New Free AI Art Generator Does Something Unique / A Look at Prome.AI
Vložit
- čas přidán 11. 09. 2024
- Today we're taking a look at an interesting new AI Art Generator that does something fairly unique in terms of its prompting methodology.
This is PromeAI, from the team at Cutout.Pro. I think Prome is a really interesting tool for those of you who have been looking at doing AI Generated Comic Books, and we'll be taking a look at how the AI fares by first taking images and sketches from comics I have worked on in the past and turning them into photo-real, or stylistic images. It works a bit like Midjourney's Image Prompting, but takes your images less as suggestions, and more as direction.
Additionally, if you AREN'T great at drawing, I tried out one of my daughter's sketches (Look, I'm not saying she's bad, she's just ten!) and using Prome to render out one of her pieces of art.
Prome.AI is free, but only allows you to download the lowres images. They do have a Membership tier that will allow access to the high-res versions. So, give it a shot and see what you think!
As an FYI, this video was sponsored by Prome.AI and Cutout.Pro, but I honestly think the platform is interesting and worth checking out, so I wanted to bring it to you.
LINKS:
PROME.AI: www.promeai.co...
CREDITS TO CUTOUT.PRO: First Five Comments to ask get 10 free credits to Cutout.Pro
My Daughter's AI Nightmare: • My Daughter had an A.I...
AFFILATE LINKS:
Camera: amzn.to/3yXMDY2
Microphone: amzn.to/3K1jIZm
Audio Interface: amzn.to/3lDX9kf
Coffee: amzn.to/3JZuBeq
Follow Me on Twitter: / theomediaai
#promeai #midjourney #stablediffusion #AIGC #sketchrendering
-------------------------------------------------
Thanks for watching Theoretically Media! I cover a large range of topics here on the Creative AI space, Technology, Tutorials, and Reviews! Please enjoy your time here, and subscribe!
Your comments mean a LOT to me, and I read and try to respond to every one of them, so please do drop any thoughts, suggestions, questions, or topic requests!
.
Try out PromeAI here: www.promeai.com/?vsource=youtube_share-theoreticallymedia202305 Let me know what you think of it!
👋
@@LouisGedo Hi Louis!!
@@LouisGedo 👁
@@-TheRealThing- 👁
Sadly.... it does not seems to work.....
Keep on getting Generation Failed. :(
check here the new domain!!!!!!!!🐱
Ah! So it has been updated!
yes!brand new pro@@yubumao3165
finally find this one
grrreat ur profile
works now with pro
I can see this being a wonderful little bridge between synthographers and traditional artists. Thank you for sharing this. 👍
Agreed! I'm really excited for the folks that have been wanting to do AI powered comic books. I think this could be a real gamechanger for them. With the in-painting, it solves a lot of those wardrobe problems that we run into in Midjourney. Faces seem pretty consistent too, depending of course on the level of fidelity in your original sketch.
If anything, I think this could serve as a real bridge between traditional artists and AI Artists. You could hire a traditional artist for layouts and sketches, and then finalize in something like this.
I don’t believe this can count even as a wooden beam bridging two sides. The war between a parasite and a host cannot be bridged, as AI bros only sees traditional artists as fuel and resources to exploit, while traditional artists and their skills will be lost to time. The whole point of generative AI is to be a vending machine with as little human interaction needed as possible. Future AI models would just incinerate this feeble little bridge and completely devour traditional artists.
Personally I feel that Stable Diffusion image-to-image does a better job, and even more amazing if you combine it with controlnet modules.
I can't say I disagree-- but I think this is a good tool for those who are newer to AI Imagery and don't have the computing power (or the technical skills) to get SD running. I see this as a good place to start (past Bing Dall-E) for those who are looking for more control of their output. Ultimately, some will be happy with this, some will move on to the next step. Different tools for different skill sets.
@@TheoreticallyMedia That's fair. Eventually, with any luck, SD will have much simpler access to running locally. It's all so exciting!
@@not-alot-of-options i have used most ai tools and promeai gave the quickest acceptable results.
For my requirements, obviously.
Some people have special needs and true control net is the best way yet.
I've always been someone who has been writing, never having anything close to the ability to draw. Probably a lot of people out there like me who now have access to tools which can take care of a lot of the drawing aspect. It's really quite something to experience, and reminds me of the feeling when the internet first came about and being completely overwhelmed by it haha. We really are only limited by our imagination at this point!
It is really a fantastic time. I was able to draw at one point, but kind of lost those skills over time. I think they did hone in a sense of good composition for me, which was helpful for camera work.
