|
|
Post by RAVENEYE on Feb 6, 2023 18:50:56 GMT -6
Also, Bird - I've been trying to create the trees on Elivera for you. So far I've gotten these returns:  Lots of purple, and beautiful, but I don't think they're quite right. OMG. That is awesome. The bottom one with the huge-ass trees is closer, but the buildings and infrastructure are in the canopy, and the trees grow a lot closer together (their branches tend to intertwine, since essentially the Raliok forest is one giant tree - like how aspens are essentially all one tree? Certain types of aspens grow in large clonal colonies from one seed). The Cities in the canopy are built over multiple intersecting branches. Hmm, yeah, I was hoping MJ would put the buildings IN the branches, but it never did. I'll try some of these descriptors you mention and see if it gets closer. Edit: Man, MJ is having real trouble with putting the village/city up in the branches. "Does not compute" I'm guessing. Anyway, slowly but surely, we're getting a little closer:   |
|
|
|
Post by Valhalla Erikson on Feb 6, 2023 18:57:30 GMT -6
It has a new blend feature that may help.
|
|
|
|
Post by ScintillaMyntan on Feb 6, 2023 21:00:10 GMT -6
Apart from ethics issues, I'm feeling pretty uncomfortable with the idea of machine-made media replacing media by humans. What part will artists, including creative writers, have in the world anymore if the algorithms can do it all? If a company could type some commands instead of hiring human artists? And it just feels so bleak, the thought of having our aesthetic feelings only stirred by emotionless machines creating things according to a set of rules. I do get that at least there's a human hand directing the machine and deciding it's worthy to post, but that's not that much. I comfort myself by saying that humans like to create too much. There will always be manmade art as long as we like and feel the need to make it, but then, what if humans can't compete? I'm imagining something like how it's hard these days to get non-mass-produced goods; you have to deliberately go out of the mainstream to buy handmade stuff. Would it be like that for manmade art? Like all the radio stations, game companies, movie theaters, and so on just publish AI things, and you have to specifically seek out manmade media on your own and also question whether it really is manmade because it's hard to tell anymore? I say all this without knowing much about the subject, so these are really just personal wibbles. But it gets me uneasy whenever I see this stuff posted around, no matter how beautiful it can be. The violet colors in these are really close to the violet colors of Elivera though!!! You know, I'm glad someone else distinguishes violet and purple! When I realized 'violet' is the correct word for the softer, bluer shades that I usually find much more beautiful than purple, I've been trying to remember to use the right words for both. It's hard though; I'm pretty used to the word 'purple' for all of them. |
|
|
|
Post by RAVENEYE on Feb 6, 2023 23:44:30 GMT -6
Apart from ethics issues, I'm feeling pretty uncomfortable with the idea of machine-made media replacing media by humans. What part will artists, including creative writers, have in the world anymore if the algorithms can do it all? If a company could type some commands instead of hiring human artists? And it just feels so bleak, the thought of having our aesthetic feelings only stirred by emotionless machines creating things according to a set of rules. I do get that at least there's a human hand directing the machine and deciding it's worthy to post, but that's not that much. I comfort myself by saying that humans like to create too much. There will always be manmade art as long as we like and feel the need to make it, but then, what if humans can't compete? I'm imagining something like how it's hard these days to get non-mass-produced goods; you have to deliberately go out of the mainstream to buy handmade stuff. Would it be like that for manmade art? Like all the radio stations, game companies, movie theaters, and so on just publish AI things, and you have to specifically seek out manmade media on your own and also question whether it really is manmade because it's hard to tell anymore? I say all this without knowing much about the subject, so these are really just personal wibbles. But it gets me uneasy whenever I see this stuff posted around, no matter how beautiful it can be. I felt this way when hand-drawn animated movies were replaced by computer-drawn animation ( Aladdin was the first Disney animated on a computer instead of by hand, if I recall). And when it became almost impossible to find hand-painted cover art on books and all the "characters" featured on covers suddenly became photographed human models mashed on top of some photographed/photoshopped setting that may or may not accurately depict the story's content. At that point, book covers felt as if they lost all soul. I terribly miss the old hand-painted cover art by folks like Keith Parkinson and Ciruelo and etc. So in a huge way, replacement of human artists has already happened. At the same time, in every one of these media types, the machines are working in tandem with humans. The machines would produce nothing without human creativity, even if it's not the everything-by-hand variety. One can buy a machine/factory produced coffee mug, or they can buy a handmade one from me. It's nothing new, just sad to see certain artforms transitioning into something rare and ... endangered? |
|
|
|
Post by Valhalla Erikson on Feb 7, 2023 2:41:59 GMT -6
It's an unfortunate inevitability that we can't avoid. The more technology continues to advance the more we'd gradually start to lose the more practical things.
