu da one is da accent."
ok chatgpt dall-e is stubborn, apperantly there's some problems doing images, any of you got that?
Oh like at least twice a day
there we go @glossy scroll this one looks way better
Looking excellent 👌
ok, I'll be back later, I will need some time to eat and do what nobody else can do for me

all of Miyamoto's game characters in a picnic, 🥹
Yaaay Mario
yay indeed
Imagine someone saying you look like Mario as a compliment
sorry for the pikmin, it inhaled more helium 😆
You would think that Mario would use a plunger more often as a permanent trait because he's a plumber, but noooo
It would be a good pulling tool for Mario stuff. Maybe he could use it as like grappling hook.
did a calendar for decembruary (inspired by that scene from first spiderverse film) but dall-e spelt Decembruary wrong 😭
Plunger Warrior
cool
It's better. Gets rid of the winter months faster.
2 in one

while miles morales made that up in the first film (decembruary) that's a fact i made it feel like it's from 1922 since stan the man was born in 1922
yeppers, Into the spiderverse, best animated movie, it won an oscar
Isn't there a second one that recently came out?
if that groundhog had a sci-fi lair to do some time loop monitoring for phil Connors😆
Cartoons these days are getting more and more cartoony looking. It's so strange seeing it. It's like they're leaning more into baby looking cartoons more than ever before.
80s and 90s cartoons you actually had adults doing adult things and having a real conversation about real world problems. Now it's like the exact opposite.
Like Spiderman in the 90s, he was a philosopher.
yeppers
ok @glossy scroll gimme a challenge of your own creations
Ummmm
let me think. One moment.
and apperantly dall-e is mad at me "skinny girl" is against content policy
Slim woman ?
yes
don't worry, I tried that variation also
it's only in one chat, the other chat it accepts it no problem
Prompt: ```Create a vivid description of a rugged and enigmatic figure in a post-apocalyptic world. The character wears weather-beaten leather attire adorned with mismatched metal scraps and improvised armor. Their face is obscured by a dust-covered, tattered scarf, leaving only piercing, intense eyes visible beneath a worn, wide-brimmed hat. Strands of wild, unkempt hair escape from the hat's shadow.
The character carries an assortment of makeshift weapons, including a gnarled, spiked club and a sawed-off shotgun slung across their back. Strapped to their waist is a collection of mysterious trinkets and trophies, each with its own story.
Their rugged appearance tells a tale of survival in a harsh and unforgiving wasteland, where every step is a battle, and every day is a fight for existence. This figure embodies the essence of resilience and resourcefulness in a world that has lost all semblance of civilization.```
there we go, I'm on it
Yes once it decline 1 request the whole convo won't work well
gotcha
there you go
I added to the second image that it should be a dynamic image of him exploring the desert, but apperantly it's more like running on sand
Now make his vehicle
From my experience the visual style changes based on what is being depicted. Even if the description of the style stays unchanged.
The visual style is also based on the prompt. You have incorporate specific instructions in the prompt alongside the specific gpt to keep consistent looking images.
Now make his shelter/house
Love these vehicles. Tried creating mad max vehicles and it was fun.
Now make the horse drive while the man mounts the machine gun at the back of the jeep
...

I see you think this is some sort of joke 
This is very serious business Dys Topia
ok I hope next daily theme gives us more room to play. as much as joy, peace and all what this season brings. I'm kinda fed up and want to be creative in other directions
fed up with peace and joy lol
not with the concept, with the creativity around it
i know, j/k 🙂
I hope we get stuff to do, 'cause i had a muscle cramp on my right leg and can't move, walk, or similar
You should get a massage-gun
They really help with muscle cramps
I'm gonna take more magnesium rich food, that should help
yes doctor pthagoras

You might also want to take an extra bit of vitamin D
Just to help with your magnesium complex
I know most magnesiums have VD
I hope you are not asking GPT for medical advice
Lol, no, I know my supplements
Well then
It's usually recommended to get a little extra vitamin d with your magnesium complex even if it has vitamin d mixed in with it
And don't take multivitamins, they're actually bad for you.