What I love about a tool like this is that I'm still able to compose images (roughly) and have it render. I think of these tools as artistic collaborators in a way...just an amazing time.
And next week, I'm planning on getting back to the movie side of AI! That's a WHOLE other ball of wax!
It’s awesome! I used to draw a lot 20+ years ago when I was in high school. Haven’t drawn since then. I’ve been trying to get back into it on and off for almost two years now but it seems that talent isn’t coming back and it sucks. Ai art is beginning to be a great thing now and I’m enjoying it.
The crazy Thing is you can Switch drawing and writing in your comment and it makes just as much Sense.
Absolutely blown away by PromeAI's capabilities and how it can transform sketches into photo-real images. It's incredible to see how even a child's drawing can be rendered so beautifully. The platform definitely seems promising for aspiring artists and comic creators. Thanks for sharing this! Always great to see AI's potential in the creative world. And shoutout to your daughter for her contribution! 👏🎨 #AIArtEvolution"
Awesome to hear, I actually did a follow up here: czcams.com/video/P1-VpUFHoEo/video.htmlsi=snUUe4WezzhV5Zd4
Some new toys in there as well!
just tried it out and it's amazing! i didn't think it would work so well but i'm blown away how specific you can get with the prompt
It's pretty nuts right? When they first sent it over I was like, "ehhh, maybe?" And then I tried it and was like: "I need to make a video about this!" 😀
It is amazing how quickly these tools are maturing!
It boggles my mind to think where we'll be in 2 years!
This has been an amazing tool for me. I can draw, sketch, paint and have been for years. The great thing about Prome A.I is I can just sketch, bang my sketch up and walla, gives me all I need and I don't have to think about lighting, shadows and all the thinking I used to do. It becomes my own reference to then download to my ipad and I paint on my canvas with Oil, sometimes watercolour. It turns my thumbnail sketches into paintings and one by one I lay them out on photoshop to see which one to choose, print it out, switch off computer and i grab a new canvas, clean pallet and tube paint to start on a new piece. Been waiting for this for years! I have the ideas already in my mind but the tedious planning was daunting as always but with Prome it does it much faster and does my thinking to just be creative to paint on my canvas. I still use traditional method painting in the end.
Oh, that's brilliant! You are EXACTLY the person I keep talking about. As an artist, AI tools are an amazing resource to supercharge your output. And you are proof that it isn't "handed" to you with a simple prompt-- there's still a ton of elbow grease you have to put in. Really happy you found this video, and thank you for the comment-- it really did make my day!
dam that glitch gave me an idea for a new piece of art...lol a tattoo that is continued as a pattern on the clothing...
Oh, that is a cool idea! I mean, it would be one very specific shirt that you’d have to have like 40 versions of/- but it would be really awesome
I would LOVE some free credits! I am thrilled by all of the new AI wonderfest!!
Cool! Shoot me an email with your email address and I’ll pop it over to the Cutout team!
Thanks for showing us another tool. I'll give it a try. Also, I'd love some of those Cutout credits! 👍
You're the first! Drop me an email address and I'll shoot it over to them: theoreticallymedia@gmail.com
Wow! This blew my mind. I found the interior design use case especially useful. Thanks for the great content!
Glad you found that! I wanted to highlight that part as well-- it's really cool, and I'm not even an interior designer!
Following since a couple of months. And I just love the wonderful content that you put out, your thought put into it as well. You are continously highlighting super cool tools and tips! Just wonderful. My wife was absolutely blown away right now and she directly tried it out (got a remarkable to draw with a pen -> thats a killer combo with that tool :) ).
Excellent to hear! Funny story: I actually lived in Germany when I was really little. I have a lot of those "core" memories-- namely Pretzels and Schnitzel, haha!
Anyhow, yeah- I think this is an incredibly good step to bridge traditional artists and AI. It doesn't negate any of the skill of an artist, while being able to add in details as a collaborator. I can't wait to see how this develops down the road!
@@TheoreticallyMedia haha lovely :) Then you were too little to actually try our amazing beer! And on the topic:been playing around with it more and feels even a little more like actually creating. Can see the tool combining super well with the recent announcement of generative Fill by Adobe and your Faceswap tips as well .
Hi @TheoreticallyMedia, first thank you for this video. I will try it but I will need to find some sketches as I do not know how to draw.
Honestly, try just going as rough as you like and try it out. The Molly sketch was...well, not professional? Haha, and still ended up with some good results-- you just might have to play with prompting and the different models a bit.
