found this video while trying to figure out how to get nightcafe to generate images approaching what i want. trying to describe a scenario with two people in different clothes and with different expressions is my ultimate goal du jour , but this is a great intro to the basics of what NOT to include
Oh Daniel, now you are going to get me started in another hobby. I love art, but can't really draw myself so this looks very interesting! I hope you and your family are doing well.
@@Cottage-Reader Oh.. Thank you. Yes I just made one, it will be published on Aug 15th, It goes over comparing SDXL vs. all other models, using same seed and prompts. I also go over many of the prompts that I find useful for different types of artwork. You can see my artwork at: creator.nightcafe.studio/u/ArtDreamer
I was slow to use negative prompts because I didn't know about them and when I learned about them I didn't know how to implement them correctly. So I was making a lot of bad images even though my prompts were constructed well. When I finally learned how to implement negative prompts, the extra limbs, cockeyed facial expressions, and out-of-frame images practically ceased. The overall image quality also increased. I wish I'd known about your suggestions months ago. Thanks, you've even helped me with a few extra prompt ideas!
@@selmateacher7 Thank youI subscribed 😇... I have actually spent the past 4 hours trying to get an angel holding an antique brass scale. I have tried every combination of words in the prompt and negative prompt I can think of. The style is fine, but I just cannot get the engine to come up with a an antique set of scales. And weird things will happen, like the wings are set in the lower back, or there will be an extra wing floating about. 😣Thank you so much if you have time to give any advice.
@@TatyanaValdaBelindaHill You can use the Mask tool. I used it today. I created a woman holding a vacuum cleaner, but one of her eyes were just wrong. So I masked her face, used another seed and bang! Perfect face, with perfect vacuum cleaner.
@@selmateacher7 Ohhh! I was thinking a feature like that would be just perfect. I am new to AI, so I didn't notice the functionality was already there. Will save me a ton of photoshopping and time/money with zillions of retries and credits. (I spent another 7 hours on various other images last night and had a huge problem with deformed hands - even with all the negative prompts added and refining positive prompts.) Thank you so much again. Just about to open up Night Café now. 😘
ZBrush was originally released in 1999. It is still used to create 3D models for film today. It's interface and modelling style feels familiar to sculptor's and 2D artists. It excels at creating organic and highly detailed models. Nightcafe just released a new SDXL model for stable diffusion that might be possible to create a greater variety of art in higher detail.
Beautiful video, very informative. If I may ask: with the new sdxl they have suggested to reduce the amount of prompts as much as possible (despite seeing some quite spectacular results with huge lists of them) , does anybody find a short amount of them more beneficial than a long and detailed list? Last but not least, what’s the weight you guys set on the prompts: I’ve tried both with positive and negative from their minimum to to their maximum extend (+/-3) and I cannot see any tangible difference, any feedback on this?
Thank you so much for your very kind words. I believe I show an answer to your questions on multiple prompts on this video: ruclips.net/video/nrPIYRP22VM/видео.html I still use multiple prompts on some of my work within SDXL and it turns out.. great. About weights..... it does make a big difference when you are using many prompts. I show in my videos that prompts with the higher weight number will have a greater influence on the overall creation, but lesser prompts are still represented. Also, the seed has some influence with which prompts are seen to a greater degree in the final output. I encourage you to see all of my recent videos on this topic as I try to make it enjoyable.
That's amazing! I'm trying to learn about AI so I can write about it convincingly (as a fiction author), but it's still intimidating. However, you've given me hope by using prompts to get Indiana Jones from Bono and extra-limbed Hunter S. Thompson. 😄 Thank you so much for sharing what you've learned so far.
Very interesting. Thanks so much. I'm a total beginner. Question: Can I use a photo of someone as s start image and tell AI to portray them as Indiana Jones or something? Or Do you have a video showing using real people as the start image?
You may use a start image, yes. I made an example of my own photo and opened prompt as seen here: creator.nightcafe.studio/creation/kBeEmrnOnkTg2qs6gYqH
wow! Yes you can and it works! I left my prompts open for you so you can see what I wrote, but it did take my face and put a hat, jacket, and beard on it. This is so cool!
