Yeah, that would be even better, however Flux is good enough and will make these adjustments for you if you include it in the prompt. See how I had bamboo hats on the boy in every image and it still managed to give me images without the bamboo hat. The model is great ❤️☃️
May I ask for the rooster has like arm handle the chess, is it generated after training or manually editing by your end? Just concerns whether AI can generate the action with handling action like that for the trained character or not
@@nvckenBy variations I mean generating images a few times using the same prompt. However, it's always great to learn Photoshop for better control and making small adjustments. ❤☃️
Hi brother, thank you for the kind words! I love your content. I recommend anyone reading this comment to check out taoprompts, he has really useful guides/tips.
Hi, thank you for your video, this is awesome! Just to clarify, did you create the 3 original characters all in Midjourney, or did you illustrate one of each of them yourself? Thanks!
Yeah, use the "--cref", character reference feature and include your 2 images. You need to generate a bunch of images and pick the ones that resemble your character accurately. Once you have 10-15 images go ahead and train a model for that character. Now you can generate infinite consistent images. I hope this solves your problem, otherwise send me an email and I will help you further ❤️⛄️
Hi, thanks for sharing this, I just subscribed to this channel and I have a question: Is it possible to generate all three characters together in one frame/image using generative AI, or does it require editing in Photoshop? I’m looking for a solution to generate multiple consistent characters (not just one) with generative AI. I’d appreciate your thoughts on this. Thank you!
Great question. There are possibilities to achieve this with inpainting, however if you're looking for a single prompt and getting the perfect image it's impossible currently. The closest you can get is with Stable Diffusion, it takes time and it's usually not as great. If you put your time into editing the images yourself you will have total control and it doesn't take as much time as it used to thanks to AI and model training etc. ⛄️❤️
perfect tutorial but what if i have my characters image im already created how i can create sheet and train model for it in Open art cuz i try give image and create sheet for it to start create my train model for it but fail which better way i can do this ?
I would suggest you upload your image to Midjourney and use it as character reference. Generate a bunch of images until you get 8-12 decent images and train the Flux model from there following the guide on this video. If you still can't solve it, send me an email: Ravjar.said@gmail.com Good luck! ❤️☃️
As someone who has worked with Stable Diffusion to create comic characters, I’d like to share a few thoughts. Creating your own comic isn’t hard; it’s actually quite easy. But to make it commercially viable, an artist is always essential. Current AI-generated characters simply don’t convey emotion in a way that resonates with readers-let’s be honest about that. AI can only help speed up certain tasks, like coloring. Even at that, it can’t outperform an experienced artist, and that’s the reality. So keep exploring, but don’t expect anything miraculous from AI; let it be an assistant, and take the time to learn to draw. Trust me, a character hand-drawn by a person, even if imperfect, carries more emotion than an AI-generated one. We humans can tell the difference-it's the wonder of having a soul.
I agree with you to some extent. I try not to compare things too much; I prefer to create and have fun. I would choose something hand-drawn every day as well. However, there are certain styles that artists can no longer replicate, and with these tools, you can satisfy your imagination in just a few seconds. Another example is that many people contact me to help them create stories or illustrations in specific styles that would typically require paying 20 times that amount for a real artist. In some cases, that artist isn't even alive anymore. I once worked with a mother who wanted to create a story in memory of her son who had passed away, using the style of her favorite illustrator, who is no longer living. It was a beautiful process, and if that artist had been alive to do the illustrations, it would have undoubtedly been ten times better than what we created. I try not to let ego get in the way; I just want to have fun and create things. At the end of the day, a professional artist who knows how to use AI and the most cutting-edge tools is better than the same professional artist who does not.
@@chucky_genz Actually it's true. Everything can be copied on the internet. Whatever is created via mind & external tools will always be a copy. Only the soul is capable of creating something new something that never existed before. No wonder why Art masterpieces are almost priceless or sold in millions Because they contain things that AI can never put in its creation. It's the emotions of the soul.
DUDE ! You did all the work well and quickly in Photoshop! What do you use artificial intelligence for? I think artificial intelligence is currently a useless complication. High volume, low speed, high complexity and time are among its problems.