You can also try just photobashing some stuff together in a photoeditor and see how it does!
The funny thing is, I actually liked the drawing on the left in your thumbnail the most... by far.
Oh, Ech is great! You can check out more of his art here: instagram.com/ech_illustration/
@@TheoreticallyMedia Cool! Thanks!
This isn't working for me at all, even when I try to use their sample sketches. I click GENERATE and it just gives me 3 blue squares that say "File not exist" and "please re-generate."
I ran into that a few times-- I think they might be having some backend issues. I'd say wait a few hours, or circle back tomorrow. Through the week I was testing it, things also moved around, like the site itself, so I think they're still working on everything-- like everything in the AI world, its a work in progress!
This is truly jaw-dropping! I'm absolutely thrilled to check it out.
Excellent to hear! I think this one is super fascinating amongst all the AI image generators!
As an artist, I keep saying well it can't do that or it can't do this. At this rate, we really are entering the era of best ideas. However it is not lost on me, that the process and understanding behind color theory, lighting, anatomy, perspective, backgrounds, all of this stuff that takes years to learn is going to be lost overnight to these advances. Which to me, is heartbreaking.
I'm not saying don't use it, ultimately unless Congress gets involved and actually regulates this on a copyright infringement level for AI being trained off of other people's content without their consent, it is what it is.
I honestly think that artists are going to be fine. I've got a neighbor's kid who is going to Art School this fall, she's teaching herself to integrate AI into her workflow. Fundimentals will still be taught, composition/color/etc will still be learned. Even something as basic as the Rule of Thirds, because while AI can do a lot, it can't take a photo of a wedding...or a family vacation.
Oh, and never have faith in Congress to do...well, anything. I'm not even talking about AI. Just...anything. Ugh...
@@TheoreticallyMedia I really do hope that you are right, because more and more it looks like it is going to be impossible to be valued.
0 noise, very clean image, you maybe do not remember me, but you look FANTASTIC now
Thank you! I do too! Was thinking about you when playing with settings. Honestly I think I had the ISO cranked too high. Plus, I slowed the MJ animation down by like 400%, so that helped on the compression side.
It’s pretty dialed in! Minus the overall color change because of the sunlight, but there isn’t much I can do about considering there is a ceiling window,
I love it, but it sucks for shooting at certain times
I tried it but got really non sense results. Then I realized it was dropping the images I have uploaded. I mean, the reference picture wasn't showing on the sidebar. Maybe an interface bug?
Could be. I ran into a few issues where the image wouldn't generate for an hour or two. I think they're still working on the UI. I did notice in my week of testing that modules were changing spots, and some even dropped entirely. As with all things in the AI world: It is all a work in progress!
Always keen to look at new methods of character development, many thanks for bringing this to my attention
Glad you found it!
Can you use batch sequence? If not, do you know another alternative to SD+controlnet to use batch sequencing? Thanks!
I didn’t see it, so I don’t think so. Maybe it’s something they’ll add in the future? I could see that fitting in on the cutout.pro side. So far, I haven’t seen any service that is going batch, but I’ll look around. Trust me, if I find it: I’m making a video about it!
Wow !! Crazy !! Thanks for the discovery. I guess this tool is going to give a big headache to copyright offices of the wolrd ! 😂😂
Oh, its all a mess...That's why I say have fun now, before the grownups show up and take our toys away!
Good results for me. Definitely understands line art better than most right from the start.
Agreed. I’m not sure what extra secret sauce this one has going on, but the results I get from image references are way more accurate than most of the others I’ve played with.
I used to render images like 20 years ago but it was very demanding in every way and very slow so I abandoned that, I still made tons of crazy scifi skecthes, I threw away a bunch of them...never thought it would would be possible to feed them into a program and render them. I still kept some so I'm going to test PromeAI, for scifi landscapes. Thanks for the tip!
That program will get me back into drawing.
Accurate. I went out and bought a sketchbook after playing with it. Its been AWHILE since I've had a sketchbook!
Hi can I use this to just make my simple art pop out more, like all my pencil lines become animated and I can download the image to digital?
I’ll have to take a look, it’s been a minute since I’ve checked in on Prome! I’ll keep that in mind for a follow up video!
Thanks for all the guides love seeing someone with an art and film background
Thanks for watching!! Any good movies you've seen lately? Long weekend and I actually think I might have some free time!
@@TheoreticallyMedia im into sci fi stuff like dune. Looking to use mj to generate concept art for an epic landscape journey based with a floating city of mist and stone. I agree with you on Barry Lindon and the Napoleon movie. Such a shame we never got it.