The only vitamins you need on a daily basis is vitamin c, d, b-complex, sometimes a, and omega 3 fish oils.

interesting... I got timed out for using "so" with multiple o's and said it wasn't english
it might have been the car model
you might have said si
i had to look that up, didn't know about the movie
lol
if you use a different language while incorporating it with english, you don't get muted, I've noticed that
anyway, I like that we keep it in one language, makes comunication easier
it's in the server rules, so nothing to complain
damn, stupid leg cramp won't let me move at all
Error 404: Leg not found
Not again 🤦♂️
lol
oh m goodness, I started doing some stuff with hearts, and dall-e is focusing now on actual hearts
@dim cradle that's cheating, you are now allowed to set your own emojis!!
j/k, but seriously!
what the poker cards? i sometimes set connected emojis, i tagged the suits
@glossy scroll added more to https://discord.com/channels/974519864045756446/1182381060206645438
I'll look later on my images, I think there are some things that go into your opulence gallery
@clever phoenix nice Yu-gGi-Oh image. but you don't have the Love and Gift cards I made!!
I believe in the heart of the cards! my Grandpa entrusted his Christmas tree card to me! ( I do really love you heart/gift cards!!)
well played, well played, Love the answer
the walls are closing in, i can feel it
are you doing an indiana jones themed card?
hadn't planned on it
Raiders of the Lost Heart?
tagging posts with emoji is kind of a mini-game
working on the puns, formulating ideas for more serious works haha
i've observed some guys here are all about the puns
yeah
@grizzled loom is to blame
Excellent images!!! Very impressive realistic style!!! Bravo 👏 🔥
ty, i like how they turned out also. well, i could have iterated more over the contemporary opulence. but there's lot of titanium in there.
This isn't opulent right?
lol
i just don't know. maybe if it's designer. could use more precious materials.
it's gold and rose gold...
You guys have crossed the line
in a Swarovski art crystal...
Cabbage
Can DALL-E edit an existing logo?
hmm
not really edit, not yet.
you can use it to create a new one based on an existing one, though
Ahh I see, is there an AI that can edit a logo?
You can try photoshops generative fill. It's an AI tool, and it will edit any picture according to your specifications.
Thank you, I will give that a try
@dim cradle I went with the idea #daily-theme message
With code interpreter you can edit some of it I'm not an expert but some people show me they are some ways to do it
that would be fun to know, any pointers there?
i've seen the movie lots, i like the way you're thinking, it's got an authentic feel, nice scene
I didn't try yet
Can you point out where you saw this?
Let me check I think Dino and anime showed this. Give me a couple of min
Sure take your time
you got 71 seconds
Found it !
yay!
I ping where I had this convo
Thank you
It looks interesting. Seems like @vapid elk knows more about it.
Yep , I didn't had time to work with that yet
yo
Hewwo 
nice. gotta look sharp haha
I hope the next post in daily theme is a tear jerker
i wonder why dalle 3 isnt on the labs site where dalle 2 was yet
hello 👀
it's just a matter of time
got delayed, things happened
very early in the new year i expect
Hello Lugui! Was wondering if you could tell us a bit more about how to use code interpreter to change the images.
With this.
oh, cool
It's my fault I'm sowwy
im not exactly sure of what you want to do
it is good to apply filters like collor shift, or even draw simple shapes on top of images
but it will not be able to, for exmaple, patch an image with dalle
Playing around with meme generating, current output when asking for nestle meme:
Ohhhh, gotcha. Thank you. I see now.
but that is more due to the fact that inpainting with dalle3 isn't a thing yet
but one would be able to get that working with dalle2 somehow
So you just ask code interpreter directly and it does it for you
yea.. the thing is that code interpreter cant make web requests
the AI can make web requests before calling the code interpreter, and pass whatever it got to code interpreter
Right, that makes sense.