|
|
|
|
Post by havekrillwhaletravel on Feb 7, 2023 3:51:20 GMT -6
Part of the argument against the AI training sets is how the sets work:
1. The training set takes the EXACT COPY of the work EXACTLY AS IT IS. It does this for MILLIONS of artworks. There is usually no asking for permission. The equivalent of this with writing would be if you took several fantasy book verbatim and reprinted them under your name. All of this artwork is then part of the database that the AI will pull from when it gets its queries, meaning all of the art taken without permission exists within the AI engines in their original form.
2. The AI algorithms then go through the training sets again and again to build up profiles of certain tag words. So when users input those specific words, the AI knows to pull up that piece(s) of art to smash together with another piece(s) of art. This is the recombining and creating anew. But to get to this point, millions of artwork was taken without permission and used to train the AI enough for it to actually try to create something different.
The issue is with step 1. That's where the legality of it all is starting to come to a head. And I think it's right to examine the ethics and morality of step 1. When people create things, they don't create them to be used to make fancy-tech-AIs. They created them to be enjoyed by others. There's a consent element here, that big tech refuses to acknowledge, but we MUST acknowledge if we are to build a better world.
Ooh, yeah. I wasn't aware of what went into the training process. Hmm, that sounds a lot harder to justify and much more ickier than I initially thought.
Apart from ethics issues, I'm feeling pretty uncomfortable with the idea of machine-made media replacing media by humans. What part will artists, including creative writers, have in the world anymore if the algorithms can do it all? If a company could type some commands instead of hiring human artists? And it just feels so bleak, the thought of having our aesthetic feelings only stirred by emotionless machines creating things according to a set of rules. I do get that at least there's a human hand directing the machine and deciding it's worthy to post, but that's not that much. I comfort myself by saying that humans like to create too much. There will always be manmade art as long as we like and feel the need to make it, but then, what if humans can't compete? I'm imagining something like how it's hard these days to get non-mass-produced goods; you have to deliberately go out of the mainstream to buy handmade stuff. Would it be like that for manmade art? Like all the radio stations, game companies, movie theaters, and so on just publish AI things, and you have to specifically seek out manmade media on your own and also question whether it really is manmade because it's hard to tell anymore? This is something I'm also worried about, but I'm cautiously hopeful AI art will remains its own genre, and people will continue to seek out and consume human art. Computers have been able to consistently beat us at chess for decades, but we're still interested in chess games between people. Nobody really wants to watch two computers go at each other on a board. Maybe once the novelty of AI art has worn off, it'll be used much more selectively and sparingly.