@@JS-zd4yp I cannot wait for Dune 2!! Huh, I wonder if I’d like Barry Lyndon more if it were set on Arrakis? I should roll up some images to find out!
Smaller scale, but have you seen Silo yet? It’s pretty good!
@@TheoreticallyMedia that is a hilarious idea
crazy renders.
Sometimes crazy cool, sometimes crazy insane, sometimes crazy terrible. I love the unpredictability of AI generation!
This is the best kind of tool for people who can draw and sketch but can't paint or render.
Same thought. I'm not even that good at sketching anymore, but I can play around enough with it via AI to get it there now!
From this video .... That AI SOFTWARE _looks_ amazing. The computer *makes* a brand new detailed image just from either a rough sketch or from a contrasty photograph. Technology is superb
Sir This is really nice tool even i used it for production for Animatic Sketch to Image , and i genrated some background from sketch after getting image i tweak it in photoshop but it cuts off my 60% of work and may be its free ..really thanks for your valuable suggestion of this AI tool..love from India
Excellent to hear! I’ve been kind of wondering how the lowRes images upscale. Have you been doing that as well?
Impressive! It doesn't seem to have a huge sample database but when it works it works well.
They're pretty new, but I think they've got something pretty good going on-- As with all things AI, in a few weeks it'll probably be 10 times what it is now. Things are so crazy...
Very good information sir it helped me quite a lot thank you so much your all contents are very helpfull 🙏🙏
excellent! Thank you for the watch and comment!
Anyone knows of any good ways to control the lights and camera angles on ai art generation?
Currently, everything is just hit or miss.
Camera angles I feel an image generator like this is kind of doing, like, you can sketch your scene or camera angle.
Lights? Kind of? I generally do prompts like “horror movie lighting” or “Film Noir Lighting” to give it an idea of what I’m looking for.
But full control won’t happen until we move more into a 3D space and image generators are acting more like Blender or 3DSMax.
7:42 If you do get to play around with photobashing, I would love to see a video on that 😊
I actually did do one! but not in Prome, more MJ based: czcams.com/video/qeRVBKi1QMQ/video.html
Very nice and informative video sir. I'd like some credits to try cutout :)
Cool, shoot me an email with your registered account and I'll pop it over to the Cutout crew!
Brilliant stuff! I tried it out and yes, this really worked out the way I expected it to be. Awesome. The pro version is a bit too expensive tho. 39 per month is kinda steep. #99 (dollars) per year is pretty grandiose too. But the demo/free plan is doing a great job if you just want to play around and gather some ideas on how your scribbles can turn out if they were rendered in 3D for instance.
Yeah, I hear you on the cost. Between MJ, CharGPT, Kaiber and a few others-- it's starting to rack up. I kind of see all the various AI tools to be a bit like streaming services. Cancel Netflix, grab Disney, cancel that, grab Amazon-- etc...Just gotta figure out a workflow for when I'm using each.
But, in general, you're right on the money (no pun intended) with Prome.AI-- the free to pro tier is a really smart move on their part. That way, when/if you decide to go paid, you aren't waffling around with that time!
awesome thing, thank you again!!
100% And happy birthday to the good dog!! Sorry, didn't get back to that comment earlier! 😀
Do they censor prompts (text and image) as well?
So, I didn’t run across anything, but my prompts were fairly benign. That said, I’ve run across some weird language model triggers in MJ in the past, like “cutting a cake” where “cutting” was banned.
But, I didn’t run into that with Prome.
Thanks so much for taking the time to reply :)
Cool video! How good is it at creating consistent characters? So if I wanted to turn a sketched storyboard into something nicer, could I generate the same character in different panels? Or is it a roll of the dice each time?
A little A, and a little B. I think it depends on how much fidelity you have in your sketch. Like, I ran some experiments with Batman comic pages, just the pencils-- no inks, no color-- and it came out pretty solid. I mean, but that's Batman. Give it a shot and use the precise mode. Experiment with your prompt-- once you get it, you should be able to repeat it.
One thing I like about Prome is that it is free on the low-res version, so you can really spend some time honing in before deciding if it is worth it.
Nice but nothing new. This can be done with any stable diffusion image to image model
Indeed. I think this is the kind of thing that is really useful for MJ users, or those who do not have the hardware/computer power to run SD. There is Leonardo.AI, which does similar things (I've covered them on the channel in the past, and I really should circle back there), but I have heard a number of people say they were overwhelmed by Leo (and SD in general)'s ramp up.
