DRAWING Bleach Speedpaint [CLIP STUDIO PAINT] | Draw Yoruichi Shihōin Anime Fanart 

…??? did I miss something? This was the full artwork.

oh never mind, I just saw the time that you posted your comment. you posted before the video aired, that makes sense. sorry, I pulled an all nighter to get this posted today and I only just woke up, lol. I’m glad you like how it looks, I hope you enjoy the video!

As a warning, if you decide to watch Bleach, it has so many fillers… just so many. They also make a habit later of interrupting primary story arcs with massive groups of fillers. They will literally just pause the primary arc and slap in a non-contemporary filler story before resuming right back where they were in the primary arc. But the main story is fun and enjoyable despite all of the fillers.

Actually if you look up one of those lists that tells you which episodes are fillers, you could probably shave a few hundred episodes off the runtime.

Well I’m glad to hear that you really enjoyed the video. Without giving anything away, I’m hoping I can do something special with this painting… Something I haven’t done before. Now this isn’t something I’m officially announcing. This unrevealed process is going to require me learning yet more new software, but if I pull it off this might not be the last video I post with this artwork in a fun way! Just something to look forward to.

@@JRsJunkDrawer Oh, my spidy-sense is tingling :D I'm looking forward to whatever is to come and I wish you all the luck and fun!

Also, to give you a few tips about using Stable Diffusion - it would have been nearly impossible to get it to provide useful references if it wasn’t for having fed it my grayscale to start with. So make sure you try out image to image mode if you haven’t given that a shot already. Feeding at a grayscale with a hint of what your lighting is going to be will help it keep that lighting on track for what you have in mind. FYI if you find it is only outputting black and white images, it happens sometimes when you give it a grayscale, go back to your art program and place a color blending mode layer above your grayscale and splash in rough color. It only has to be enough so that stable diffusion realizes you want color. My second tip would be to make heavy use of the negative prompt section to feed it information that might sound basic but it does help the program stay on track, broken anatomy, third arm, badly drawn, ugly artwork etc. If you’re having trouble with it listening to some of your positive prompts, try enhancing the value assigned to it. If you highlight a word or descriptive phrase, you can press control and the up arrow to have the program put more value behind it. FYI you can also use control and down arrow to decrease the value. These also work in the negative prompt section. Also don’t be afraid to tell it to do the same thing in multiple ways, if it’s really being stubborn about listening to you. Because my computer is on the slower side, I like to keep my sampling steps Low around 20 when I’m just generating random images, but when I find something I like, I lock the seed and bump those sampling steps up to 50 or even 100. This will get a lot more detail out of your details I like to keep my Denoise strength between .4 and .6 And be careful not to push the CFG scale to high or else it will over sharpen your image. Anything above 11 would be pushing it. And my final tip would be to not be afraid of taking multiple images that it outputs bringing them into the art editing program of your choice and mashing together the aspects of each that you like. If you feel the face on one output is closer to your desired results, but the arm on a different one is what you want, combined them in your art program and then send that image back into stable diffusion’s image to image mode as your new source. I know that was a lot, those are some of the basic things I learned, I hope it helps.

@@JRsJunkDrawer That you so much! Feeding it a greyscale image with a little colour shpuld be a no brainer but i honestly didnt think of that. Something I'll defnetly tryin the future :D The CFG is something i never thought off, maybe my problem is indeed also there and the emphesis on promt is something I always forget ^^" it's so cool to be an artist and being able to really use the ai through splicing images together! ^°^ Or putting one down as a base and overpainting it is also cool :D (Pretty lazy shortcut I know, but i dont feel guilty because it ends up all covered in the end anyways ;) ) Oh and there was somewhere a vid of someone who used special tricks to generate a pretty good turnaround sheet that made it possible to make a 3d model out of the desing. so cool :D but i think that's all local run stuff

