Leonardo AI Tips & Tricks | Prompt Structure
HTML-код
- Опубликовано: 26 май 2024
- We hope you enjoy this tutorial on basic prompt structure for generating images on Leonardo.Ai
Don't forget our Resource Center, Discord and Facebook group are a great place for inspiration and support.
🔗 Resource Center: intercom.help/leonardo-ai
💬 Discord: / discord
🔗 Website: leonardo.ai
🐦 Twitter: / leonardoai_
🔗 Facebook Group: / leonardoaigroup
🙏 Special thanks to Wade from Creator Impact for contributing to this tutorial.
🔗 Leonardo.Ai playlist by Creator Impact: • Leonardo.AI - AI Art 
Развлечения
There should be something it creats all characters same looking in all scenes.
Yes, give it a name ( his name is Gary )
How do I give it a name?
I was trying to create images of Paul and Silas in one of my Bible stories.
Your AI generated different images of Paul and Silas in each scene. Which were different from the ones the AI generated in the first scene. These makes my work looks poor.
love it! Quality info presented in a very digestible way
Thank you, that was very informative
Thank you, great explanation . Very good.
Nicely explained, thank you so much
Brilliant and helpful - thank you!!!
wow, very good. really appreciate this
Hi im newbie here..thank you for a very clear explanation😊
Excellent tutorial.
this was useful, thank you!
THANK YOU SO MUCH...OUTSTANDING!!!
Nice tutorial❤
Thank you very much for your help. Beautiful voice.❤❤❤❤❤
Wow very helpful
You're are the best grandma
There is so great get this app to learn how to do the AI Image , so nice
Thanks for your sharing ❤
Thank you.
Thank you
Thankyou
Go go team Leo 👍
I love it .
Knowledge definitely increase your will power! I certainly am now aware of a key coponent to AI generation ~ communication!
Head on over to the Discord server. We have a great community there.
Id love to make comics with Leonardo ai. or book covers, I hope that feature will be added.
It kind of is a feature already. You can get great results with just prompts and the correct models. You can even get text in your images if you don't mind pre preparing your text and using Controlnet with either Edge to image or Text to image.
My tips: create the elements of your comics separately: scenarios without characters, for example. Then compose the scenes in a software like gimp or whatever. And finally use image to image feature to get the final or intermediate results.
😍🥰🤩 unbelievable
This is really helpful as a total and complete newbie, thank you. How do you go about preventing unwanted changes in what's generated for eg in this grumpy guy one, how would you prompt/select settings if you always wanted to see all of the man from his feet to the top of his head (I note it changed when you used a different model)? And could you retain the pose from a particular generation and use adjusted prompting to retain the 80% of the image that's perfect for you, and just make changes to specific other elements?
Thank you. Getting the results you are asking for with just prompting would be very hard even if you re-used the seed and settings. If I had an image that was 80% what I wanted I would take it into the Canvas Editor on the app and make the changes there.
Thank you!
Is there a tutorial for that canvas editor?
Bcause I think I am looking for the same thing.
I would like to wrap my design on different objects.
Like a flat tattoo design wrapped on an arm or leg. And vice versa
For example in this video when you deleted the beard, also the old man changed. I would like to keep exactly that same old man but without the beard
A big requirement that lots of people have is that "IS IT POSSIBLE TO GENERATE A CHARACTER AND USE IT TO GENERATE OTHER IMAGES IN THE FUTURE"? for example, if I wanted to have images of the same man in this video, while driving a truck in desert, and also the same man walking with his dog in the city. is it possible in Leonardo AI?
Yes, you can train a model for a face. Its not 100% but you can get great results.
How is it possible to create it with the same character? Is there a video which explains how to proceed?
Thank you! Is there a way to give more weight to some words or phrases?
If you use Alchemy then you can add weight with () the more brackets the more weight or you can use a number as well ie (red dress 1.5). If not using alchemy you need to put the words that you most want near to the front of the prompt or double up on words.
I can't seem to have long shots or very distance shots of a specific prompt/ character...its always just a medium shot or extreme closeup.
You can try creating the scene you want and taking that into the canvas editor to add your character to it. If you want to do it with prompt alone the best place to get advise is over on the Discord server.