I actually like this one due to the simplicity.
i was searching for this for quite a long. Thanks a ton ! Subscribed and liked. My question is after i generate such a realistic pic from a sketch, how can i upload it somewhere and opt for variations? Is there any FREE site where it can be done and how? Please help . I am from India.
So, I would say to take your image and head over to Leonardo.AI and run an image to image on it for variations. You may or may not get exactly what you're looking for, depending on how character accurate you're aiming for. I might have some Leonardo content coming up soon, so keep an eye out for that!
As a novelist, I can't wait to create illustrations of all my characters.
I played around a bit with putting character descriptions from famous novels into Midjourney. It was interesting that just putting the description (no name) of Lisbeth Salander (Dragon Tattoo) it came out looking like a David Fincher character.
Has been around in Stable Diffusion for a while now?? Using ControlNET which is also free :)
True. But I’ll say that a lot of the audience focuses on MJ, so it’s new to them. And a widely requested feature as well.
The only other service I know of that offers this is Leonardo, which I have covered here before as well (actually need to circle back to them at some point), but there are some who feel overwhelmed by Leo’s interface. I actually like Prome for its simplicity.
@@TheoreticallyMedia Im one one who never uses MJ or haven't heard or Leo or Prome so.... maybe you dont know your audience so well ;)
@@PhotoRestorationVideos ha! You’re one of the rare birds in that case! Are you running SD local exclusively?
I tend to shy away from that area, only because it’s so well covered by many other bright folks on YT. And there is this weird rabbit hole you fall down where things can quickly move into the heavy technical side as well.
I kind of like working in this middle area of cool tools and artistic inspiration. Like, “here’s a cool tool, you don’t need to spend 10 hours configuring it, go make that thing you’ve always wanted to make!”
@@TheoreticallyMedia Im into cool tools that i can use and am willing put the time in learning for it make the thing "exactly" as id like, not similar or ish or near to.
really interested in how to get consistent images and styles of different characters and scenes. I find that the most challenging in all these tools.
often by wrestling with seeds, although that's still an uphill battle. I think with PromeAI, it'll be a matter of a good input source, and then using the inpainting (here called Erase and Replace) feature.
It’s no better than wonder so I’m sticking with that
Wait, Wonderstudios? Or am I missing another generator?
How did you get decent results with this? I had an artist make some nice inked line drawings for me years ago. I was hoping this would let me bring them to life in color. No matter which prompts I used or which style filter I selected, almost all of the results with promeai looked like nightmare fuel. Although it would get the general pose correct, it really distorted the hands, feet, and faces. Only one drawing turned out ok after 30 tries and I'm guessing that's because it was a profile image with fewer details to get wrong.
Let’s check it out. When you get a minute, pop over to the channel discord and drop your ref image in Project Help. I’ll give it a shot on my end and we’ll see what we can do: discord.gg/bZJmJK9y
I have a question for you sir... But it doesn't pertain to text to image AI...
I'm a young writer/illustrator. I have two FULL LENGTH novels I've completed, and I'm not really sure where to go with them.
You mentioned that you had published your work, and I know a novel is vastly different than a comic book, but I'll take advice wherever I can get it.
I'm looking to have my novels published, but I'd also be open to a collaboration with you if you are interested in my stories and want to convert it into a comic.
Essentially, I just want to get my work out there. And I need to start pursuing this sorta thing as a business because my financial health is dwindling at the moment.
Any words of wisdom you can offer to me would be greatly appreciated..
Do me a favor, pop on the discord here: discord.gg/yHzGgWtb and drop this in Project Help. I think advice/communication would probably be easier there.
Thanks for sharing this tool ! It's exactly what I was looking for :)
Thank your, nice video!
I’m ready to publish all of those comic books i never launched
Right there with you!
StarryAi does this too, maybe many others so I don't think it's unique at all.
I’ll check it out!
Problem with all these generators I have is that they focus so much on people/characters. For my art I would need a generator that can handle fantasy and sci-fi item/objects and make them look realistic but on a black background without any scene behind it. But I haven't found any generator that can give good results in that area even with very refined prompts. Either look horrible or makes them drawings/stylized. Wonder if this can do it or if it also tries to make a sketch into persons at all costs.
You should give it a shot. Particularly since this inputs via sketch, as opposed to prompt (it has prompts, I think it just weighs the sketch heavier). Alternatively, I think you might be looking at MJ, but then needing to bounce it to Leonardo for in and out painting.
Are you just looking for items/objects? Like, you are building assets for a game?
