CONSISTENT Characters Still Work in Midjourney V5!

so you started playing around with my journey version 5 but for some reason none of your old version 4 prompts seem to be doing what you want them to do and worst of all you just can't seem to get consistent characters anymore don't worry we're all in the same boat here version 5 is clearly putting a lot of us to the test because let's be honest version 4 did so much the heavy lifting but version 5 forces us to grow and improve our skills and that's a good thing I was curious to find out whether any of my previous methods for creating consistent characters still work in version 5. I also wanted to see whether I could bring Carla Caruso back to life so if you're curious to find out what worked and what didn't then stick around so I started off by doing what everybody would do I simply entered my old V4 prompt in order to find out what would happen here's a very simple prompt for Carla Caruso imagine frontal close-up on white background beautiful woman bright red braided bun hair only red hair in the style every Marvel comic book and a version parameter of five and just for reference here's what that very same prompt gave us in version 4. however in version 5 we get these images and I know exactly what you're thinking meh sure these are super super realistic but coherence isn't the issue here it's the style this simply isn't what we want and of course it's super frustrating when you add a style to your prompt but then it just gets ignored by my journey so I did a little bit of experimenting it's become clear to me that version 5 of my journey requires a different prompt structure than before I decided to switch things around a little bit and simply move the Style part of the prompt to the front of the prompt imagine Marvel comic illustration and Then followed by headshot of a beautiful woman with bright red braided bun hair and voila this is already much much better however it's still not the exact same type of style that we got before when we were using version 4. it's true that this style is much closer to a Marvel comic book I'll grant you that and I guess this also says a lot about the additional stylization that version 4 has been inserting into our images just by default but I would really like to try to get our old Carla back or at least something that looks even remotely similar so I had another closer look at the original images the illustration is much more polished here almost like a 2d version of a 3D character I know I know this is kind of an oxymoron but let's try the following prompt imagine blend between Marvel comic illustration and 3D animation style headshot of a beautiful woman with bright red baby bun hair so what exactly did I try to do here well I'm trying to tell my journey that I would like it to use a style that is a combination between a Marvel comic book style as well as 3D animation and if we look at our initial results you can see that while we're still not quite there it's definitely had an impact on the image in fact these are just a little bit too 3D like one thing that I did notice in the old images is that Carla does have some very very subtle Asian features and I mean really really subtle but some of them do remind me of Anime characters so what if we add anime into the mix imagine style like Marvel comic illustration and 3D animation and Anime and Then followed by the rest of our prompt as you can see I've added anime to the prompt and here's what the images look like I would like to point out that we're finally seeing a lot more people of color when we're using prompts that use the word beautiful these women may not look anything like Carla however the latent bias of some of these AI models is clearly being addressed so if we just look at the style we can clearly see that this is much more like the polished very stylized look that we've gotten used to from version 4. this also demonstrates some of the stylistic range of version 5. prompting might be harder now however you can still cook up anything you want as long as you're using the right words in your prompt here are a few more examples if I simply re-roll the very same prompt to be honest I feel quite comfortable just sticking with this style that we have I am however going to make a few adjustments to our prompt nonetheless I've added on white backgrounds to our style because I'm trying to replicate what we did with Carla in version 4. I've also added close-up just to make sure that we get the closest possible portrait shot ever I've changed beautiful woman to attractive European woman simply because our Carla was of European descent and finally I've added the stylized parameter with a value of one thousand in version 4 I basically never used the stylized parameter simply because the images were already extremely stylized to begin with however in version 5 I feel like I may need to add this just in order to get the same kind of style that we had with our old Carla but careful I'm not saying you should always do this this is very specific to this particular example that I'm showing you here here's the first set of images from these I'll keep the top left one and then I'll keep re-rolling until I find a few images that I like and here's the second set of which I'll keep the top left and the bottom right images and now it's time to start creating consistent character in version 5. so just to recap here are the three image references that we're going to use I made sure to pick slightly different styles and different poses I'm going to repeat the very same prompt except that I'm going to add our image references as image prompts imagine and then we start off with our three image prompts followed by the exact same prompt that we just used before technically we could use the image weight parameter in order to control the influence that these image prompts have on our overall prompt however we're going to skip this part for now I'll show you how that works a little later on here's the first set of images if you ignore the character for a moment and simply focus on the style you can clearly see that it's much much closer to the original style that we had so from these I'm going to keep the top left image because what we're doing here is actually just creating more consistent references that we'll use later on we also need slightly different poses again so let's see if we can control that as well notice how I've specified that I want a frontal headshot and that Carla should be looking straight into the camera this gives us this very coherent set of images I'm going to keep the bottom right image