Do you struggle with prompting? 🌟 Download a sneak peak of my prompt guide 🌟 No Membership needed: ⬇ Head over to my Patreon to grab your free copy now! ⬇ www.patreon.com/posts/sneak-peek-alert-90799508?Link&
I really appreciate your positive feedback, thank you! I completely understand your frustrations with inpainting - it can be a bit of a challenge sometimes. With this in mind, I'm considering creating a comprehensive tutorial on inpainting in the future. Stay tuned for that, and hopefully, it will help ease some of those frustrations!
Love these videos about tools. I didn’t even know they were out on tools for stable diffusion. One request, think of the newbies and keep them in mind. Always show how to install each tool when you create a video about a new tool. It helps so very much. Thank you for your generous content. I’m going to be heading over to your patreon page.
The commenter's message is very positive, showing appreciation for your videos and providing a helpful suggestion. A response that acknowledges their feedback and thanks them for their support would be appropriate. Here's a suggested reply: Thank you for your kind words and valuable suggestion! I'm glad to know that my videos on tools for Stable Diffusion are helpful. Your point about considering newbies is well-taken; I think this is one of the view videos where i don't show the installation. All my new Videos show how to install. Thanks for considering supporting me on Patreon.
This tool seems like it would be a big help - I'm going to try to figure out how to use it with my workflow. I've got it installed, just need to do some experimenting with it. Thank you for highlighting this really cool tool!
You're very welcome! I'm thrilled to hear that you've installed the tool and are excited to experiment with it. As a quick tip, if you're working with images involving multiple persons, you might find this tool particularly useful when inpainting faces. You can simply mask out the face that should be inpainted using After Detailer. Experiment, play around, and see how it fits into your workflow.
@@AIKnowledge2Go I've been using it for the past couple of days and it's been a game changer for speeding up my workflow! Thank you for introducing us to it! I'll try that tip for images with multiple people as well. Thank you!
I love discovering new helpful tools, and this is one of the best so far. Saves an amazing amount of time indeed! So many, many thanks for posting this video! 💛
I'm thrilled to hear that you found this tool so useful and that it's saving you time! Your feedback means a lot. Thank you for your kind words and for watching the video!
Danke für das Video, du machst das sehr gut, obwohl es nicht deine Muttersprache ist, mach bitte weiter. Ich kann immer bei deinen Videos immer was mitnehmen was mich weiterbringt, bin neu in dieser fazinierenden Welt mit AI und KI lol. Klares Abo.
Vielen Dank für das positive Feedback! Es freut mich zu hören, dass meine Videos hilfreich sind. Herzlich willkommen in der spannenden Welt der AI und KI! Bleib dran, es gibt immer Neues zu entdecken. Danke für das Abo!
Great, exactly what I was still looking for! I'm not a big fan of all these face restorers, they often make things even worse, so I don't use them. The only reliable way for improving faces I found so far was to upscale the image in latent space.
I'm glad you found the tool helpful! You're right, scaling up the image often helps, but Adetailer does a fantastic job too. Thanks for sharing your experience!
I totally agree! Those blue highlights turned out to be a pleasant surprise. Sometimes what might seem like an unexpected outcome can enhance the final result.
I'm delighted to hear that you found the post beneficial, especially in enhancing AI's drawing capabilities. Thank you for your lovely feedback, it truly brightens my day! ^.^
Great videos and channel. Glad I found you! One question, can you please add in the description the prompt (full generation data) of the images you create? Like the prompt you used in this video. Then it's easier to recreate and follow your steps.
Thank you for the kind words! I'm glad you're finding the channel helpful. I completely understand your request and will make sure to share prompts in future videos for free. However, for this specific prompt set, it was exclusively shared with my Patreon supporters. I hope you understand, as I want to honor their support. Thanks for your understanding and continued support!
Vielen Dank für deinen Kommentar! Es freut mich zu hören, dass du die Informationen hilfreich findest und vorhast, sie auszuprobieren. Ja, du hast recht, es hat Ähnlichkeiten mit der Control reference only, aber es ist spezifischer auf das Bild abgestimmt. Ich hoffe, dass es dir bei deinen eigenen Projekten hilft. Lass mich wissen, wie es läuft. Viel Spaß beim Ausprobieren!