@@TheoreticallyMedia Not game assets, I create digital stencils that I use in lightpainting techniques to merge them into real world scenes. So for example a unicorn gets placed with camera tricks in a forest. But it is a real photo in the end, no digital manipulation at all.
And for that I want to generate objects like magic crystals, ghosts, mythical creatures, strange machines and all kind of other stuff. Bit specific since I need that black background and it needs to look photorealistic to work with that technique.
Who's the target market for this AI tool? Like who are the people that will need this more specifically?
Not sure I understand your question? I mean to my mind, anyone? Like if you're a novice artist or if you're a pro-- I can see a use for it. It might not be a specific cup of tea, depending on the style one is aiming for-- but I can see a use for it.
@@TheoreticallyMedia thanks for your response. I mean which business would have the need for use of the art transformation? Like authors etc?
wow , this is pretty impressive
It is a LOT of fun!
This is very cool! Do you know what material these models were trained on?
I do not. It looks (to me) to be based in Stable Diffusion, so...really, it could be anything.
And when you feed it cubist art ...? :)
oooooh, that sounds fun to try!
Great 😀 thanks
100%!!
you helped me man, thanks very much!
Interesting and interested. My son does several drawings daily. I've tried tracing manually in inkscape and then coloring but it's very time consuming and not the only thing I have to do in a day 😂. I have a huge backlog of drawings. I will take one of his raw scanned drawings and see what happens. I am working to get enough of his drawings digital so I can get him his own pod store. Heidi
Ha! My kids are the same-- Basically, I have 0 printer paper left, and I can never find a pen or pencil because they've all been stashed somewhere in the house. But WHERE?!!!!
@@TheoreticallyMedia i know where his latest drawings are as I have been storing them in clear plastic shoeboxes. Have 3 full now. He's been drawing since he could hold a crayon. He's quite good and better than I could have ever done. 😆 i save my paper by buying him sketch books. They are at walmart and about $6 each. It does help. Buy them their own and try to get them not to take yours. I remember once my mom sent us each a huge gift set of sharpie markers. I was sparingly using mine for art etcetera but he used all his up quickly. 72 markers. I would use mine maybe once a month or so. I went to go use them and found them used up. He had been taking mine when one of his colors ran out and putting the dried up one in it's place. We had a looong talk about that. 😆
Come on , you dont need to pay to make it. Use stable diffusion but change few parameters ... Google it
This paid apps all based on the same technology
True. But I think there is an audience for platforms. I think of it as PC Gaming vs Console. Not everyone has a beefy GPU, not the time/inclination to install and upkeep.
Thanks!
No, no! Thank YOU so much!! Happy new year to you!
Great content, subscribed!
Excellent to hear! Looking forward to seeing you!
I would like to know if what you upload to this site remains yours or do they take ownership without asking you of what you upload?
That said, IA is for those who would like to be what they will never be.
I don’t think that would be worth it to them. In all honesty, think about the amount of images that are uploaded to any of the various AI tools. I just don’t think it’s worth it to them in just storage space to hold onto all of it.
actually i have failed to generate one of my sketches because there is not good site to do so
"It doesn't act like a standard prompt generator", it's crazy we're already saying this 😅
3 months from now “It’s crazy that it doesn’t work like a standard brain wave prompt generator, but rather travels from the future to give you images you hadn’t thought of yet…”
Can this be used to sharpen a photo?
So, I tend to think it will alter the photo too much. Like, it will stylize it into a different theme. Which, might be interesting, but I don’t think that’s what you’re aiming for.
You might have luck with using something like Leonardo, using your blurry photo as an image prompt, generating an AI photo version, and then use InsightFace to face swap your subject in.
@@TheoreticallyMedia Thank you.
Are there any AI image generators or tools that can complete a partially drawn image? (e.g. generate feet, legs and hips to an image containing just a head and torso)
A few. You could use Midjourney and use your particularly drawn image as a reference-- but things wouldn't come out exactly the same. You could try Prome, and use a heavy text prompt along with it? I didn't try, but on one of my images (the dragon shirt girl) it did generate an arm and a hand-- so it is possible.
you must say from the beginning that major function need to paid. like hd download
I wouldn’t call that a “major function”
The major function is that it creates images from reference images. The HD Upscale is an upgrade:
I just tried it. It's horrible. I tried converting my troll sketch into... well anything and it couldn't interpret it any other way than a trope fantasy girl. I see why it's free.
What was your prompt? But yeah, I can see if there is a fantastical creature, it might not be in their library.