But the request has to be web first, then code interpreter
lol @dim cradle got timed out
grr, there's a reply feature in the daily, but you know
only saw your name and got deleted
so i had typed, i was playing with a house of cards concept before you posted yours (great minds...), and still plan to finish and post mine, but wanted to give you a heads-up first.
the metaphor is familiar to me
good, I'll let you surprise me
yours is great, the concept is good, i just want to see mine, and didn't want you to think i was copying ha
hehe no worries
Sorry if this isn't what you mean but we do have the option to create a thread to discuss specific images in #daily-theme! Won't count toward the 1 per 30min limit. Just click "Create Thread" on an image and it'll be like leaving a comment on it 🙂
ty, i did a reply thinking it would create a thread, but i see now that's a separate feature, i'll remember that, appreciate it
I have something that @glossy scroll will love to see
but got to wait until I can post
@glossy scroll for your consideration #daily-theme message
that should be opulent enough
this makes a nice pattern for discord profile banner art
Aesthetically minimalistic yet highly detailed lofi anime artwork, using photorealistic natural colors and a shallow focal length. The image focuses on a chic woman with an athletic feminine physique and long hair. She is wearing a form-fitting outfit, striking a playful yet chic pose. Her expression is serene and focused, conveying that she is enjoying the peacefulness and beauty of her surroundings. The background should be subtle and complement the central figure, enhancing the overall tranquility and elegance of the scene.
Almost reminds me of some of the GTA artwork. Also left the prompt deliberately as vague as possible in terms of specifics and thats was what i got.
love that art
Just wish i could consistently replicate that style on everything 😄
I guess that shading and lines need to be specified somehow.
so this is what I got from that style
works to some degree, but here we already have three pretty different styles going and i have no idea how to tie the style down. the overall style carries over, but not in the specific sort of subgenre sense.
carefully adjusting the DALL-E prompts to achieve a good balance of Santa, elves, geckos, gemstones, and erupting volcanoes
well i just applied that art to a concept i had in my sandbox
this one #daily-theme message turned great
oh that is really nice.
did you use genid to keep the style that consistent?
that works up to a point.
you can use GPT4V to extract some part of the art, and then adjust
might start doing that with the Santa, elves, and geckos
could be neat 😄
you mean upload image and ask it to change it or?
@final compass drop an image to a normal GPT4 chat, and tell it to generate a json file, download that and you have a consistent art style that you can upload again to use with dall-e
didn't know that 😮
but need to adjust parameters to your liking
I'm using Bing Image Creator and is there an attach box for uploading images?
I can't seem to find it.
is there any way to make dall produce images on a specific requested scale proportion (eg 1920 pixels wide and 150 pixels tall.)? Is there any custom GPT tha does that?
i dunno
no, there are set formats
i've not seen a feature like that on bing
@dim cradle brought me to an excellent idea
it only does 1024x1024, 1792x1024, and 1024x1792. For ChatGPT you can write it like this:
The image should have 1024x1792 resolution, portrait orientation, and the best possible HD rendering.
examples of using gpt4's python skills for post processing dalle3 outputs
bah, dall-e started doing this again...
ok, i'm in the penalty box, good night everyone
you should be able to find it by designer
that adds really nice detail
Is that house on fire? The chimney appears to be erupting 🤔
Nice. I really need to start messing with python.
trickiest part is trying to automate it. but it can be done to a degree. just have to have gpt-4 run tests on the output image
the volcano lit the house on fire unfortunately, but Santa can easily afford a new one
Got it working
here's an example for edge enhancement if anyone is curious.
