Muchas gracias. Me alegra que te suscribas a mi canal. Haré más videos sobre eso. Mejoremos nuestra técnica juntos. Traduzco usando Google. Espero que exprese mi opinión. Que tenga un lindo día.
Thanks for your comment. I have not yet checked the function. Recently it update frequently. Always has some new functions added. It is really fascinating. I will try it out later. And share the information with you. Have a nice day.
I have one specific issue that's been bugging me for a very long time, when I generate casual images the results always come up with a few real life images which what I was aiming at and a few cartoon lookalike images, i mean they're not totally cartoon like but they showed that cartoon type drawings look which I hate. I want to completely remove/filter out the cartoon/drawing results how do I do that? any tips? I've tried a thousand ways to filter them out with different categories, methods or prompts but nothing seems to work, it only seems to work if I change to cinematic photo category but that creates whole new problems altogether it screw up the poses and lighting
Thanks for your comment. As you said, use the prompt like long black hair with red ends first. And then add more modifications. More and more. Even the new SDXL can give good images using simple keywords, more specific definitions can help get better images. Good luck. Have a nice day.
Thanks for your comment. I am sorry that I also find the latest version can not use DAAM now. It may be fixed later. Recently SD webui update a lot. Some of the extensions may work properly now. We can report the bug and hope they will fix it soon. Good luck, and have a nice day.
Thanks for your comment. I am glad that you subscribe to my channel. I will keep going. Also hope you will make an excellent video. Let's make progress together, have a nice day.
I always think I am not a native English speaker, it is a problem not so fast and fluent. I find you. I am a little relieved. I can slow down in the following video. Thanks for your comment. Have a nice day.
first, try using negative prompts such as bad anatomy, bad hands, missing fingers, extra fingers, etc. along with the weight such as (bad hands: 1.5) etc, If not work. Then use ControlNet, from img2img, paint the wrong area, and try. Finally, you will get it. Good luck.
Good guide except for the negative prompts. Most of them do not and can not work for various reasons and in fact actually limit the possible good outputs. For example "cropped" does nothing, it was a desperate attempt in the beginning to reduce the number of cropped images. But those came from bad training where right before the processing the image was cropped to be square. So the "AI" doesn't even know that the image was cropped in first place. What is "ugly" and what not is undefined and depends on context, it can therefore prevent pretty creations too. It's also safe to assume that no training occurred with images labeled as "bad anatomy", mentioning hands simply reduces the chance to generate images with visible hands and so on. My advice would be to get rid of all the clutter, start with simple prompts and add to them as necessary.
Thank you very much for your comment. Your opinion is correct. Most of the models were trained using improved samples. Now they are already quite good even without using negative prompts. As you said I also find when you use bad hands or bad anatomy the AI just cleverly does not use hands or covered hands with something. It decreases the creativity of ART. Really not use them only when something is wrong. Let's try out and share the experience. Have a nice day.
Ive just started using a1111 and have made a long neg prompt based on other images and yesterday i tried removing some of the prompts, but they all see to be working in a beneficial way, using a fixed seed and trying out a litle can be used to pin out whats usefull or not.
@@maitotechlab9035 Thanks for your comment. I am glad that you also try using WebUI automatic 1111. It is really nice. Try it yourself first, it is the way to learn. Have fun with it.
@@itzpaco5539 Really, I am not a native English speaker. In fact, I usually read the content displayed on the screen. It makes me like a ROBOT. Sorry for that. But you know improving my English is not so easy. I will try harder. LOL. Have a nice day.
Extremely well put together, with some great resources.
Legitimately helpful.
Thanks!
Thanks for your comment. I am glad you like my content. I will continue. Have a nice day.
Loved it. Thank you so much as we need as many resources as we can get without all the flashy stuff just to the point. :)
Thanks for your comment. Have a nice day.
Finally, someone making something useful. This is an extremely helpful resource. Thank you for providing other resources too!
Thanks for your encouragement. I will keep going. Have a nice day.
Muchísimas gracias!!! He estado buscando un vídeo como este por tanto tiempo! He aprendido bastante! Me suscribirse a tu canal! Gracias!!!
Muchas gracias. Me alegra que te suscribas a mi canal. Haré más videos sobre eso. Mejoremos nuestra técnica juntos. Traduzco usando Google. Espero que exprese mi opinión. Que tenga un lindo día.
Excellent video. A lot of very good resources and 2 extensions revealed that I didn't know about.
Thanks for your comment. I am glad that help you a little. Good luck. Have a nice day.
Excellent video for beginners. Thank you!
Thanks for your comment. I just start learning the software. It is really fun. Let's make progress together.
thank you for this informative sharing👍
Thanks for your comment. Have a nice day.
That was a great video thank you. It's given me a lot to look through! Subscribed. Have a great day :)
Thanks for your subscribing. I am glad it helps you a little. I will keep going. Have a nice day.
This was the best prompt guide video which i ever seen,❤ keep doing you are the best
Thanks for your comment. I am glad that you like it. Have a nice day.
A lot of you tube channels publish same content, but I prefer to watch those made for people who seems nice, which is the case of this video.
Thanks.
Great tutorial for those of us just starting with SD! Keep bringing the videos :))
I am glad you like it. Please share with your friends. Thanks for your comment. Have a nice day.