Thank you. It's fun to use AI and you can create some things very quickly and satisfy your imagination from time to time. AI is advancing pretty fast, it might be a great opportunity to get comfortable with it's capabillities. We'll never know what the future will look like. ❤️⛄️
Yeah, that's a really fun thing to do. I suggest you pick "face" instead of "character" when training the model. You can play around with a ton of different styles and haircuts that way. It allows you to experiment much more in my opinion.
A few tools in this video requires a fee however there are free alternatives to many of the tools. I just try to use the best for the sake of the video. Openart has a free option and you can pay for more usage, photopea.com is free, Midjourney costs etc.
☃❤ Hi, don't forget to subscribe! 95% of my viewers haven't subscribed yet.❤☃
You covered everything I needed to know in such a clear way! This video is a fantastic resource!
Thank you. That's great to hear. ❤️⛄️
Hey Snowball AI, Thanks so much for this and i will take your Course ! Your Photoshop and AI integration is so good !
My pleasure! I am sure you'll find it useful. Don't forget to share your progress with me on Discord if you find the time for it.
That's some impressive consistency, man! Seemingly works great with illustrations, I wonder how soon we'll reach the animation stage with ai
Thanks, Flux is a great model and a little editing goes a long way. Animations will be very fun! ❤⛄️
Love this video, I was captivated looking at you photoshop the images!
Now I want to visit Vietnam! 🇻🇳
Thank you for taking your time and watching the video.
It's amazing what you do combining AI, traditional tools, and imagination!
Thank you Excelerator, we live in amazing times ⛄️❤️
Fantastic tutorial. Thank you
Thank you for the kind comment! ❤️☃️
Great Video!
Does it makes sense to train the model with images of the Character with many different outfits?
Yeah, that would be even better, however Flux is good enough and will make these adjustments for you if you include it in the prompt. See how I had bamboo hats on the boy in every image and it still managed to give me images without the bamboo hat. The model is great ❤️☃️
May I ask for the rooster has like arm handle the chess, is it generated after training or manually editing by your end? Just concerns whether AI can generate the action with handling action like that for the trained character or not
It is generated by Flux entirely. I generated 2 variations and the one on the video is one of them. ❤️☃️
@@SnowballAI sorry what kind of variations you mean? Did you need draw the arms for the rooster to train to model at first?
@@nvckenBy variations I mean generating images a few times using the same prompt. However, it's always great to learn Photoshop for better control and making small adjustments. ❤☃️
great guide man!
Hi brother, thank you for the kind words! I love your content. I recommend anyone reading this comment to check out taoprompts, he has really useful guides/tips.
Great tutorial, I am a follower of Leonardo AI, do you think the steps can be applied?
I haven't tried it out on Leonardo AI, might create a video on that. It all comes down to the AI model behind it. ❤️☃️
Hi, thank you for your video, this is awesome! Just to clarify, did you create the 3 original characters all in Midjourney, or did you illustrate one of each of them yourself? Thanks!
Thank you! Midjourney did all of the characters, I had a sketch of the boy in a scene but the real process started in Midjourney. ❤️☃️
Is there a possibility to generate images of a character if i only have 1 or 2 images of a char? Would really appreciate it. Thx a lot for your video.
Yeah, use the "--cref", character reference feature and include your 2 images. You need to generate a bunch of images and pick the ones that resemble your character accurately. Once you have 10-15 images go ahead and train a model for that character. Now you can generate infinite consistent images.
I hope this solves your problem, otherwise send me an email and I will help you further ❤️⛄️
which would you recommend for better results between flux and hailuo?
Flux is incredible for training a model without much effort. ❤☃️
@@SnowballAI woww thanks for replying...I will use flux and let yu know
@@SnowballAI New subscriber here!!!!
Great vid mate
Glad you enjoyed it! Appreciate you.
Excellent work! I imagine Affinity Photo will work just as well as Photoshop.
I am sure it's as great, atleast for the steps required here. ☃️❤️
Hi, thanks for sharing this, I just subscribed to this channel and I have a question:
Is it possible to generate all three characters together in one frame/image using generative AI, or does it require editing in Photoshop? I’m looking for a solution to generate multiple consistent characters (not just one) with generative AI. I’d appreciate your thoughts on this. Thank you!