I asked ChatGPT about this:
Yeah, okay. Flattery will get you nowhere, ChatGPT. *stares daggers at the algorithm*  |
|
|
|
Post by RAVENEYE on Feb 22, 2023 15:45:23 GMT -6
So Bird posted in a different thread (Litterbug Junction) an image created by the AI Art creator NightCafe. So I thought I'd give it a shot. I'm probably doing it wrong, bc the results I'm getting over at NightCafe lack the drama and beauty of MidJourney's results. (Bird managed to get a great result somehow, but mine have been just gross. Using the same prompts this is what I get: Prompt = "a majestic white castle city surrounded by a lush forest, the forest is on fire, smoke billowing, Octane render, gloomy, violent motion, cinematic lighting, in the style of Donato Giancola" MidJourney: (drama, threat, gloom, texture, SCALE) NightCafe: (weak-sauce, like, for one, where's the effing fire? Is that happy glow behind the tower a fire? Or a sunrise?) Actually, the second pic is the SECOND attempt with this prompt at NightCafe. The first was even worse:  (I don't even know what this is. Is that a racecar???) Instructions at NightCafe say to evolve an image with more modifiers. ... The prompt I used already has TEN modifiers to "a castle, a forest on fire." I'm not sure how much more specific that prompt could be. I've even told it what style and lighting to use. So I'm not yet sold on buying more credits at NightCafe. |
|
|
|
Post by Valhalla Erikson on Feb 22, 2023 15:49:59 GMT -6
Using MidJourney I've managed to create a perfect character image for my main character in my novel Shepard Mysteries.  The prompt I use is a good-looking 13-year-old Japanese boy with short black hair and wears glasses on dark golden eyes. He has a muscular build and is a lot wider than the other boys in the class. He has a somewhat stoic appearance but can be very funny. |
|
Bird
Counselor 
Posts: 350
Custom Title: World Creator and Destroyer
Preferred Pronouns: they/them/their
HARD: 1700
MEDIUM: 400
EASY: 110
|
Post by Bird on Feb 22, 2023 16:18:32 GMT -6
So Bird posted in a different thread (Litterbug Junction) an image created by the AI Art creator NightCafe. So I thought I'd give it a shot. I'm probably doing it wrong, bc the results I'm getting over at NightCafe lack the drama and beauty of MidJourney's results. (Bird managed to get a great result somehow, but mine have been just gross. Using the same prompts this is what I get: Prompt = "a majestic white castle city surrounded by a lush forest, the forest is on fire, smoke billowing, Octane render, gloomy, violent motion, cinematic lighting, in the style of Donato Giancola" It's not descriptive enough on the art style. I added a few more modifiers and adjusted some of the advanced parameters to get these four options:
(What I used for modifiers and prompts: "a majestic white castle city surrounded by a lush forest, the forest is on fire, smoke billowing, Octane render, gloomy, violent motion, cinematic lighting, in the style of Donato Giancola photorealism digital illustration fantasy Unreal Engine 8k resolution." I then set the parameters to Fantasy Style, weighted the prompts used at 70% - if I went higher I might overcook the image. Then I upped resolution a bit, added in a higher run time, used model Stable Diffusion v2.1, sampling method of K_LMS, and hit the render button):
Midjourney does a lot of that for you, but at the same time, it doesn't let you alter any of these advanced options to the degree that NightCafe allows. So if you want to test out different sampling and model render techniques to find the ones best suited to your preferences, NightCafe is one of the few that allow that.
Also, you get free 5 credits a day. (I thought it was four, but then checked my account, and yes, it's 5.) So if you just wait and do your prompt-render-runs once a day, you can get some high quality stuff for free.
There's a guide on the site on how to really up the ante to get even better results than MidJourney. But again, it does have a higher learning curve. So your mileage may vary.
EDIT: Also, you can take any image that you created and then evolve the image to transform it to closer to what you desire too.
|
|
|
|
Post by RAVENEYE on Feb 23, 2023 8:44:49 GMT -6
So Bird posted in a different thread (Litterbug Junction) an image created by the AI Art creator NightCafe. So I thought I'd give it a shot. I'm probably doing it wrong, bc the results I'm getting over at NightCafe lack the drama and beauty of MidJourney's results. (Bird managed to get a great result somehow, but mine have been just gross. Using the same prompts this is what I get: Prompt = "a majestic white castle city surrounded by a lush forest, the forest is on fire, smoke billowing, Octane render, gloomy, violent motion, cinematic lighting, in the style of Donato Giancola" It's not descriptive enough on the art style. I added a few more modifiers and adjusted some of the advanced parameters to get these four options:
(What I used for modifiers and prompts: "a majestic white castle city surrounded by a lush forest, the forest is on fire, smoke billowing, Octane render, gloomy, violent motion, cinematic lighting, in the style of Donato Giancola photorealism digital illustration fantasy Unreal Engine 8k resolution." I then set the parameters to Fantasy Style, weighted the prompts used at 70% - if I went higher I might overcook the image. Then I upped resolution a bit, added in a higher run time, used model Stable Diffusion v2.1, sampling method of K_LMS, and hit the render button):
Midjourney does a lot of that for you, but at the same time, it doesn't let you alter any of these advanced options to the degree that NightCafe allows. So if you want to test out different sampling and model render techniques to find the ones best suited to your preferences, NightCafe is one of the few that allow that.