This was exactly what I was looking for! Thank you so much.
I am glad I can help a little. Have a nice day.
I loved the fact-packed delivery of this very informative guide. Exactly what I needed to really improve my skills!
Thanks for your comment. Have a nice day.
Very good! Thanks e please continue.
Sure, I will go on. Thanks for your comment. Good day.
Fantastic video my friend! Thank you for sharing your knowledge 🙏
Thanks for your comment. I am glad to share my experience. Have a nice day.
ARTIST FROM EGYPT HERE ,i like what u do man
Thanks you very much. Have a nice day.
this is amazing, thank you and please make tutorials too
Thanks for your comment. I will make the video soon. Please keep an eye on it. Have a nice day.
thank you sir great video, explaining the prompts very good
have you try the text to video in SD ? i played with it with and got nice result
Thanks for your comment. I have not yet checked the function. Recently it update frequently. Always has some new functions added. It is really fascinating. I will try it out later. And share the information with you. Have a nice day.
Nice video, thanks for making.
Thanks for your comment. Have a nice day.
Awesome video! Thanks alot!!
Thanks for your comment. Have a nice day.
Thank you for sharing your knowledge
You are welcome! Have a nice day
I have one specific issue that's been bugging me for a very long time, when I generate casual images the results always come up with a few real life images which what I was aiming at and a few cartoon lookalike images, i mean they're not totally cartoon like but they showed that cartoon type drawings look which I hate. I want to completely remove/filter out the cartoon/drawing results how do I do that? any tips?
I've tried a thousand ways to filter them out with different categories, methods or prompts but nothing seems to work, it only seems to work if I change to cinematic photo category but that creates whole new problems altogether it screw up the poses and lighting
The prompts can change a lot of definitions. Please try using different prompts and using negative to avoid the Cartoon image. Good luck.
You're doing God's work. 😄ありがとうございます！
ご褒美ありがとうございます。ちょっとお力になって嬉しい。これからもよろしくお願いします。
Thank you ❤
Thanks. Have a nice day.
i'm just learning and think i'm trying to be too complicated. what kind of prompt would I put in for long black hair with red ends?
Thanks for your comment. As you said, use the prompt like long black hair with red ends first. And then add more modifications. More and more. Even the new SDXL can give good images using simple keywords, more specific definitions can help get better images. Good luck. Have a nice day.
Great video! Unfortunately I couldn't get DAAM to work.
Thanks for your comment. I am sorry that I also find the latest version can not use DAAM now. It may be fixed later. Recently SD webui update a lot. Some of the extensions may work properly now. We can report the bug and hope they will fix it soon. Good luck, and have a nice day.
Hey man, cool guide :). I'm creating content around the same topic too. Subscribed and looking to following your journey
Thanks for your comment. I am glad that you subscribe to my channel. I will keep going. Also hope you will make an excellent video. Let's make progress together, have a nice day.
Thanks for the video. A little fast, but still good stuff
I always think I am not a native English speaker, it is a problem not so fast and fluent. I find you. I am a little relieved. I can slow down in the following video. Thanks for your comment. Have a nice day.
Thanks 🤩
Thanks for your comment. have a nice day.
any tips on fixing hands?
first, try using negative prompts such as bad anatomy, bad hands, missing fingers, extra fingers, etc. along with the weight such as (bad hands: 1.5) etc, If not work. Then use ControlNet, from img2img, paint the wrong area, and try. Finally, you will get it. Good luck.
Good guide except for the negative prompts. Most of them do not and can not work for various reasons and in fact actually limit the possible good outputs. For example "cropped" does nothing, it was a desperate attempt in the beginning to reduce the number of cropped images. But those came from bad training where right before the processing the image was cropped to be square. So the "AI" doesn't even know that the image was cropped in first place. What is "ugly" and what not is undefined and depends on context, it can therefore prevent pretty creations too. It's also safe to assume that no training occurred with images labeled as "bad anatomy", mentioning hands simply reduces the chance to generate images with visible hands and so on. My advice would be to get rid of all the clutter, start with simple prompts and add to them as necessary.
Thank you very much for your comment. Your opinion is correct. Most of the models were trained using improved samples. Now they are already quite good even without using negative prompts. As you said I also find when you use bad hands or bad anatomy the AI just cleverly does not use hands or covered hands with something. It decreases the creativity of ART. Really not use them only when something is wrong. Let's try out and share the experience. Have a nice day.
Ive just started using a1111 and have made a long neg prompt based on other images and yesterday i tried removing some of the prompts, but they all see to be working in a beneficial way, using a fixed seed and trying out a litle can be used to pin out whats usefull or not.
@@maitotechlab9035 Thanks for your comment. I am glad that you also try using WebUI automatic 1111. It is really nice. Try it yourself first, it is the way to learn. Have fun with it.
Thank you cutie😍
Thanks for your comment. Have a nice day.
ai human??
Not AI human, just generate a beautiful photo I think. Have a nice day.
I was thinking that you is ai human, the way you are talking LOL
@@itzpaco5539 Really, I am not a native English speaker. In fact, I usually read the content displayed on the screen. It makes me like a ROBOT. Sorry for that. But you know improving my English is not so easy. I will try harder. LOL. Have a nice day.
@@TubeUnderdevelopedyour English is very good sir, keep doing❤