Great question. There are possibilities to achieve this with inpainting, however if you're looking for a single prompt and getting the perfect image it's impossible currently. The closest you can get is with Stable Diffusion, it takes time and it's usually not as great. If you put your time into editing the images yourself you will have total control and it doesn't take as much time as it used to thanks to AI and model training etc. ⛄️❤️
@@SnowballAI Thanks for your kind response
@@SnowballAI Thanks for your kind reply, looking for more of your videos...
Awesome, keep it up
Thank you for watching ☃️❤️
perfect tutorial but what if i have my characters image im already created how i can create sheet and train model for it in Open art cuz i try give image and create sheet for it to start create my train model for it but fail which better way i can do this ?
I would suggest you upload your image to Midjourney and use it as character reference. Generate a bunch of images until you get 8-12 decent images and train the Flux model from there following the guide on this video. If you still can't solve it, send me an email: Ravjar.said@gmail.com
Good luck! ❤️☃️
As someone who has worked with Stable Diffusion to create comic characters, I’d like to share a few thoughts. Creating your own comic isn’t hard; it’s actually quite easy. But to make it commercially viable, an artist is always essential. Current AI-generated characters simply don’t convey emotion in a way that resonates with readers-let’s be honest about that. AI can only help speed up certain tasks, like coloring. Even at that, it can’t outperform an experienced artist, and that’s the reality. So keep exploring, but don’t expect anything miraculous from AI; let it be an assistant, and take the time to learn to draw. Trust me, a character hand-drawn by a person, even if imperfect, carries more emotion than an AI-generated one. We humans can tell the difference-it's the wonder of having a soul.
I agree with you to some extent. I try not to compare things too much; I prefer to create and have fun. I would choose something hand-drawn every day as well. However, there are certain styles that artists can no longer replicate, and with these tools, you can satisfy your imagination in just a few seconds.
Another example is that many people contact me to help them create stories or illustrations in specific styles that would typically require paying 20 times that amount for a real artist. In some cases, that artist isn't even alive anymore. I once worked with a mother who wanted to create a story in memory of her son who had passed away, using the style of her favorite illustrator, who is no longer living. It was a beautiful process, and if that artist had been alive to do the illustrations, it would have undoubtedly been ten times better than what we created.
I try not to let ego get in the way; I just want to have fun and create things. At the end of the day, a professional artist who knows how to use AI and the most cutting-edge tools is better than the same professional artist who does not.
Be patient bro, next year you will delete your comment 😂
@@SnowballAI 100
@@chucky_genz Actually it's true. Everything can be copied on the internet. Whatever is created via mind & external tools will always be a copy. Only the soul is capable of creating something new something that never existed before. No wonder why Art masterpieces are almost priceless or sold in millions Because they contain things that AI can never put in its creation. It's the emotions of the soul.
@@chucky_genzto be honest, no
DUDE ! You did all the work well and quickly in Photoshop! What do you use artificial intelligence for? I think artificial intelligence is currently a useless complication. High volume, low speed, high complexity and time are among its problems.
Thank you. It's fun to use AI and you can create some things very quickly and satisfy your imagination from time to time. AI is advancing pretty fast, it might be a great opportunity to get comfortable with it's capabillities. We'll never know what the future will look like. ❤️⛄️
Спасибо за видео!
Thank you my friend! ❤️⛄️
Is it possible to use my own photo and make is consistent?
Yeah, that's a really fun thing to do. I suggest you pick "face" instead of "character" when training the model. You can play around with a ton of different styles and haircuts that way. It allows you to experiment much more in my opinion.
Excellent
Thank you ❤️⛄️
How about animations?
Not quite there yet, with enough generations maybe, but not to count on yet.
❣
A.I. is the apotheosis of modern materialism. It is a being without a soul. Purely package, no content.
is that free
A few tools in this video requires a fee however there are free alternatives to many of the tools. I just try to use the best for the sake of the video. Openart has a free option and you can pay for more usage, photopea.com is free, Midjourney costs etc.
Why does every one of these videos say it's easy when it really isn't. You had me confused within the first 5 minutes of your video:(
You joke like AI... lol!
Are there AI tools that can place multiple characters in the same scene, for example, a rooster and a puppy playing chess together?
Love this video, I was captivated looking at you photoshop the images!
Now I want to visit Vietnam! 🇻🇳
Amazing
Thank you! ❤️⛄️