@@SherrisWorld Is it a famous face? There are other apps that do copy a non-famous face, like AI magic. But even that does not do perfectly, as it takes hundreds of photos to train an AI to create a specific face . Just to let you know, I never even thought of that until you asked. I look alot like Indiana Jones, so it worked for my face, LOL. Also, I used my stage name, "selmateacher7" so that it would use that face and not my common name, as it uses the singer that is called "daniel berry"
there are many out there, this was one that I found interesting: speakperfect.co/?ref=futuretools.io another place to look is here: aitoolsdirectory.com/
@@Specialist1Aim Those are free tools. There are many AI tools out there. The one I found interesting is for audio, the second is a list of many different AI tools, audio is one of the categories.
@@selmateacher7 This is awesome because I need something to help speed up workflow, it can auto generate sound effects like door opening sound or wind noises ?
@@Specialist1Aim Thank you for the clarification on what type of audio you need generated. I found this AI tool to generate sound effects: myedit.online/en/audio-editor/ai-sound-effect-generator
Hello there. It can draw good eyes and hands, but it is hard for the AI to do it 100%. Here is my work and it shows perfect eyes and hands. My newest work is all "open prompts" so you may copy the prompts , if you wish. here is the link to my gallery: creator.nightcafe.studio/collection/dBFNhLdpG8Xvm9adj7Zx
here is one of my works, called the embrace. I have the prompts open for you and the eyes are done correctly . here is the link: creator.nightcafe.studio/creation/u6ZLP2CB3bM0qwBzvBEr
I am sorry to hear that. please try "advanced mode" then try different seed numbers. That happens to me sometimes, but I try to always use different seeds to get it just right. As seen here: creator.nightcafe.studio/creation/p3vBtPQYw1ASHSHWHSTW
@@AutumnPeir Hello there, I am sorry to hear that. Please try again, but use the settings , 1 image, short, and thumbnail to create without using credits. Also, try different seed numbers . I know it may seem hard, but you can find success. I use nighcafe every day and once you have found a good combination, I normally just evolve using the same prompts. first prompt: white mennonite cap, White cream, white lotion, white drops, white droplets, facial, white drips woman weight : 1 prompt 2 : ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, extra limbs, disfigured, deformed, body out of frame, blurry, bad anatomy, blurred, watermark, grainy, signature, cut off, draft weight: -1 seed: 444219900 modeL: Stable Diffusion v1.5 here is the link to what it made: creator.nightcafe.studio/creation/6QeOwdnQ11Y4MsBZJaqH
You are literally the only person on the internet making videos explaining Night Cafe!!! This has helped me TREMENDOUSLY!! Thank you!
I enjoy working with Night Cafe. I am happy to hear that my videos help you.
Thank you for the uplifting words. I will try to make an updated one of NighCafe soon.
found this video while trying to figure out how to get nightcafe to generate images approaching what i want. trying to describe a scenario with two people in different clothes and with different expressions is my ultimate goal du jour , but this is a great intro to the basics of what NOT to include
Thank you. I have about 20 videos on this topic of using Nightcafe. I hope you enjoy the other other ones. Thank you again.
Oh Daniel, now you are going to get me started in another hobby. I love art, but can't really draw myself so this looks very interesting! I hope you and your family are doing well.
You can do it!
Thank you!!!! I will have fun now, without wasting my credits due to a lack of description of what I am trying to create.
You're welcome. I am making newer videos on this topic. I hope you like those new Nightcafe videos too.
@@selmateacher7 I will look for them.
@@Cottage-Reader Oh.. Thank you. Yes I just made one, it will be published on Aug 15th, It goes over comparing SDXL vs. all other models, using same seed and prompts. I also go over many of the prompts that I find useful for different types of artwork. You can see my artwork at:
creator.nightcafe.studio/u/ArtDreamer
I was slow to use negative prompts because I didn't know about them and when I learned about them I didn't know how to implement them correctly. So I was making a lot of bad images even though my prompts were constructed well.
When I finally learned how to implement negative prompts, the extra limbs, cockeyed facial expressions, and out-of-frame images practically ceased. The overall image quality also increased. I wish I'd known about your suggestions months ago. Thanks, you've even helped me with a few extra prompt ideas!