Same issue for me.
Thank you for this, but very confused as to why my left sidebar displays all of the AI Tools and Advanced Tools starting from Image Generation, Motion, Realtime Canvas, etc, then down to Finetuned Models, Training and Datasets, Texture Generation, and so on all the way to the bottom with FAQs and Help.
BUT...., when I want to generate an image, all of that disappears and I can't access those features any longer. It's blocked off with a box saying "Image Inputs have Moved to the Image Guidance Tab - Show Me!
Can't I access any of the former left side bar features/tools when I'm generating an image instead of those tools disappearing? Thanks so much for your kind help!
Image inputs have moved now, look next to generation history on the generation page for image guidance
Thank you kindly for taking the time to reply, but frustratingly that Image Input floating box is still there blocking all of the left side bar tools, but will figure it out eventually. Thanks anyway for your time.@@leonardo.ai_community
This is a good tutorial, but actually the systems gets confused when it's prompted/required to create a picture with 2 or more persons in it. Is there any guidance on this case?
Head over to the Discord for a great community who can help with all sorts of tips and tricks. My advise would be to separate your subjects with a "," rather than using "and", its not a catch all but can help.
I really love Leonardo's new feature image to motion but I am having a problem with my characters eyes. The motion on the eyes are off and look very bad. I have tried everything and I can't seem to fix it. Any suggestions?
There are many updates coming for motion in the future but for now there are no fix's for your issue, sorry.
@@leonardo.ai_community is there any way you can direct where the motion should be or does Ai choose it?
This was very helpful! Any tips for avoiding female images revealing too much cleavage? I have used modest or professional attire in the prompt, but it hasn’t worked. Any word suggestions for the prompt or negative prompt that I could use?
Thank you! You can try describing the clothes you want her to wear. For example... "Beautiful woman wearing a sweater". The best place to get great tips and tricks is over on the Discord server though.
@@leonardo.ai_community thanks!
Имам чувството, че Леонардо по подразбиране създава женски образ с деколте, голи рамене или гол гръб.
Налага ми се, например, да пиша за дрехата, че е с ръкави, или да пиша негативен промпт, примерно "голи рамене". И не винаги се получава желан резултат 😄
Is it possible to add the created character in different scenes? I tried several times but it generated a different character image each time
Yes, head over to the Discord for a guide
Where is the discord?
Leonardo teria como ter outros idiomas para executar o comando? sou do Brasil e tenho muita dificuldade, para entrar com os comandos na criação da imagem
Outras línguas funcionam às vezes, mas o inglês é melhor
@@leonardo.ai_community vou usar o Tradutor, consegui resultados impressionantes, com alguns textos tudo muda e fica mais real ainda, seu site é incrível, seu site é o Futuro
Is there a reason why no artificial imaging process can create regular polygons? for example an enneagram or ninegon!
That could be limitation of Stable Diffusion and how it was trained. You may be able to train a model to do this though.
Greetings, community, thank you very much for providing us with this majestic application. Can someone tell me why when I publish a video on RUclips with images of Leonardo, it is deleted along with the audios? Be careful, I have the free plan with the free bonuses, so it's the same with the Facebook page, does anyone know why this happens? Please.🙏🙏🙏
Hello and thank you for commenting. You own the rights for images you create using the Leo webapp, I don't know why your RUclips videos are being deleted. The Facebook Group needs you to post at least 10 characters with your image or it removes your image by default.
@@leonardo.ai_community Do you think that's because you have to pay a subscription? I thought it was free, and of course I already mentioned it in the WhatsApp group
How to write prompt for multiple characters, for example if i want 3 characters in image so how i can write because ai tool is not creating 3 characters in image
Head over to the image help channel on the Discord server to get the best advise for your images
A handsome grumpy old man hahahah
What is the difference between model, style and element. I think those thinks make Leonardo a bit confusing to use.
Hi, Model's are Fined tuned to get certain results. Style helps to guide those model's a little further. Elements are like a super finetuned model to add on top if you choose to. I hope this very basic description helps. There is a great community over on Discord that could help clear anything up much better than I can here.
Can you export these projects as layered files?