@@AIKnowledge2Go leicht ist es nicht auch mit den Detailer nicht der braucht auch seine ganz speziellen Feineinstellung und selbst dann bin ich nicht wirklich zufrieden vor allem wenn es um Hände geht, das Problem ist ja auch dass man nur spezielle Bereiche automatisch inPainten kann, und so stößt man eben sehr schnell auf die Grenzen vom ADetailer
Batch size determines the number of images your GPU processes at once in a single batch. A larger batch size means the GPU processes more images simultaneously, but it also requires more memory. If your batch size is too large for your GPU's memory, you may experience out-of-memory errors. Batch count, on the other hand, specifies how many of these batches should be processed. So, if your batch size is 4 (processing four images at once) and your batch count is 3, you would process a total of 12 images (in three separate groups of four). So, it's all about balancing the processing capacity of your GPU (which determines your batch size) with the number of images you want to generate (which is determined by your batch count).
In prompts, parentheses can be used to increase the weight of a specific word or phrase. Each word in a prompt typically has a weight of 1.0. By enclosing a word or phrase in parentheses, its weight is boosted to 1.1. Double parentheses increase this to 1.2. Alternatively, you can also specify the weight directly after a term, like "example:1.2". I hope that clarifies things!
Is there a setting to keep the faces from becoming too blurry? My original faces look way better than after running ADetailer. Thanks for the tutorial.
Indeed, it can be tricky to maintain facial details. As a deviation from my tutorial, you might want to try turning off the "restore faces" option. That was an oversight on my part during the tutorial. Additionally, you could consider using the image2image feature for upscaling, with the After Detailer option enabled. Another way is inpainting the facial area. Ensure that the "only masked" option is switched on. This might help maintain those all-important facial features. Happy creating.
Depends if you are happy with the result you are having atm. Under Settings -> Adetailer, you can set: Save images before ADetailer. Then you get two images with before/after for your comparison. Hope that helps. 😉
Amazing video, thanks. Why it generates a completly new image when I try to use adetailer on a already generated image, how to use it to work on already generated image
Thank you for your positive feedback! When using After Detailer on an already generated image, you cannot directly apply it as a separate step. However, you can incorporate After Detailer into your workflow by following these steps: Send your image over to Inpainting. In the Inpainting settings, select "Masked Only" to focus the process only on the masked area, which is typically the face. Once the face is masked, apply the prompt in After Detailer. In After Detailer. By following these steps, you can use After Detailer to specifically work on the face while keeping the rest of the image intact. This allows you to fine-tune the details and improve the quality of the face without altering the entire generated image. I hope this explanation clarifies the process for using After Detailer on already generated images.
Thanks for the kind words! As for your question, you might want to give ControlNet a try with either the SoftEdge or Canny model. The choice between these two depends on the complexity of the object you're trying to copy. For detailed machinery, the Canny model would be more suitable, while for simpler or more organic shapes like wings, SoftEdge would be a better fit. I have two tutorials on my channel, AIKnowledge2Go, that go more in-depth about using ControlNet. Here they are for your reference: Basic: ua-cam.com/video/dgRtt_yt3yA/v-deo.html Using 2Controlnet Models at the same time: ua-cam.com/video/RY5XqF38DFg/v-deo.html While these tutorials may not specifically cover your exact use case, they should provide you with a good understanding of how to use ControlNet effectively
When using 'After Detailer', it sometimes inpaints the entire face, which can lead to the loss of specific details like a 'grin'. While simply typing 'grin' in the positive prompt can work, to ensure the desired result, I often include it in the 'After Detailer' prompt. However, in the newer versions of 'After Detailer', if you leave the prompt empty, it will automatically take cues from the positive prompt.
Very helpful, thank you very much. This tool is great for editing faces when generating new images. However, what if I want to change the faces of 50-100 images that I already have? Is there a quicker way instead of inpainting them one by one?