If you want, feel free to pop over to the discord and drop your sketch into the project help channel and I’ll see what I can do: discord.gg/yHzGgWtb
@@TheoreticallyMedia Thanks. I just try these things out to see if Ai finally can be good and make good convincing work, but it seems it's a looooooong way to go. Even in your examples. I've worked in visual design for 15 years now. You can spot AI from a mile away... for now at least.
I like the control it offers but I'm not convinced yet on the quality of the images to pay for it and it seems you can only assess that quality by upgrading to a paid plan. The thumbnails are too small and the low res outputs aren't that good. I don't want to be nickled and dimed to death on these new platforms while the technology is still developing. But thanks for the heads up! I'm gonna bookmark this and check in from time to time to see how it's developing.
Good route! I'm not sure there will ever be an (ahem...adopting a mighty Fantasy Wizard Voice): "AI Tool to RULE them All!" and to be honest, I don't think I'd want just one platform dominating. I actually see in my near future, a thing where I sub to MJ for a month, then sub to PromeAI for a month, then hop over to Leonardo or whatever comes down the pipeline.
It's almost like streaming services...
@@TheoreticallyMedia I just created a ai motion graphic novel with MJ Leonardo, D-ID Eleven Labs and Photoshop BETA then edited it on my cell phone. It came out pretty good. But I'm searching for the generative holy grail the consistent image. And of course video as I'm a storyteller by trade.
Yeah exactly same sentiment here. I like the UI and ease of use but it's really hard to validate what it's capable of when you don't have access to those capabilities and it's still way too early to be shelling out that much money for each and every one of these new AI popups.
So far I tested going back and forth between Promeo and MJ copy/pasting from each to the other and it's pretty fast. I loved the shape cohesion in it though...really good at keeping the composition of your drawing if you're just throwing out variation ideas.
@@jenkem4464 to be honest, I think that’s where the real strength of these AI tools are. Ping ponging back and forth, generate an image in one, use it as an image prompt in another.
You sort of end up with the strengths of all of them working for you…well, with a ton of experimentation at least!
You don't need much of a commission to pay for a months worth of this tool.
I have to admit this quite killed my intention to improve my 2D skills by practicing.
haha, I still think its worth improving! I think the better you can sketch, the better your result will end up being! My problem is, every time I have a good pencil or a good micron pen, my kids steal it!
👏👏👏👏👏
Thank you!
Dr. Plague? Kids of the covid era ...
I seriously wonder the same. I'm like...you do know what this is, right?
Apparently they're learning about the Middle Ages in school...so, yeah-- I got a lecture on it. From a 10 year old. 😃
i'm not sure if to consider this as a threat or a useful tool
I think more B. If you’re looking at client work. Well clients never actually know what they want, and can’t be bothered to learn stuff like this. And I think it’s less threat, considering you still have to create an image to input into it. So, the better your art skills, the better your output will be!
Keep on getting Generation Failed?? :(
I’ve seen this happen a few times. My guess is that there is some kind of update that got borked. Try again in a few hours, or maybe tomorrow (considering it is Sunday, depending on what part of the world you are in) and it’ll probably be working.
Bet it’s a bad night for someone on their tech team right now!
@@TheoreticallyMedia Thanks for the info. :)
This is so objectionable. I have been a henchman for 30 years, it is a full time job, now Ai is going to replace us. How will I support my family? Hench them?
Get ahead of the curve: Start selling child size Domino masks for all the other Henchkids that will soon need to work!
Make sure you get the Horizon DLC 😎
Burning Shores? Sadly, I'm still in the stone age with a PS4. So, I'll have to wait a bit. And to say, I didn't hate it, but it had a bit of that Ubisoft thing of too much to do.
Also, Lance Reddick, RIP!
Thank you. The video was well made and comprehensive. It’s amazing what can be done with AI.
Thank you! Glad you enjoyed it! Yeah, it's a golden age for sure!
I dont like the aromatherapy beads in the case ill tell u that
I want the ten free credits
Oh, sorry, I think they were all allotted. I’ll check in with them to see if I can get a few more!
AND TESTED IT , AND IT'S HORRIBLE, the results are 100% random, maybe with a lot of effort sometimes they give u a deform face and some weird body and absolutely different cloth, but that's very rare, usually if u put a man standing in a field it gives u a chinese woman lying in the ground in a mansion, the worst IA ever till now.0% control of the results
You think? What kind of images are you initializing with? I mean, I was doing a test for another person here with a Batman sketch, and consistently getting Batman.
Now, that said: that’s Batman. Who, along with a prompt containing the keyword of Batman is probably fairly easy for the AI.