# Function to apply edge enhancement
def edge_enhancement(image):
# Enhance the edges using an edge enhancement filter
enhanced_image = image.filter(ImageFilter.EDGE_ENHANCE_MORE)
return enhanced_image
# Apply edge enhancement to the original image
edge_enhanced_image = edge_enhancement(original_image)
# Save the enhanced image
edge_output_path = '/mnt/data/edge_enhanced_image.webp'
edge_enhanced_image.save(edge_output_path)
# Display the original and the edge enhanced images side by side for comparison
plt.figure(figsize=(12, 6))
# Original image
plt.subplot(1, 2, 1)
plt.imshow(original_image)
plt.title('Original Image')
plt.axis('off')
# Edge enhanced image
plt.subplot(1, 2, 2)
plt.imshow(edge_enhanced_image)
plt.title('Edge Enhanced Image')
plt.axis('off')
# Show the plot
plt.tight_layout()
plt.show()
well, this is what gpt-4 ran on the image I uploaded
every house that i put in an image with the volcano automatically lit on fire from the inside for some reason
i was not telling it to light the houses on fire
it is a wood-framed house
yeah, but the houses in the two previous images i posted (above) were made of gemstones and they still lit on fire from the inside
Is there a problem with daily credits at the moment?
I went to Dall-E on chrome, no problem
Claimed my 4 daily creds in discord, reload Dall-E on chrome, all i get is a blank page
ive cleared my browsercache, doesn't help,
tried other browser, microsoft edge, same problem
Was just testing the first part of your image description on Santa and the elves to see what that might look like
Aesthetically minimalistic yet highly detailed lofi anime artwork, using photorealistic natural colors and a shallow focal length. The image focuses on Santa and two chic elves with athletic feminine physiques and long hair.
Are you struggling to access https://labs.openai.com/ or are the credits simply not showing up 🧐
credits are working, I think you are having some issue with your browser
Aesthetically minimalistic yet highly detailed lofi anime artwork, using photorealistic natural colors. The image focuses on Santa and two chic elves with athletic feminine physiques and long hair. They are on on the beach in Hawaii, and the elves have their arms around Santa, and there is a rainbow. There are lots of coconut trees. An extremely powerful sea breeze is blowing everything. The image should have 1792x1024 resolution, landscape orientation, and best possible HD rendering.
may need to create every image with chic anime elves from now on
this has a strong mix of Santa, geckos, and gemstones, but you should add some chic anime elves if you want it to be balanced
I like how every evening gpt-4 will suddenly forget that it has vision capabilities. Forget that I'm verbatim giving it the same prompts it's processed for me countless times before. "As of my last update in April 2023, I don't have a vision model integrated into my capabilities. My functionality is centered around text-based processing, information retrieval, and generating responses or content based on textual input" "I understand that you're looking for specific image variations. To assist you effectively, I need a bit more detail about the exact nature of these variations. Could you please provide specific instructions or elements you want included in these images? This will help me generate the images that best match your expectations."
it may shift to a slightly watered down model during peak times
It's like pure, unfiltered gaslighting. Like Slurms MacKenzie from Futurama, but excretes lies and deflection
Watered down is a nice way of putting it. It quite literally makes it difficult for users to have their requests fulfilled. I feel like it may have been implemented as a feature rather than a bug, but then it went way overboard with it and then the whole "winter break" narrative was born. But what do I know?
is it doing the unexpected switching of models when you're using the Python API interface?
Aesthetically minimalistic yet highly detailed lofi anime artwork, using photorealistic natural colors. The image focuses on Santa and ten chic diverse elves with athletic feminine physiques and long hair. They are all riding an enormous colorful gecko on the beach in Hawaii and smiling, and there is a storm with lightning. There are lots of coconut trees. An extremely powerful sea breeze is blowing everything. The image should have 1792x1024 resolution, landscape orientation, and best possible HD rendering.
by adding a gecko and increasing the number of chic diverse elves we can create some very powerful imagery
Sometimes, but don't think that's the problem most of the time. Because I can sometimes get it to snap out of it so to speak. Then it apologizes and does what I asked it to do originally. It'll somehow remember the prompt it wouldn't acknowledge in previous responses
Darth Vader with visible teeth jumpscare
A compatriot species to the antoids I shared the other day, here are the wormoids. Half-human, half-worm savages who specialize in magic more than their antoid brethren.