Thank you for your comment! I'm glad you find the tool helpful for editing faces in new images. If you have a large number of existing images and want to change the faces in all of them, manually inpainting them one by one can be time-consuming. In such cases, you might want to explore batch processing options. In Automatic 1111 go to the image2image tab -> From there to the batch tab. From there you can process a whole folder. For me after detailer works great even without a prompt. if your images are low res like 512 x 768 id suggest upscaling them first. i have this tutorial for a general beginner / intermediate workflow where i discuss upscaling. ua-cam.com/video/RiN-zrUlneQ/v-deo.html
That's great to hear that you're considering getting an Nvidia 4080 GPU to help with rendering in Stable Diffusion. The 4080 RTX is a powerful graphics card that can significantly improve rendering speed and performance. However, please note that in my videos, I often fast forward the rendering process to save time and focus on the key concepts and techniques.
Thanks for the tutorial - seems very powerful! Do you know how it reacts when we add a lora or a textual inversion into the adetailer prompt? I mean, is there a possibility for it to distort the image if the same lora/textual inversion is NOT in the main prompt?
Thank you so much for your insightful comment and kind feedback. I truly appreciate it. As for your question, from my understanding, the process functions more like the "mask content original" option, essentially in-painting the image based on the existing pixels. This means that the AI detailer doesn't necessarily create anything new but rather, adjusts itself according to the existing 'pixels'. While you may not witness any severe distortions, it's possible that some aspects like faces may appear unusual. For instance, if you were to incorporate a lora of a celebrity's face into the 'Ad Detailer' prompt that was not used in your main prompt, it might result in some odd outcomes. I hope this clarifies your question, and I'm here to assist further if you have any more queries!
@@AIKnowledge2Go so I tested it with textual inversions - if the same face ti is used in the prompt as well as adetailer - it works well but if you have multiple people in the image, say a man and a woman, both will end up with almost identical faces. Maybe a bit of inpainting will help solve that 😁
@@AIKnowledge2Go Yes - I ended up using ADetailer in the inpainting section and was able to get way better face results with a bit of prompt tweaking - managed to get multiple characters, with different faces though it took a few extra steps - 1 additional step per person but worth it!
Looking to craft AI-Art but GPU's holding you back? Dive into my latest video: ua-cam.com/video/mnjMUIBu33k/v-deo.html. I've teamed up with a stellar (and free) AI Online generation platform that even supports Loras, ControlNet and SD XL! 🎨🔥
Hi, thanks for your comment. If the faces you're working with are far away, it's possible that your image resolution might be too low to preserve enough detail. Try using a higher resolution image and see if that improves your results.
Hi is your After Detailer extension working right now? And which version is it? My v23.7.5 version isn't working. I have installed v23.5.19 which is released over 1 month ago and it doesn't work too.
Hello! I'm currently using the After Detailer extension without any issues. Have you considered updating your Automatic 1111? The latest version available is 1.4.1. If updating doesn't resolve the issue, a fresh reinstall could be worth trying.
@@AIKnowledge2Go I found the solution. Just delete the --skip-install in the web-user.bat (I forgot its actual name but it's in the SD folder) and install the extension again. Other extension of mine is working now thank to this fix too.
Thank you, I'm quite flattered by the 'stable diffusion hero' title. The idea of being a pagan god in this realm is intriguing. Now, I just need to figure out a cool power and name. Any ideas? Have a great day as well!
Sorry I need to get it out my chest, please stop using restore face, it ruins your images. It's defeat the purpose of using Adetailer, you meybe dont see it. if you dont believe me go to PS and adjust the curve to see the brightness value and see it yourselves there (it's destroy the value composition between dark and light, I'm bad at explaining stuff. sry). I notice people use Restore face a lot without knowing what it does to their image. it's less noticeable (but I can still see it) with realistic model when using restore face. but yeah stop it guys. :( just use Adetailer and input the separate prompt there and adjust the denoising instead, it will be more cohesive with overall composition. nice content btw, and I need to figure out how to adjust the config for Adetailer, so I dont need to adjust any parameters each time I want to use it.