Try giving it a shot with a loose sketch and a famous analog of what you are looking for, and see if you can build out from there. Might be something?
Hello Sir I have lots of comic character and cerachtor ideas but I need a platform and a chance to explore my unique art work in the 🌎
Not free. $39.00 a month
Yeah, for pro. But (currently) unlimited low res to try. And that’s something I don’t think any other service offers.
And depending on what you’re going for, the low-res stuff can be useable.
Mostly a temp service 🤣🤣🤣
Oh, it was such a fun comic. Like, at their HQ they had different henchmen outfits for every villain. There was going to be a gag in a future issue where the main character shows up in the wrong henchman outfit. He thinks he’s going to get fired for it, but when the Hero sees the costume, he thinks the other villain is in the area, which allows the client-villain to escape.
Our main character ends up getting a bonus instead of getting fired.
It was also a heartfelt story about trying to get your life together because he was going to be a dad. Still bummed the comic fizzled out, but maybe I’ll be able to finish it with AI?
I tried prome. It was a complete joke. Nothing rendered correctly and it was often incredibly bizarre.
I think a lot has to do with your initial image, and I did use the technique of reinputting the output back into the input a few times. Try the Realistic and Precise modes, I seemed to have the best luck with those two. Also, did you prompt? I wanted to spend more time exploring that, I think it prefers the Stable Diffusion style of longer prompts.
very cool
Thanks for checking it out!
I was interested in something like the photos you took and manipulated to be B&W only, suited for typical coloring, as with Photoshop or Clip Studio. But something beyond just high-contrast processing. Something that was actually rendered, the way a good inker would work. (I'm putting MYSELF out of work!) I haven't seen that anywhere, and it may be unattainable.
Ha, Klaus Janson is going to be so mad at us!! So, I got a kinda ok result from a Marc Silvestri Batman sketch with the prompt: "Black and white, comic book, ink, precise lines, batman, no color, white background" It still comes out with kind of an ink wash look, that kind of grayscale Tim Sale B/W look--
Shoot me an email at theoreticallymedia@gmail.com and I'll send you over the results!
$40.00 a month unlimited use😞
That’s just for the pro though. I’d say play with the low-res until you get results you like, then pop in for a month and run them all! Alternatively, you might be able to take the low res images and uprez them on your own.
FYI this is a paid endorsement which he only disclosed in the description, as well as the rendered images are not free as he suggests in the video. Very dishonest channel.
It’s marked a paid promotion and I do disclose it verbally in the video. Also, a lot of these images were generated from the free version. I indicate when I use the pro feature.
The website is GONE.
So it's like a diluted version of stable diffusion img2img + controlnet, but with lower resolution and long wait times. lol
It's not that bad, honestly, especially for a free service. A very good option for people that can't run SD locally.
You got it. I always bring up the analogy of PC vs Console Gaming. I got out of PC gaming years ago, particularly when I noticed that I was spending more time optimizing my system or installing mods than actually playing the stupid games.
They do offer higher res at the subscription level, which I can't really argue. It is a service-- and in some ways, it offsets the cost of upgrading a GPU, or in the case of someone just starting out: The cost of a PC build, plus the time to learn to do it.
AI-created crap is not art even a bonobo would be capable of it.
I would like to hang out with that bonobo.
yeah cool and all. we will all loose validation for art in general that way- thats for sure. Im tired already of this generic AI content everywhere. first its recognizable as such, looks generic and soulless and finally whats the point of doing it at all when everyone can do it with a few clicks? Once art was validated as something extraordinary done by skilled talented artists. Now Art is a waste product done by AI everyone can do. Worst: people think they can call themselves creatives with such tools. NOPE, you aren't.
promea ai is already gone thanks for shilling this vaporware please delete this video
If it’s done by AI it’s not art. Nuff said.
Eh, I know where you're coming from-- but I take exception with this tool. Mostly since it requires an initial image to start. The idea here is that the user draws/sketches to the best of their ability and the AI cleans it up. But here's the thing: As the user keeps working, they're actually practicing, they're (hopefully) getting better at sketching-- learning to get better at anatomy, perspective, etc.
In a weird kind of way, something like this almost acts as an instant art teacher. "Oh, you drew that? This is how I'd do it"-- and if the person is willing, they can analyze the output and see what they need to work on.
by the way: Nice art on your page!! And keep up that physical grind! I just got back to my 5am boot camp after a few days off from a pulled hammy-- feels great! We might disagree on AI art, but I think we can both agree on the importance of keeping in shape!
@@TheoreticallyMedia Tks man