because I really like that dual sort of braided bun style next we'll change the prompt one more time I've removed the earlier editions and only changed the shot to headshot from the side here's the new set of images so clearly we already have quite a bit of control over our new character from these I'll keep the bottom left one let's do one more because I want a really good one of her looking to the right I've added looking to the right at the end of The Prompt but I probably could have just re-rolled the image as well as you can see it doesn't really make a big difference so from this set I'll take the bottom left one again okay so now that we have our final set of image references it's time to start using them to create lots of different portrait shots okay so let's start off by rehashing our full prompt first off notice the four New Image references at the beginning the rest of the prompt is still roughly the same the only difference really is the image wave parameter which I've set to a value of 0.75 I mainly did this because I felt the image references were coming through a bit too much however what it also does it allows us to retain more control over the portion that is the text propped so here's what Carla looks like now remember you'll still have to re-roll every now and then and pick out the ones that you really like but the overall consistency is really really great in these so how about we give Carla a smile all I've done is ADD smiling with a half open mouth at the end and what we get is this very set of images right here but we can also make Carla laugh and of course we can make her slightly upset as well remember it's all in the prompt so now it's time to have some real fun I'd really like to show you what version 5 of my journey is capable of we're going to continue working with our core prompt and all of its image references let's create a portrait shot of Carla standing in a Parisian Street yeah Paris that's what I mean by Parisian or whatever just let's look at the prompt so what's going on here I've added wide angle shot of full body this part won't matter just yet but it will play a role a little bit later I've also added standing in a small Parisian Street finally I've set the aspect ratio to 16 to 9 which will for the most part block us into a close-up portrait so here are our four images and I personally think they look absolutely stunning I almost feel like our color corrector even though she looks different is more consistent now than she was in version 4. so let's change things up a little bit and try to get more of her body into the shot I've done two things here first of all I've added with her feet on the Cobblestone Road this is meant to signal to Mid journey to show me more of her legs we'll probably get a cowboy shot in the end but if I don't add this her body will be cut off much further up I've also changed the aspect ratio to four to five to provide more space to actually display her body and here's what the images look like and if I change the aspect ratio to 9 to 16 then we get this set of images let's make another slide change I'd like to change the color of her pants I've added wearing beige colored pants into the prompt however unfortunately in mid Journey version 5 still isn't very good at assigning certain colors to particular objects and elements in the prompt or at least not unless they're like the primary element within the image itself it clearly did pick up on the pants because in some of the earlier images Carla was wearing a skirt so let's try something else I've replaced the earlier change with wearing jacket in blue color wearing pants in white color now this time a journey properly assigns the color of the jacket however the white color is still not given to the actual pants but to her shirt instead however what we've just done with the jacket also works with a black jacket or in New York rather than in Paris notice how her clothing style is also changed to fit into the concept of a woman standing in New York rather than in Paris now I have one last surprise for you you might be wondering why didn't I just use the original image references that I used in version 4. I probably could have spared myself tons of work right well let's have a look at these sort of images that I get if I use these references with the very same prompt you can clearly see the look and style of our original Carla shine through in these images and yes if you wanted to bring back a particular character from your work with version 4 then this is an approach that you could use however trust me this is not going to work unless your text prompt already generates images that are relatively close to the type of style and image that you want just try it out for yourself and you'll see that it's not as simple as you might think it works in this case simply because my text prompt is already pretty good on a standalone basis I actually have one last piece of advice that I would really like to share with you so the process I just showed you will work well for portrait images and reasonably well for images that primarily focus on the character if you want to place your characters into a larger scene with action then you can technically still use this process however you'll have to repeat the very first steps that we did in order to create a full body character concept first otherwise my journey will tend to continue to focus on portrait style shots of your character remember just because version 5 Works slightly differently that doesn't mean that all of your previous knowledge about character design is suddenly completely obsolete in fact I didn't really change all that much about how I did things in this process so if you haven't already watched my full series on character design then check out some of the best ones or just binge on the whole series right away if you learned something today please leave a comment like And subscribe that way you can keep on learning and I'll see you next time take care

Channel: Tokenized AI by Christian Heidorn

Views: 79,658

Keywords: midjourney, midjourney v5, midjourney version 5, midjourney tutorial, midjourney prompts, midjourney consistent character, midjourney consistent style, midjourney character design, midjourney character prompts, midjourney character creation, create character in midjourney, midjourney character art, midjourney text weights, midjourney weights, midjourney multi prompts, consistent character midjourney, midjourney storytelling, graphic novel midjourney, ai character generator

Id: 72r40EahL_s

Length: 13min 39sec (819 seconds)

Published: Sat Mar 25 2023