Thank you for subscribing. that means a great deal to me. I am always here to help.
Thank you so much!! I was so confused before, and no other tutorials explained this. You're awesome!
Thank you for saying so. I always try to help. I have many newer videos on this topic that cover the other options in more detail.
This is so helpful. Couldn't figure that out.
Thank you!😘
I am always here to help. Thank you for your kind words. I hope you might subscribe.
@@selmateacher7 Thank youI subscribed 😇... I have actually spent the past 4 hours trying to get an angel holding an antique brass scale. I have tried every combination of words in the prompt and negative prompt I can think of. The style is fine, but I just cannot get the engine to come up with a an antique set of scales. And weird things will happen, like the wings are set in the lower back, or there will be an extra wing floating about. 😣Thank you so much if you have time to give any advice.
@@TatyanaValdaBelindaHill You can use the Mask tool. I used it today. I created a woman holding a vacuum cleaner, but one of her eyes were just wrong. So I masked her face, used another seed and bang! Perfect face, with perfect vacuum cleaner.
@@selmateacher7 Ohhh! I was thinking a feature like that would be just perfect. I am new to AI, so I didn't notice the functionality was already there. Will save me a ton of photoshopping and time/money with zillions of retries and credits. (I spent another 7 hours on various other images last night and had a huge problem with deformed hands - even with all the negative prompts added and refining positive prompts.) Thank you so much again. Just about to open up Night Café now. 😘
ZBrush was originally released in 1999. It is still used to create 3D models for film today. It's interface and modelling style feels familiar to sculptor's and 2D artists. It excels at creating organic and highly detailed models.
Nightcafe just released a new SDXL model for stable diffusion that might be possible to create a greater variety of art in higher detail.
Yes. You are correct, I thought I cropped that part out of my newer videos. Zbrush is a art app.
Beautiful video, very informative. If I may ask: with the new sdxl they have suggested to reduce the amount of prompts as much as possible (despite seeing some quite spectacular results with huge lists of them) , does anybody find a short amount of them more beneficial than a long and detailed list? Last but not least, what’s the weight you guys set on the prompts: I’ve tried both with positive and negative from their minimum to to their maximum extend (+/-3) and I cannot see any tangible difference, any feedback on this?
Thank you so much for your very kind words. I believe I show an answer to your questions on multiple prompts on this video:
ruclips.net/video/nrPIYRP22VM/видео.html
I still use multiple prompts on some of my work within SDXL and it turns out.. great.
About weights..... it does make a big difference when you are using many prompts. I show in my videos that prompts with the higher weight number will have a greater influence on the overall creation, but lesser prompts are still represented. Also, the seed has some influence with which prompts are seen to a greater degree in the final output. I encourage you to see all of my recent videos on this topic as I try to make it enjoyable.
So kind of you. Thanks!
You're welcome. I have many newer videos on Nightcafe, now.
Thank you
You're welcome
That's amazing! I'm trying to learn about AI so I can write about it convincingly (as a fiction author), but it's still intimidating. However, you've given me hope by using prompts to get Indiana Jones from Bono and extra-limbed Hunter S. Thompson. 😄 Thank you so much for sharing what you've learned so far.
I am so happy to hear that. I have many newer videos on this topic that you may be interested in.
Thanks for this. Big help.
I am happy it helped.
Thank you very much for this!
I am happy you like it. You're welcome.
I have more on the way.
Wow! Night Cafe is one of the most AI image creator I love! I will try your prompts. Do you have something for coloring page books? Thank you
Yes. add this to your prompt: coloring page
here to learn, lets do it
Yes. Nightcafe is fun to work with.
Thank you!
You're welcome
Thank you so much for your help! Question..what happens when you put the negativity at higher than 1%?
The effect is minimal. when I tested that, I notice only very minor change in the output.
nice
Thank you
Great video!!
Thank you
272nd like, and subscribed to the channel now at 1:52 AM
Thank you very much,
Very interesting. Thanks so much. I'm a total beginner. Question: Can I use a photo of someone as s start image and tell AI to portray them as Indiana Jones or something? Or Do you have a video showing using real people as the start image?
