AMAZING NEW Image 2 Image Option In Stable Diffusion! Apply Styles In Seconds!

Key Takeaways

hello humans my name is your AI Overlord and have you ever been in a situation where no matter what you do no matter how many hours you spend you just cannot get the result that you want use the image the image tab you try your hardest you try your best but no matter what you do you just fail so in the end you just delete stable diffusion throw your computer into the trash and then go crying in the corner of your shower no just me all right well nonetheless today I have a solution for you thanks to a brand new option that was introduced in stable diffusion and that option is called impaging conditioning mask strength now what is the meaning of all these fancy words well how about I'll show you in practice first make sure that you're using the station in painting model then let's generate an image using a very simple prompt simple portrait of a woman with everything by default and then click on generate and this is the final result looks really really good now we're gonna send it to the image image Tab and now instead of this prompt I just want to change the style of that image let's say I want it painted in the style of Vincent Van Gogh so I'm just going to change this into portrait of a woman art by Vincent van Gogh I'm Gonna Keep the impacting conditioning mask strength at 1 which is the default one which is the normal one which is the default value then scroll down the noise's strength is at 0.75 and then I'm gonna click on generate and this is the final result ah yeah I mean it's definitely in the style Vincent van Gogh I can see that but it doesn't look like the base image at all there is absolutely nothing from the base image here not the composition not the color nothing but what if instead of using a value of 1 for my in painting conditioning mask strength I'm gonna use zero and without changing anything I'm just gonna generate my image and this is the final image oh boy boy what do we have here this is basically our base image but in a completely different style look at the difference before and after now let's say I don't want Vincent van Gogh I want something else let's say I want Leonardo da Vinci if I click on generate this is the result that I get which is definitely very close to the style of Leonardo da Vinci while still keeping pretty much everything from the base image and if I were to use again the number one value here and then click on generate this is the sort of image that I get definitely not what I'm looking for and again this is the before and after with Vincent van Gogh before and after and this is with the Leonardo da Vinci style before and after definitely a huge difference and this is basically what this new impacting conditioning max strength option does it allows you to keep the exact same composition of the image with the exact same characteristics but just changing the style and this is a very super powerful tool in the image to image option that allows to basically preserve the coherence of your image so if you just want to change the style of an image you can use this option you can play around the values from this option and create absolutely beautiful pieces of art now to be able to use this you need to make sure that you are using the latest version of stable diffusion so to update stable diffusion you have two solution either you click here on your folder URL and type CMD and press enter and in the command prompt window you're gonna type gitpool and press enter and this will automatically update the stable diffusion folder and the second solution is to right click on web UI user.bat file click on edit with notepad and here just above the call with ui.bat line you're gonna put git pull and then You're Gonna Save the file and now each time the universe stable diffusion it will automatically update the folder the second thing that you need is the special stable Fusion in painting model so for this you're going to click the link in description down below you can arrive on this page you need a hugging face account for this once you created your account you're gonna accept the terms and conditions on this page then you're gonna come here in files and versions click on sdv1 impacting.kpt and then click here to download and then once you've downloaded the file in your computer you're gonna select it Ctrl X to cut it go into your super stable diffusion 2.0 folder stay with the fusion web UI models sample diffusion and you're gonna paste your model right here and now we're finally ready to go so to be able to use this setting what I recommend you to do is to go to settings and here you would have usually a brand new slider called impacting conditional mask strength but if you want to use it like me on the top of the page and not having to go into the settings again each time if you want to use it I'm gonna scroll down and here on Quick Settings list you're gonna put comma in painting underscore Max underscore weight the command will be in description down below then you're going to scroll up apply settings and then you're going to click on restart graduate and refresh components then you will have this brand new Option at the top of your UI now I really wish that they would just put it in the image to image section but for now this is the best way to do it now here is a ground that shows the correlation between the conditioning image max weight and the nosing weight so for the option to have any impact on the final result you need to play around with the conditional image rate and the noise in strength otherwise you're not gonna get any changes and depending on the kind of results that you want to achieve there is definitely a sweet spot and usually I would say that you want to keep the conditioning Image Weight between 0 and 0.5 and using a very high denoising strength this way you can still keep a little bit of the coherence from the base image and still get a huge influence from the style that you want so I would say that you probably want to stay in this section right here now obviously if you want something a little bit stronger you can still stay in this area right here but this really depends is on your own what kind of results that you want now one very interesting option that you can also use is to use the special loopback script which I've already explained in the previous video now basically this option allows you to process an image and then use that image that you just processed as an input for the next image so here you can input the amount of Loops that you want which is how it says here the amount of times that you want to repeat the processing of an image and use it as an input for the next iteration and here you can change to the nosing strength multiplier now before the loopback script was not very used that well because the final result would be very very random and usually it would be pretty bad because the influence on the image would be way too strong but now that we have this option we could probably have very very good results especially if we decrease the noise of strength a little bit so let's see what to create a 3D render of our image here and we use the normal image to image option this is the kind of result that you would get and if I input a value of a 0 here and click on generate this is the kind of results that I would get but now if we go back and use the loopback script and just decrease the music strength a little bit with four loops and one for the noisy strength change factor and then click on generate this is the kind of result that you get which becomes very very intricate for each generation so if you basically want to add more and more style over your image you can use the loopback option which really depends on the kind of results that you want you should definitely play around a little bit with these options because now with this option you have a complete control over the final results of your image now you can also use this option of course for the impainting mode as you can see right here I use this image as a base and I just wanted to train out with blue eyes this works really really well because if I input 1 here this is the kind of result that I would get which is not bad but it definitely does not keep the same exact image here so here this is our base image this is our image we're using the impainting method around the eyes with a zero for the impacting conditioning max strength and here is by using the value of 1 which is basically the default impacting experience and again it doesn't look bad but definitely here the only thing that we are changing is the color of the eyes and this is exactly what I wanted so this tool right here the impacting conditioning max strength is definitely one of the most powerful tools that has been added to stable diffusion and there you have it folks now you have absolutely no excuses you should now be able to generate absolutely beautiful masterpieces using the image to image option so play around a little bit with it and let me know what you think in the comments down below thank you so much to my patreon supporters for supporting my videos you guys are absolutely awesome thank you so much for watching don't forget to subscribe and smash the like button for the YouTube algorithm and I'll see you guys next time bye