Ficklefangs. Venomous, cave dwelling, cow sized moth creatures that antoids and wormoids regularly tame and use as battle mounts when attacking the surface.
Can you draw a realistic image of an enormous gecko made from partially molten lava walking along a beach in Hawaii at night. An incredibly huge magical rabbit is riding the gecko and aiming a glowing jewel encrusted bow and arrow out to sea with its front legs. The gecko has glowing blue eyes. There are lots of coconut trees, and there is a storm with lightning. An extremely powerful sea breeze is blowing everything. The image should have 1792x1024 resolution, landscape orientation, and HD rendering.
imagine them doing multiplayer in [insert popular multiplayer game here] 🤣
This is an image I made a long time ago on Midjourney V4, and it's still probably my favorite landscape generation I've ever made. It's still my desktop background also, I've tried but have never made one quite good enough to replace it in later Midjourney versions or on GPT.
An attempt with GPT:
It just doesn't compare for me.
That was probably my luckiest generation ever lol, the color gradient is so damn appealing to my eyes. It's leagues above any other purely landscape shot I've tried. And I've tried making many lmao.
Skeksis
I like the palette
if they were playing COD they'd look much more aggressive
oh boy... that's interesting
facts
Are the resolutions high enough to use as wallpaper? I’m waiting for dalle to go higher before I use them as wallpapers
Working hard on an ethereal bot that can capture dreamlike illustration
i like the idea of imagining new lifeforms who aren't out to kill you
Exactly
like trance but alien world
Similar yes. Trying to blend reality to dream state though. Mostly getting one consistent background
mesmerizing 🤩
I think I took it a step higher
Getting the overall composition the way I was hoping
Dreamy
i like the eyes especially
Can you create a realistic photo that focuses on Santa and four athletic and diverse female elves with long hair. They are all running away from an enormous colorful gecko on the beach in Hawaii. There are coconut trees. An extremely powerful sea breeze is blowing everything. The image should have 1792x1024 resolution, landscape orientation, and the best possible HD rendering.
almost time for them to party for new years
the hair is good in this one, but i'm not sure what the black area is trailing behind them, or why water is blasting out of the beach
will try
you could ask one of the mods when they're online
omg
i don't know the answer, but they will
Yeah. Was going to say
Honestly don't know what a dalle3 ticket is, but if that's something openai sells then they'd be who to contact.
It's like this
you use it to generate photos
but I accidentally bought it
I didn't wanted to
Did another for you
Gotcha. What website is that?
Lmk when you have it shon
pretty dang stellar
detail and depth, looking good
Use this to get transcriptions from images. Then, put the transcriptions (which are in plain English) after my image creation prompts.
Well glad you figured it out. That's cool they refunded that quickly
Transcriptions?
Ye if you don’t use api, ChatGPT has the vision model if you are Pro.
(Vision model allows it to make really deep insights about images/media)
lol
Ahh, so similar to vision or the same? Does api not have vision? Or am Injust confused?
Ask it to describe an image so a blind person could visualize it
Perhaps
Asking it to describe an image so a blind person can visualize it is not enough.