Thank you for your detailed feedback. You're absolutely right, and I appreciate your insight. The use of 'face restore' was indeed a flaw in my workflow. I initially turned it on to demonstrate the before and after effect at the start of my video. However, I understand the issue you're pointing out regarding the impact it has on image quality. I'll certainly be more careful with this in future tutorials. Thanks again, and I'm glad you're enjoying the content!
At the 3:00 mark, I press Ctrl + Enter to kick off the render process. It's a keyboard shortcut that does the same job as clicking 'Generate' with your mouse. Handy for speeding things up a bit!
Do you struggle with prompting? 🌟 Download a sneak peak of my prompt guide 🌟 No Membership needed: ⬇ Head over to my Patreon to grab your free copy now! ⬇ www.patreon.com/posts/sneak-peek-alert-90799508?Link&
This is great. Inpaint is a bit frustrating to use sometimes. Thanks
I really appreciate your positive feedback, thank you! I completely understand your frustrations with inpainting - it can be a bit of a challenge sometimes. With this in mind, I'm considering creating a comprehensive tutorial on inpainting in the future. Stay tuned for that, and hopefully, it will help ease some of those frustrations!
Love these videos about tools. I didn’t even know they were out on tools for stable diffusion. One request, think of the newbies and keep them in mind. Always show how to install each tool when you create a video about a new tool. It helps so very much. Thank you for your generous content. I’m going to be heading over to your patreon page.
The commenter's message is very positive, showing appreciation for your videos and providing a helpful suggestion. A response that acknowledges their feedback and thanks them for their support would be appropriate. Here's a suggested reply:
Thank you for your kind words and valuable suggestion! I'm glad to know that my videos on tools for Stable Diffusion are helpful. Your point about considering newbies is well-taken; I think this is one of the view videos where i don't show the installation. All my new Videos show how to install. Thanks for considering supporting me on Patreon.
THANKYOU....I'm a couple of weeks into learning and just found this. Is amazing
I'm glad you found this helpful, especially as you're just starting your learning journey! Thank you for the kind words. Happy creating.
This tool seems like it would be a big help - I'm going to try to figure out how to use it with my workflow. I've got it installed, just need to do some experimenting with it. Thank you for highlighting this really cool tool!
You're very welcome! I'm thrilled to hear that you've installed the tool and are excited to experiment with it.
As a quick tip, if you're working with images involving multiple persons, you might find this tool particularly useful when inpainting faces. You can simply mask out the face that should be inpainted using After Detailer.
Experiment, play around, and see how it fits into your workflow.
@@AIKnowledge2Go I've been using it for the past couple of days and it's been a game changer for speeding up my workflow! Thank you for introducing us to it! I'll try that tip for images with multiple people as well. Thank you!
Thanks for tip
You're welcome! I'm glad you found the tip useful.
I love discovering new helpful tools, and this is one of the best so far. Saves an amazing amount of time indeed! So many, many thanks for posting this video! 💛
I'm thrilled to hear that you found this tool so useful and that it's saving you time! Your feedback means a lot. Thank you for your kind words and for watching the video!
Danke für das Video, du machst das sehr gut, obwohl es nicht deine Muttersprache ist, mach bitte weiter. Ich kann immer bei deinen Videos immer was mitnehmen was mich weiterbringt, bin neu in dieser fazinierenden Welt mit AI und KI lol. Klares Abo.
Vielen Dank für das positive Feedback! Es freut mich zu hören, dass meine Videos hilfreich sind. Herzlich willkommen in der spannenden Welt der AI und KI! Bleib dran, es gibt immer Neues zu entdecken. Danke für das Abo!
This is great knowledge. Thank you. I had no idea this was possible.
I'm glad I could introduce you to something new! There's always so much to discover in the realm of AI. Happy creating!
Great, exactly what I was still looking for! I'm not a big fan of all these face restorers, they often make things even worse, so I don't use them. The only reliable way for improving faces I found so far was to upscale the image in latent space.
I'm glad you found the tool helpful! You're right, scaling up the image often helps, but Adetailer does a fantastic job too. Thanks for sharing your experience!
perfect ! thanks a lot !
