Character Consistency made easy in Midjourney (Tutorial) | Photography Style + Prompt

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
creating consistent characters in mid-journey can be tricky without using celebrities names multiple image links or using mid-journey seeds but I still found a way and the only thing you'll need is patience the first step is the prompt setup you want to start with the description of your subject from describing the character to location and film type and the important part now to achieve consistency lies in the second part image split and describing different angles after you get your grid you can choose the image you like the most and then we go on to step two variations this is the part where you need patience I use both variation options to become more flexible strong variations give you more options in terms of poses and slight variations help you find a fitting pair in terms of consistency [Music] voila let's try out a few more options just to see how it changes in terms of subject description here I'm trying out a more editorial kind of look the second part of the prompt Remains the Same then I'm doing the variations and finding the best pairing of characters it's also doing fairly well with keeping clothing or jewelry consistent here's an example for wedding photography and we can see the dress remains pretty similar and also the hair and face features match very well you can now repeat this process with all your prompts describe your subject and include the Split Image and different angles you won't always find the perfect pairing right away that's why I said that you'll need patience to repeat the steps of variations as often as needed now on to step three going the extra mile changing aspect ratios you can already try and do a zoom out in mid-journey and see if this setup is already good enough but I personally like to take the images over to photoshop beta and help extend them with generative fill for that I just simply cut out the images do the selection I want to fill in and let Photoshop do its magic find the most fitting fill and repeat for the other part of the image and find a good layout for them to go together nicely let's try it for another example without previously zooming out in mid-journey and just directly going to use generative fill in Photoshop beta I'm usually going for horizontal images but you of course also can go for portrait mode here I'm expanding not just to one side but actually both and the results are absolutely perfect thanks [Music] here are the original outputs from mid-journey on the left and the expansions of the images done with generative fill in terms of subjects you're obviously very flexible and you can dream up every scenario here the only limitations now could probably be your imagination but in terms of tools we have already all we need and that's it two main steps and one optional third step to go the extra mile but I think it's one of the easiest ways at the moment to achieve consistent characters in mid-journey and you don't have to go all crazy on parameters and image links and whatnot have fun creating and if you're sharing your results don't forget to tag me on Twitter or Instagram so I can have a look at your work
Info
Channel: Julie W. Design
Views: 10,358
Rating: undefined out of 5
Keywords: Midjourney, character, consistency, photography, cinematic, tutorial, tips, prompts
Id: vTJkBfmnf-8
Channel Id: undefined
Length: 4min 32sec (272 seconds)
Published: Sun Aug 13 2023
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.