Also, you get free 5 credits a day. (I thought it was four, but then checked my account, and yes, it's 5.) So if you just wait and do your prompt-render-runs once a day, you can get some high quality stuff for free.
There's a guide on the site on how to really up the ante to get even better results than MidJourney. But again, it does have a higher learning curve. So your mileage may vary.
EDIT: Also, you can take any image that you created and then evolve the image to transform it to closer to what you desire too.
Yeah, I got somewhat better results when I evolved the image and played with the sliders. Like, DO WHAT I SAY, AI. Irritating that it can't produce a high res image without being told to and stick to the prompt unless the sliders are used. Oh, well. I do like the evolve capabilities and want to see what happens if I plug in an image from elsewhere. Like, what will NightCafe do if I upload a MidJourney image and mess with settings. That will be fun... Anyway, thanks for showing me the prompt you used! |
|
|
|
Post by RAVENEYE on Feb 23, 2023 8:46:19 GMT -6
Using MidJourney I've managed to create a perfect character image for my main character in my novel Shepard Mysteries. The prompt I use is a good-looking 13-year-old Japanese boy with short black hair and wears glasses on dark golden eyes. He has a muscular build and is a lot wider than the other boys in the class. He has a somewhat stoic appearance but can be very funny. OMG, he's so cute! Love it! |
|
|
|
Post by RAVENEYE on Mar 2, 2023 17:20:50 GMT -6
Learning how to get uber high-quality images out of MJ now. I have no idea what some of the terms mean that I'm using, but they seem to work.  Copying them from other peoples' images. Prompt: a brunette sorceress with gray eyes wearing a silver robe, lavender fire, stormy sky, lightning ::5 Premium Portrait Photography::4 sharp-focus, high-quality, ultra-realistic, realism, artistic, unique, award-winning photograph, Canon EOS 5D Mark IV DSLR, f/8, ISO 100, 1/250 second, close-up, cinematic lighting, professional, flattering, full body, glibatree style ::3 black and white, grainy, deformed, cartoon, painting, drawing, animated, watermark ::-2 --no freckles --ar 2:3 (This is Carah, by the way. She's the heroine of Falcon Saga Books 3, 4, and 5.)  Incidentally: I plugged the same prompt words (sans weight numbers) into NightCafe and get this awfulness:  This lady just woke up, put on a literal silver bath robe and walked out into the ugly countryside she lives in to watch a storm pass. WTF. Where is the sorceress I ordered? The magic? Yes, I evolved the image several times hoping for a better result, and it just didn't happen. SMH... |
|
|
|
Post by Valhalla Erikson on Mar 2, 2023 20:47:35 GMT -6
Now that's a heroine I'd follow! Interestingly enough she kind of resembles an urban fantasy variation of Rey.
|
|
|
|
Post by Valhalla Erikson on Mar 2, 2023 21:20:02 GMT -6
This is a concept of a book cover for my novel.  The prompts I used for it is a 18-year-old Japanese male with a black shirt and combat pants and short brown hair, he rides on the back of a dragon in the middle of a gothic Victorian town. For the curious his name is David King. He's the main character for my Shepard Mysteries novel. |
|
|
|
Post by Valhalla Erikson on Mar 2, 2023 21:27:57 GMT -6
One thing I love about MidJourney is how it helps me get a better visualization of my characters. Case en point. 
Prompt: A beautiful gothic asian woman in her 20s. dragon tattoo on her arm. urban fantasy style. Izumi "Izzy" Miyazaki from the same novel. She's another main character, a childhood friend who helps David when he's conducting cases within the town of Shepard. She's part of an all-female magical species who are descendants of The Fae. What makes her character intriguing is she dabbles in dark magic. |
|
Copying them from other peoples' images.
You add them to your profile as you do them around the forum.