You're welcome! I'm glad you found the tutorial helpful.
actually that blue highlights on her hair looks amazing😂😂
I totally agree! Those blue highlights turned out to be a pleasant surprise. Sometimes what might seem like an unexpected outcome can enhance the final result.
Thank you for your post, it has been very helpful in improving the drawing capabilities of AI. ^.^
I'm delighted to hear that you found the post beneficial, especially in enhancing AI's drawing capabilities. Thank you for your lovely feedback, it truly brightens my day! ^.^
cool tool. will try it tomorrow. thanks for the tutorial!
I'm glad you found it interesting! Thanks for watching!
that german accent is genius :D
Thank you! I'm glad my German accent brings a unique touch to my videos. I appreciate your positive feedback!
Thank you
You are very welcome!
Great videos and channel. Glad I found you!
One question, can you please add in the description the prompt (full generation data) of the images you create?
Like the prompt you used in this video. Then it's easier to recreate and follow your steps.
Thank you for the kind words! I'm glad you're finding the channel helpful. I completely understand your request and will make sure to share prompts in future videos for free. However, for this specific prompt set, it was exclusively shared with my Patreon supporters. I hope you understand, as I want to honor their support. Thanks for your understanding and continued support!
more tools, more tools 🎉🎉
Have you already watched this one here? Another tool video is definitely on my list.
ua-cam.com/video/8At23HEJnAk/v-deo.html
Thank you so much! ✌
You're welcome! I'm glad you found the tutorial helpful. Don't hesitate to ask if you have any questions.
@@AIKnowledge2Go I do what is the difference between hand,person and media pipe in ADetailer ?
Great 😃👍
Thank you for the feedback! It's always a pleasure to see people enjoying the content. Stay tuned for more!
Danke das klingt sehr hilfreich und werde ich auf jedenfalls ausprobieren, klingt etwas wie control net reverence nur näher am bild
Vielen Dank für deinen Kommentar! Es freut mich zu hören, dass du die Informationen hilfreich findest und vorhast, sie auszuprobieren. Ja, du hast recht, es hat Ähnlichkeiten mit der Control reference only, aber es ist spezifischer auf das Bild abgestimmt. Ich hoffe, dass es dir bei deinen eigenen Projekten hilft. Lass mich wissen, wie es läuft. Viel Spaß beim Ausprobieren!
@@AIKnowledge2Go leicht ist es nicht auch mit den Detailer nicht der braucht auch seine ganz speziellen Feineinstellung und selbst dann bin ich nicht wirklich zufrieden vor allem wenn es um Hände geht, das Problem ist ja auch dass man nur spezielle Bereiche automatisch inPainten kann, und so stößt man eben sehr schnell auf die Grenzen vom ADetailer
bro, what is a diference betwen Batch size and Batch count?
Batch size determines the number of images your GPU processes at once in a single batch. A larger batch size means the GPU processes more images simultaneously, but it also requires more memory. If your batch size is too large for your GPU's memory, you may experience out-of-memory errors.
Batch count, on the other hand, specifies how many of these batches should be processed. So, if your batch size is 4 (processing four images at once) and your batch count is 3, you would process a total of 12 images (in three separate groups of four).
So, it's all about balancing the processing capacity of your GPU (which determines your batch size) with the number of images you want to generate (which is determined by your batch count).
Thanks bro. ❤
You're welcome! I'm glad you found it useful. Stay tuned for more content. ❤
what does the ((example)) double double parentheses close and open do? on promt please?
In prompts, parentheses can be used to increase the weight of a specific word or phrase. Each word in a prompt typically has a weight of 1.0. By enclosing a word or phrase in parentheses, its weight is boosted to 1.1. Double parentheses increase this to 1.2. Alternatively, you can also specify the weight directly after a term, like "example:1.2". I hope that clarifies things!
Is there a setting to keep the faces from becoming too blurry? My original faces look way better than after running ADetailer. Thanks for the tutorial.