Well, if you are looking to follow a workflow like the one I showed here, Stable Diffusion is probably your best bet to set up your references. I understand that the program may be intimidating to learn, but any versatile piece of software is going to have its fair share of complication. Actually, in my opinion, while Stable Diffusion might have a lot of features and look complicated, once you get a handle on the basics it is very straight forward. It still takes time to refine useable outputs, but a few hours of bouncing between an art program and Stable Diffusion can net you some very useable references. I also gave someone else here in the comments some Stable Diffusion advice. To copy and paste the same tips I gave them - it would have been nearly impossible to get it to provide useful references if it wasn’t for having fed it my grayscale to start with. So make sure you try out image to image mode if you haven’t given that a shot already. Feeding it a grayscale with a hint of what your lighting is going to be will help it keep that lighting on track for what you have in mind. FYI if you find it is only outputting black and white images, it happens sometimes when you give it a grayscale, go back to your art program and place a color blending mode layer above your grayscale and splash in rough color. It only has to be enough so that stable diffusion realizes you want color. My second tip would be to make heavy use of the negative prompt section to feed it information that might sound basic but it does help the program stay on track, broken anatomy, third arm, badly drawn, ugly artwork etc. If you’re having trouble with it listening to some of your positive prompts, try enhancing the value assigned to it. If you highlight a word or descriptive phrase, you can press control and the up arrow to have the program put more value behind it. FYI you can also use control and down arrow to decrease the value. These also work in the negative prompt section. Also don’t be afraid to tell it to do the same thing in multiple ways, if it’s really being stubborn about listening to you. Because my computer is on the slower side, I like to keep my sampling steps low, around 20 when I’m just generating random images, but when I find something I like, I lock the seed and bump those sampling steps up to 50 or even 100. This will get a lot more detail out of your details. I like to keep my Denoise strength between .4 and .6 And be careful not to push the CFG scale to high or else it will over sharpen your image. Anything above 11 would be pushing it. And my final tip would be to not be afraid of taking multiple images that it outputs bringing them into the art editing program of your choice and mashing together the aspects of each that you like. If you feel the face on one output is closer to your desired results, but the arm on a different one is what you want, combine them in your art program and then send that image back into stable diffusion’s image to image mode as your new source. I know that was a lot, but those are some of the basic things I learned, I hope it helps. Regardless of how you decide to setup your workflow, I wish you the best of luck. I know how difficult it can be to learn a new program without the aid of someone looking over your shoulder giving experienced advice... like stumbling around a dark cave without a flashlight. If you need more help or tips don't hesitate to drop me another comment or DM me on Instagram. If I know the answer, I'll help however I can!

As a follow-up to your edited comment - Thanks! I'm glad to hear that you enjoyed the process! It makes me happy to hear the musculature left a positive impression. I put a lot of work into it. As a follow-up to edit 2: It would be though for me to say exactly one thing that jumped my skill. It would be more accurate to say that taking more time on some paintings, that I feel show more potential, gives me the opportunity to practice more advance techniques and therefore improve my skill over time. Of course, using proper references is an essential part of learning new techniques, but that isn't an instant process, it still takes time to practice what you observe. Balancing a learning process with efficiency is an ongoing struggle, and usually comes at the cost of more time put into each project in the short term for better results in the long term.

@@JRsJunkDrawer Thank you for the in-depth response! You've convinced me. I'll look into Stable Diffusion over the weekend and see what it does to some of my sketches with greyscale values. I'm already excited.

Well, if you need any assistance, I can help you with the prompt and settings. Your first step working with your artwork will be finding a model that produces an art style you're comfortable using as a reference. I used civitai (dot com) to get some assets (models and stuff). I'm warning you now that Stable Diffusion and civitai have no aversion to... how should I put it... forgetting the clothing... it just comes down to your prompt, model, and random seed. So if that is something that will bother you, well, I did my part, I warned you in advance. Second step should be bringing your sketch/greyscale into "image to image" (may need to add color first) Then load your model - setup the scale of the output, just don't exceed what your computer can handle - setup your prompt and negative prompt (I can help with that) - there are a few other settings that if you run into issues with, I can tell you what I used. I look forward to hearing how it works out for you!

I plan on it eventually, but right now I'm still refining the settings and diversity of the brushes I have. I actually made and experimented with a dozen new brushes for this painting, but I still want more from them. I'm busy for the next few weeks, but since you've expressed interest in getting my brushes I'll take some time, when I can find some to polish them and get them ready.

sorry I haven’t been posting anything lately. I am still alive… things have just been crazy around here, but now the new roof is on and a lot of the other projects are fixing up nicely too. i’m trying to get back into painting in full, but I keep having other projects pop up and take away my time. thank you for asking though, and I do hope to be back to posting soon.