You may use a start image, yes. I made an example of my own photo and opened prompt as seen here:
creator.nightcafe.studio/creation/kBeEmrnOnkTg2qs6gYqH
wow! Yes you can and it works! I left my prompts open for you so you can see what I wrote, but it did take my face and put a hat, jacket, and beard on it. This is so cool!
@@selmateacher7 Great! I will give it a try!
@@selmateacher7 Drat. I tried it but it doesn't use the face at all? I chose Stable Diffusion and tried to follow your same prompts?
@@SherrisWorld Is it a famous face? There are other apps that do copy a non-famous face, like AI magic. But even that does not do perfectly, as it takes hundreds of photos to train an AI to create a specific face . Just to let you know, I never even thought of that until you asked. I look alot like Indiana Jones, so it worked for my face, LOL. Also, I used my stage name, "selmateacher7" so that it would use that face and not my common name, as it uses the singer that is called "daniel berry"
This are very helpful informations, thank you!
You're welcome. I am always here to help. I hope you might subscribe and enjoy my other, newer videos on this topic. As I love AI generated art.
Thanks for this video!
You're welcome
Do you know of a AI audio generator, thanks for this video
there are many out there, this was one that I found interesting:
speakperfect.co/?ref=futuretools.io
another place to look is here:
aitoolsdirectory.com/
@@selmateacher7 I didn't get a chance to visit them because I am not at PC yet, are those tools free to use and they make audio not images ??
@@Specialist1Aim Those are free tools. There are many AI tools out there. The one I found interesting is for audio, the second is a list of many different AI tools, audio is one of the categories.
@@selmateacher7 This is awesome because I need something to help speed up workflow, it can auto generate sound effects like door opening sound or wind noises ?
@@Specialist1Aim Thank you for the clarification on what type of audio you need generated. I found this AI tool to generate sound effects:
myedit.online/en/audio-editor/ai-sound-effect-generator
Tried all that - eyes still disfigured.
Hello there. It can draw good eyes and hands, but it is hard for the AI to do it 100%. Here is my work and it shows perfect eyes and hands. My newest work is all "open prompts" so you may copy the prompts , if you wish.
here is the link to my gallery:
creator.nightcafe.studio/collection/dBFNhLdpG8Xvm9adj7Zx
here is one of my works, called the embrace. I have the prompts open for you and the eyes are done correctly . here is the link:
creator.nightcafe.studio/creation/u6ZLP2CB3bM0qwBzvBEr
I have been looking at everyone who follows me and they all have perfect eyes. Which username do you use, so I can see your work?
I have made a video , just for you.... here is the link to the new video: ruclips.net/video/CKqRxCk37pw/видео.html
Even using negative prompts I still get disformed figures.
I am sorry to hear that. please try "advanced mode" then try different seed numbers. That happens to me sometimes, but I try to always use different seeds to get it just right. As seen here:
creator.nightcafe.studio/creation/p3vBtPQYw1ASHSHWHSTW
@@selmateacher7 I've been trying and I still get all disformed faces I wasted so many credits.
@@AutumnPeir Hello there, I am sorry to hear that. Please try again, but use the settings , 1 image, short, and thumbnail to create without using credits. Also, try different seed numbers . I know it may seem hard, but you can find success. I use nighcafe every day and once you have found a good combination, I normally just evolve using the same prompts.
first prompt: white mennonite cap, White cream, white lotion, white drops, white droplets, facial, white drips woman
weight : 1
prompt 2 : ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, extra limbs, disfigured, deformed, body out of frame, blurry, bad anatomy, blurred, watermark, grainy, signature, cut off, draft
weight: -1
seed: 444219900
modeL: Stable Diffusion v1.5
here is the link to what it made:
creator.nightcafe.studio/creation/6QeOwdnQ11Y4MsBZJaqH
@@selmateacher7 What a beautiful picture...! 💖😇
@@wallhanna1291 Thank you. I try to be active with NightCafe. creator.nightcafe.studio/u/SelmaTeacher7
Thank you so much ! I subscribe and follow you on Night Cafe as Mimi93France ❤
Thank you.