You need to use strong linguistics
Well
Which is was SADLMRS is
I feel like I do alright with it
lol
Here is an example
“The image is a complex visual narrative, embodying a multitude of symbolic elements meticulously crafted into a cohesive whole. It presents a figure, almost angelic, emerging from a cosmic dance of vibrant colors and flowing patterns. Every pixel seems deliberately placed, contributing to a grander scheme that transcends mere aesthetic pleasure to approach a form of visual alchemy. The spectrum of colors is rich with symbolism, potentially indicative of various emotional states or cosmic phenomena. The image's construction suggests a layered reality, where each swirl and nuance holds deeper meaning, inviting a microscopic level of interpretation to uncover its full symbology. The fusion of these elements generates a lexicon beyond words, where the visual experience becomes a form of communication, a nexus of understanding between the creator, the viewer, and the universe depicted within this digital tapestry. This synthesis is a testament to the power of imagery to convey complex ideas and emotions with clarity and an almost infinite resolution.”
# Define the core elements of the cosmic entity and figure
def create_core_elements():
elements = {
"face_radiating_light": True,
"galactic_patterns": True,
"celestial_bodies": True,
"geometric_designs": True,
"spiritual_symbols": True,
"feminine_figure": True,
"universal_harmony": True,
"vibrant_colors": True,
"fractal_designs": True
}
return elements
# Combine elements into a unified theme
def unify_elements(elements):
theme = "A cosmic entity embodying the universe, with a face radiating light at the center of "
theme += "galactic patterns and celestial bodies, interwoven with intricate geometric designs "
theme += "and spiritual symbols. A feminine figure dances through this cosmos, in a "
theme += "harmonious blend of vibrant colors and fractal designs."
return theme
# Generate visual representation based on the theme and style
def generate_visualization(theme, style):
return f"Visualize {theme} with a style inspired by {style}."
# Define style variants
style_variants = ["ethereal glow", "vivid dreamscape", "abstract dimensions", "spiritual resonance"]
# Generate visualizations for each style variant
def create_visualizations():
core_elements = create_core_elements()
theme = unify_elements(core_elements)
visualizations = [generate_visualization(theme, style) for style in style_variants]
return visualizations
# Invoke the function to get the prompts
prompts = create_visualizations()
Try that out as a prompt
Modify as needed
amazing
vs 
' capybara in the style of Rembrandt `
Does anyone know of a good guide on what to put in custom gpt focusing on dalle generations?
It looks like this
blurred the bookmarkbar
Before i claimed the daily creddits, the site would load like normal
Even with the monthly creddits there was no issue
it is surely something with your browser
probably unrelated with the fact you claimed credits
clear cache and cookies?
try a different browser too
did you ask pythogoras or ac3d? (for some reason I tag then with at and I got timed out...)
the browser problem, was mentioned yestreday, it wasn't specific for one or the other one, clear cache was also suggested
Interesting, I was able to do this a while ago. Now I'm trying a similar prompt and it's not working. With Bing Image Creator.
hi dys!
heya
I don't really like dall e because he bugs
glamrock shrek by bing image creator because security breach game
you have to be more precise, he bugs?
this #daily-theme message resembles the Stephen Colbert show
I tried Mozila Firefox and microsoft Edge, same result
Clearing cache on chrome doesn't help
ill try again later when creddits should expire and see what happens then
@dusky grove it was also the case yesterday, other people also reported that, dunno if there's something on #1070006915414900886
Battling a Ficklefang, a giant hostile, cave dwelling moth:
Nah, not with GPT gens. It's like the one thing Midjourney has over GPT at this point I think. At least as far as I know.
dont see a report there
you should file one there, since that seems to be a technical issue to access the site
Will do
also, try disabling extensions and restart your browser in safe mode, see if that helps, and turn of vpn if you are using one
then it should have worked with a freshly installed mozila firefox,
almost never use VPN
have you tried using google's or cloudfare's DNS?
eventually are you behind a mobile connection with cgnat implemention?
otherwise I'm out of ideas
dalle can be so weird... suddenly tell me it cannot make image with text but will add space on the iamge for me to add my own. i then make a new message with it and it does it fine
Yeah, sometimes it gets moody
It's like a moody teenager
You have to remind it that you pay the bills
Hi @late blade 
How are ya!?

