Consistent Characters in Midjourney v5

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
so mid Journey 5 is out and there's some new tricks to creating consistent characters for comic books story books and even stock photography so I'm going to show you how I created this short story all with AI in just a few hours in order to get started with me Journey you'll need to set up a Discord account take a look at the start of this video here for steps and how to get started quickly the first step in the process is creating the original inspiration for the characters that you want to create so I just asked Chachi PT to write me a story and then to describe the character's physical attributes Chachi PT gave me full scene breakdown and individual descriptions for an Auburn haired heroine called Leela and her AI companion named Sage I'm going to leave a reference to this reference sheet which you can find any different artists that you'd like to use as your inspiration using the same artists across all your prompts is key for consistency I'm using Lois van Barrel a really amazing artist that's done some work for Horizon zero Dawn mid-journey works by crawling millions of artists and images all over the web over the last couple of years is this plagiarism or is this the same as just using any other Artist as a form of inspiration this is hotly debated and outside of the scope of this video for now let's start with the prompt for Leela as an Auburn haired 20 year old girl slender frame white cotton t-shirt linen pants here are the outputs we get if you don't see exactly what you want you can modify the prompt or you can hit the reroll button to see if there's other variations you might like when these images are generated a seed number is used as the basis for the generation the randomization is the reason why every time you put in the same prompt twice you don't get the same image so when you find an image that you like we copied the seed so that model uses the same randomization to generate the outcome and hence we end up with a more consistent output to get the seed in mid Journey 5 it's different from version 4. to do that we go to the reaction button and if you've not used it before you need to type in envelope to find the envelope Emoji this tells me journey to give us the job number and the seed which will be sent back to you in the chat now we can take the image that we like as a basis and paste that into our prompt along with the seed I found the best results by pasting in the image that I like along with the original instructions put any variation that I want in at the end and then of course adding the seed [Music] okay so this one is pretty consistent I like it now what we want to do is create various different emotions and create a character sheet so by doing that I just add in frightened expression or whatever additional expression that I want at the end of The Prompt and keep rolling through those until I get what I want so now that I have a consistent lead protagonist I want to make my way through the script that Chachi PT has given me to pick out the various different scenes and characters that I need to recreate all the time I use the same artist style and this gives a real sense of consistency all the way through the imagery if I want to get detailed or far away shots of Lila what I do is I actually drop out the detailed information about what she looks like I just put in Auburn hair girl I don't even add in a seed in order to get the consistent images so you can get these far away shots which look pretty much close to what she looks like but we don't have the facial details that we need to worry about pulling this all together let's see what's possible with just a few hours of prompting time in the eco-friendly city of Solara a young horticulturist named Lyla was known for her Innovative and sustainable gardening techniques one day while researching ways to enhance her City's green spaces she stumbled upon an old deactivated AI module in a forgotten corner of the public library being able to create a rough story visuals music and imagery would one person in just a few hours is still incomprehensible to me the positive and negative effects this will have on artists musicians and writers is still hard to Fathom and we still are in the early generations of what's possible on one hand this opens up massive creativity and on the other there are deep sociological economic and philosophical hurdles to surmount as we move forward if you're interested in keeping your finger on the pulse subscribe for more of this content I'll do my best to keep you up to date
Info
Channel: Rob Shocks
Views: 24,389
Rating: undefined out of 5
Keywords: ai art, midjourney, midjourney v5 prompts, chat gpt, eleven labs, mubert, midjourney tutorial, midjourney ai
Id: vlqDG6PVz4w
Channel Id: undefined
Length: 4min 45sec (285 seconds)
Published: Tue Apr 11 2023
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.