Indeed, it can be tricky to maintain facial details. As a deviation from my tutorial, you might want to try turning off the "restore faces" option. That was an oversight on my part during the tutorial. Additionally, you could consider using the image2image feature for upscaling, with the After Detailer option enabled. Another way is inpainting the facial area. Ensure that the "only masked" option is switched on. This might help maintain those all-important facial features. Happy creating.
@@AIKnowledge2Go thank you for the additional advice and insight. Good luck with your Channel.
Hey when i'm upscaling using the ultimate SD upscale, should I keep the ADetailer on?
Depends if you are happy with the result you are having atm. Under Settings -> Adetailer, you can set: Save images before ADetailer. Then you get two images with before/after for your comparison. Hope that helps. 😉
great
Thank you! I'm glad you enjoyed it.
Amazing video, thanks.
Why it generates a completly new image when I try to use adetailer on a already generated image, how to use it to work on already generated image
Hi, same problem here.
Have you figured out a solution by any chance?
Thank you for your positive feedback! When using After Detailer on an already generated image, you cannot directly apply it as a separate step. However, you can incorporate After Detailer into your workflow by following these steps:
Send your image over to Inpainting. In the Inpainting settings, select "Masked Only" to focus the process only on the masked area, which is typically the face. Once the face is masked, apply the prompt in After Detailer. In After Detailer.
By following these steps, you can use After Detailer to specifically work on the face while keeping the rest of the image intact. This allows you to fine-tune the details and improve the quality of the face without altering the entire generated image.
I hope this explanation clarifies the process for using After Detailer on already generated images.
Great video learning alot! Do you know any extension for A1111 that can detect and copy objects too that can be used in different characters? Cheers
Thanks for the kind words! As for your question, you might want to give ControlNet a try with either the SoftEdge or Canny model. The choice between these two depends on the complexity of the object you're trying to copy. For detailed machinery, the Canny model would be more suitable, while for simpler or more organic shapes like wings, SoftEdge would be a better fit.
I have two tutorials on my channel, AIKnowledge2Go, that go more in-depth about using ControlNet. Here they are for your reference:
Basic: ua-cam.com/video/dgRtt_yt3yA/v-deo.html
Using 2Controlnet Models at the same time: ua-cam.com/video/RY5XqF38DFg/v-deo.html
While these tutorials may not specifically cover your exact use case, they should provide you with a good understanding of how to use ControlNet effectively
But why don't you just type the word "grin" in the positive prompt, and it will generate it ??
When using 'After Detailer', it sometimes inpaints the entire face, which can lead to the loss of specific details like a 'grin'. While simply typing 'grin' in the positive prompt can work, to ensure the desired result, I often include it in the 'After Detailer' prompt. However, in the newer versions of 'After Detailer', if you leave the prompt empty, it will automatically take cues from the positive prompt.
Very helpful, thank you very much. This tool is great for editing faces when generating new images. However, what if I want to change the faces of 50-100 images that I already have? Is there a quicker way instead of inpainting them one by one?
Thank you for your comment! I'm glad you find the tool helpful for editing faces in new images. If you have a large number of existing images and want to change the faces in all of them, manually inpainting them one by one can be time-consuming. In such cases, you might want to explore batch processing options. In Automatic 1111 go to the image2image tab -> From there to the batch tab. From there you can process a whole folder. For me after detailer works great even without a prompt. if your images are low res like 512 x 768 id suggest upscaling them first. i have this tutorial for a general beginner / intermediate workflow where i discuss upscaling. ua-cam.com/video/RiN-zrUlneQ/v-deo.html
@@AIKnowledge2Go Awesome. Thank you again. This stuff is really fun and exciting.
What's your GPU? I'm thinking on buying a Nvidia 4080 to help render Stable Diffusion faster.
That's great to hear that you're considering getting an Nvidia 4080 GPU to help with rendering in Stable Diffusion. The 4080 RTX is a powerful graphics card that can significantly improve rendering speed and performance. However, please note that in my videos, I often fast forward the rendering process to save time and focus on the key concepts and techniques.
Thanks for the tutorial - seems very powerful! Do you know how it reacts when we add a lora or a textual inversion into the adetailer prompt? I mean, is there a possibility for it to distort the image if the same lora/textual inversion is NOT in the main prompt?