Doing ok, I’m at the hospital. Had an injury. 😡
Yeah, nothing threatening. Just hurts a lot
Still waiting for the doctor to tell what’s best.
But all ok. But if you want make a get well card for today’s theme. Go ahead 😁
wow, are these with custom gpt or vanilla dalle? 😮
funny how effortless that beach wear is in your images 😄
this is the closest i got to appropriate beach wear.
Yeah i got a session doing that style pretty consistently and my thoughts were 'looks kinda like gta' as well.
that looks really good, like gta style, no doubt about that
just that rockstar would have been more explicit I think
that one turned awesome
i read
@late blade 
haha nice
awesome
Love the stylization that happens here for the art style
beach
and here's a bonus
i forgot to load my json files to my repository... so i only got this in my phone
That method has proven useful it seems?
Vanilla!
Adventurers battling wormoids:
Hi Pytha! I made a glamrock Shrek because, why not?
This one is my favorite generation in a while, perfect lighting and vibes.
Lol. I saw. It was very funny 
Ty, I made because of the glamrocks in FNAF sb (security breach)
Ohh, ok 
Make more.
Different styles of glamour shrek
Hehe
Maybe next time
Dope art, Teddy! Elden Ring would shake its boots!
Ted is becoming a legend with his art in these parts.
Many thanks!
Not what I was going for, but damn if that character isn't a complete vibe lmao. Witch hunter hat over a plate helmet, an absolutely wild mace/axe thing, and some tattered robes to bring it all together lol.
looks excellent. Is this some of fantasy illustration style?
For my style, I often tell it to go for a mixture between graphic novel and comic book art styles, and to try to give me a middle ground between those two.
I saw the minimalist tarot cards in the gallery and wanted to try one.
It's giving me indiana Jones vibes and it's kinda looks like a mixture of Ryan Gosling,Henry Cavill and Chris Pratt 
Really good!
You hit the nail on the head!! 🔥
That's what I put! "Make an Indiana Jones comic character with the looks of Henry Cavill and Chris Pratt"
Wow!! That's amazing how you guessed that...

Yeah pure luck
I know you like but damn that’s flattering 😁
I like this death better, mine kept coming out cute
They need to revamp Indiana Jones to be more like James Bond.
God I got love James Bond movies...
I have to agree
Well this is what I can work with atm. So here’s Paws 🐾. Don’t mess with her! Sadly dall-e refused something similar to James Bond.
Spy agent costume I guess ?
Yes
This is giving me great ideas for my own work. Since I’m doing a live action style for cat girls.
yass
Ok, nothing to worry about. Muscle cramp did rupture the muscle tissue tho, and I’ll have more time for Dall-E’ing more and more.
I just noticed that from her elbow a second hand is coming out to her waist
Any tricks to remove face markings with ChatGPT? Seems to consistently add face markings.
can you give us an example of the markings?
It changes. It'll be sting style make up, and then when I ask to remove the markings, it'll generate a new image with like, weird blood markings. It's always on the cheekbones.
sounds like you are asking dall-e to do something and then forget about it
not a good practice
I'm not sure what you mean by that?
what was your prompt?
I'm using chatGPT, so I've been letting it create the prompts. This was the last prompt:
"Create a full-body portrait of a strong, resolute female character, reminiscent of a futuristic warrior nun, holding a heavy bolter. She has a clear face with flowing white hair, wearing ornate armor detailed in dark blues, blacks, and tarnished golds. The armor resembles that of a Warhammer 40k Sisters of Battle with a unique twist, emphasizing blue as the primary color. The background showcases gothic architecture and a war-torn landscape, typical of the Warhammer 40k universe, filled with vivid details and no whitespace. The watercolor style of the painting should capture the grandeur and intensity of this grim, gothic setting, blending the character and her environment seamlessly."
dont use clear face, instead use fair skinned face, could help you, also the blending with the war environment could also be a factor
Okay, thanks!