Not at the moment but there are updates all the time so it could be a future feature. If you head over to Discord we have a suggestions channel that the team check regularly.
I'm practicing making cartoon characters, creating human-like shapes. Can you give me some command suggestions? Thank you very much
Anthropomorphic and Humanoid are good key words to get human like shapes.
I can't seem to find Lesson #1. Link?
ruclips.net/video/66apNN0C4dI/видео.html Hope this helps
How do I keep the same character when generating different scenes and moments?
Head over to Discord for some tips and guides on achieving this.
@@leonardo.ai_community Yes, I did. Thanks! Will you be able to compete against midjourney v6?
Sem opção de legenda para português é impossível me ajudar!! Sinto por isso....
Desculpe por isso. Você pode acessar o Discord e obter ajuda da comunidade portuguesa. Usei um programa de tradução para este comentário, desculpe se não for bom.
I have a free account and I started creating realistic art but now it’s only letting me do cartoon style art unless I start paying for upgrade.
I thought the advert said you can use everything free using the amount of tokens you get each day. I’m confused what I can and can’t do
No sorry, you get a free trial of all the features for the first week. After that you need to be on any of the paid plans to open up all the features.
How can you create many prompts with the same character? So that you have the same man in different spaces? Is there a tutorial?
The best method at the moment is to train a model. Consistent characters are not easy to get yet though. If you head over to Discord we have a large community that may be able to help more.
1080 Image to Video please
HOW TO GET SAME FACE,how we get similar results in Leonardo Ai
I'll ask about getting a guide made for the channel. You can head over to Discord for a few guides until then.
i think if you give your character name like (old man Sam ) Try to use names that are not related to feelings, places, foods, etc
I tried this. I got rubbish in the outcome. How can you stuff up a eucalyptus tree with a hole in the trunk and a piece of paper in the hole. I got everything but that prompt. Incredibly frustrating.
Head over to the Discord server, its a great place to get tips on how to get the results you want.
Ние ли обучаваме Леонардо да генерира образи или той нас тренира, започвам да се запитвам... 🤔🙃
Leonardo and other image ai generators seem to work well with single persons. But it is so difficult to prompt it to do 2 persons. And forget about ethnicity
Yes this can be an issue, head over to the Discord for some great tips and tricks
I am from india
Leider kann ich kein Englisch. Wäre ein tolles Programm !
I think I forgot to confirm that I am over 18 years old when setting up my free account using the app version !
How can I update my profile and confirm I’m over 18?
Its in the ToS that you have to be over 18 so no need to have confirmed that. You are good to go.
@@leonardo.ai_community thanks for the reply.
It’s just when I generated a woman in a tight dress the picture was censored.
I couldn’t see it unless I clicked on it
أكتب في وصف الصورة ولكن يطلع لي بالصورة شئ لا أريده حاولت عدة مرات بدون جدوى
إذا توجهت إلى Discord، فهناك الكثير من الأشخاص الذين يمكنهم مساعدتك. يمكنك استخدام المطالبات السلبية للمساعدة في إزالة الأشياء التي لا تريدها في المطالبات الخاصة بك
@@leonardo.ai_community أنا أكتب في مكان التوجيه السلبي ما لا ريده في الصورة ولكن نفس النتيجة ولكن الجيل الفوري كان يعطيني الصورة كما أردتها ولكن توقف لم يرد إعطائي ما أريد مرة أخرى توقفت خدمته
@@leonardo.ai_community أكتب امرأة يتبين معي رجل جننت بالفعل
Kinda like being a film director writing a very precise brief
Por favor né eu não falo inglês
Desculpe, este vídeo está apenas em inglês
When I create an image, the image is not the same as when I write about clothes, the poor man is wearing old clothes, so the image is created.
Why does the man in the image come in a pant coat?
Head over to Discord for some great tips and tricks and a community that can really help by sharing images.
Comprei uma coisa enganosa qero meu dinheiro de volta
Você pode usar o botão de intercomunicação no aplicativo para entrar em contato com o suporte
How can i fix the issue with fingers. It will show 3 fingers(looks alien0 or the fingers look warped
Using the Canvas Editor feature id the best option for fixing hands or any issues with images.
Nicely explained, thank you so much
No, thank you for viewing