Thank you so much for your insightful comment and kind feedback. I truly appreciate it.
As for your question, from my understanding, the process functions more like the "mask content original" option, essentially in-painting the image based on the existing pixels. This means that the AI detailer doesn't necessarily create anything new but rather, adjusts itself according to the existing 'pixels'.
While you may not witness any severe distortions, it's possible that some aspects like faces may appear unusual. For instance, if you were to incorporate a lora of a celebrity's face into the 'Ad Detailer' prompt that was not used in your main prompt, it might result in some odd outcomes.
I hope this clarifies your question, and I'm here to assist further if you have any more queries!
@@AIKnowledge2Go so I tested it with textual inversions - if the same face ti is used in the prompt as well as adetailer - it works well but if you have multiple people in the image, say a man and a woman, both will end up with almost identical faces. Maybe a bit of inpainting will help solve that 😁
@@AIKnowledge2Go Yes - I ended up using ADetailer in the inpainting section and was able to get way better face results with a bit of prompt tweaking - managed to get multiple characters, with different faces though it took a few extra steps - 1 additional step per person but worth it!
Looking to craft AI-Art but GPU's holding you back? Dive into my latest video: ua-cam.com/video/mnjMUIBu33k/v-deo.html. I've teamed up with a stellar (and free) AI Online generation platform that even supports Loras, ControlNet and SD XL! 🎨🔥
Ur from Germany right? 😊 VG aus FRA
Ich kann meinen Akzent wohl nicht verstecken, oder? 😂 Richtig erkannt, komme aus dem Ruhrgebiet. Wie ist es in FRA?
I just discovered this tool as well. But I'm having terrible results for faces that are far away
Hi, thanks for your comment. If the faces you're working with are far away, it's possible that your image resolution might be too low to preserve enough detail. Try using a higher resolution image and see if that improves your results.
Hi is your After Detailer extension working right now? And which version is it? My v23.7.5 version isn't working. I have installed v23.5.19 which is released over 1 month ago and it doesn't work too.
Hello! I'm currently using the After Detailer extension without any issues. Have you considered updating your Automatic 1111? The latest version available is 1.4.1. If updating doesn't resolve the issue, a fresh reinstall could be worth trying.
@@AIKnowledge2Go I found the solution. Just delete the --skip-install in the web-user.bat (I forgot its actual name but it's in the SD folder) and install the extension again. Other extension of mine is working now thank to this fix too.
By the way, you are now my stable, diffusion hero. If you were a pagan God, I would sacrifice a chicken to please you. L O L. Have a great day.
Thank you, I'm quite flattered by the 'stable diffusion hero' title. The idea of being a pagan god in this realm is intriguing. Now, I just need to figure out a cool power and name. Any ideas? Have a great day as well!
Sorry I need to get it out my chest, please stop using restore face, it ruins your images. It's defeat the purpose of using Adetailer, you meybe dont see it. if you dont believe me go to PS and adjust the curve to see the brightness value and see it yourselves there (it's destroy the value composition between dark and light, I'm bad at explaining stuff. sry). I notice people use Restore face a lot without knowing what it does to their image. it's less noticeable (but I can still see it) with realistic model when using restore face. but yeah stop it guys. :(
just use Adetailer and input the separate prompt there and adjust the denoising instead, it will be more cohesive with overall composition.
nice content btw, and I need to figure out how to adjust the config for Adetailer, so I dont need to adjust any parameters each time I want to use it.
Thank you for your detailed feedback. You're absolutely right, and I appreciate your insight. The use of 'face restore' was indeed a flaw in my workflow. I initially turned it on to demonstrate the before and after effect at the start of my video. However, I understand the issue you're pointing out regarding the impact it has on image quality. I'll certainly be more careful with this in future tutorials. Thanks again, and I'm glad you're enjoying the content!
3:0 whay key you press?
At the 3:00 mark, I press Ctrl + Enter to kick off the render process. It's a keyboard shortcut that does the same job as clicking 'Generate' with your mouse. Handy for speeding things up a bit!
Базарь по-русски)
😄👓💜💛✌️