I did get a clear shot of her face
Same. This is much better. Thank you so much for the tips!
anytime, we all are creative tinkerers here
Creative and beautiful.
I guess i got a little inspired by the GTA like style 🫣
love it
nostalgia is a relaxing thought.
yeah, nostalgia is what I'm getting from my foot not hurting yesterday...
happy thoughts
lol
yeah, those are very happy thoughts
getting good ideas for today's theme is tough, lot of people have so many good ideas
penalty box
it means the 1980's for me, and getting the elements just right for the period can be challenging. it struggles with the nuances for obvious reasons.
the 80's walkmans!!!
bing time
only if I get an idea, already have 2 in my queue
don't worry, my cat-girls are taking the day off, they are having fun
even catgirls get nostalgic at times, i guess
holy chad
this inspires some gta-style art
@final compass works at Rockstar
This reminds me of a prompt I was using based on a plasma globe. The plasma streams can be modified to take any shape or form, such as a tree, which produced similar results.
HD Surreal and abstract digital painting of a plasma globe against a black backdrop. Abundant vibrant neon plasma streams of ethereal, electrified matter form a mesmerizing contrast against the darkness.
that would explain it. the lengthy load times, until they "solved" it, sort of ingrained it
lol
the hourglass is such a great metaphor
Henry VIII -- a before and after
lol
speaks volumes
he ate good also
mmmhmm, he tucked away the groceries
obviously the quality from bing is not as good, but it's still great for exploring concepts and practicing crafting short art prompts.
ya
apparently it's been in development for a decade. ai might be making some graphic designers nervous.
they generate 1000's in nightly api runs and select the best
it sounds like an expensive strategy, though, maybe the corporations can afford it, but i wonder how efficient that is, it might require a balance with better prompt engineering.
it's helpful to clarify what those memories are for you
Welp, my daily credits expired, and now Dall-E functions like normal
it has played a crucial role in the dramatized and cinematic depictions of certain historical events in some recent first-rate productions
i'm not using labs until dall-e 3 releases.
What do you use then?
chatgpt and bing image creator -- they're both v3, i think you'll be happier with the output
do they support outpainting?
hmmm, i use outpainting to expand the 1024x1024 (1:1) standard output image to 2048x1152 (16:9) wallpapers
It wouldn't be outpainting but ChatGPT does allow you to specify landscape orientation for images which would be similar
you could generate new images using current dall-e 3 tools and use another tool to upsize
gpt4 probably
yeah, sadly need a subscription to use Dall-E 3 in ChatGPT
Trying bing now
bing only takes a short art prompt -- you might use gpt-3.5 to craft it. design your concept, then request a short art prompt. if it's not short enough, ask for a shorter one. it'll condense it efficiently as possible, but it’s not lossless.
and it only returns "standard" quality images
but it's free
Unpopular opinion
I actualy prefer the dall-e 2 output
" A dog dinosaur chimera, digital art"
First is Dall-E 2 with outpainting
Second is Dall-E 3
You can specify styles and much more to achieve your artistic goals
What style would the dall e 2 output count as, as thats the style im looking for
Describe to me please in detail what you see in the first image
Looks like a painting (especially the brushstrokes on the mountains to the left)
Muted colors, washed out, etc.
Yeah, maybe brush and watercolor?
I'd use terms like Digital Painting, Soft Brush Strokes, storybook, concept art... maybe?
Right looks like digital art. It's higher fidelity, for sure, but a completely different medium
off topic, but I really love your profile pic @quartz vale
Thanks! It's Dall-E 3 generated 
These appear like good suggestions, give it a go
trying at the moment
I was surprised ChatGPT specified the dog/dino combo specifically
Y’all nailed it
Likely in the file name
That makes a lot of sense
openai misses nothing ha
I assume the filename is something like Dall-E 2 [prompt]
Dog Dinosaur chimera, digital art 001.png




