Midjourney for Architects: The Ultimate Workflow for Design and Photorealistic Renders

if you're like most Architects you probably wished  for a magic wand to instantly create perfect   renders well I can't promise you a wand but I do  have something better at the moment AI tools AI   will transform our rendering process in the blink  of an eye in the next few months and years so the   sooner we start learning it the less likely it  is to replace our jobs just kidding it's already   started replacing our jobs now it won't be long  before your program of choice has a text option   that will let you generate whatever you wish for  in seconds but until that day comes we still have   to overcome a few challenges with these tools such  as copyright position and originality so let me   show you how far mid Journey has come in the past  few months how professional Architects and artists   are using it in a step-by-step tutorial on how you  can use it for your architecture projects a few   months ago these were my results when creating  architecture imagery from mid-journey although   it looked pretty cool it wasn't nothing to be  completely afraid of or excited about just yet   however since the release of mid-journey version  5 and the future release is coming the next few   months the images passed from looking like this to  looking like this there's no longer a mid-journey   look now I can achieve different styles and be  really hard to tell from a real photo the level   of detail in realism for a program in its fifth  version is crazy and I can't even begin to imagine   what the future versions will bring many people  are already using it along other AI tools like   chat GPT to optimize their workflows and create  entire businesses by it if you would like to see a   tutorial on how I use chat GPT for my architecture  workflow comment down below some Architects and   creatives are already working with me journey in  their day-to-day and the results are astonishing   like for example Jose Sanchez architect theorist  and game developer from Michigan he started   working on mid-journey in creating iterations  of different Suburban house projects these are   some of his results the surprising thing about  this is the level of detail in some images and   how how real these images look there's also  Andrew Cutlass principal of the art studio   Matty's design who has been creating mesmerizing  architectural images and exploring with infinite   amount of words and prompts to generate different  kinds of results through his Instagram you can see   the variety of the images and settings he has  created to such detail and position that you   would think is something generated by something  different than mid Journey his work with fabrics   and soft materials serve as an inspiration and  reference for your own architecture projects now   there are other creatives like for example Nick  floats who have been exploring with chat DBT and   mid-journey create additive prompting where  basically he tells chat GPT how he wants his   prompts structured what are the categories and  parameters he needs and how many he needs then   Chad gbt creates infinite amount of prompts and  you can modify them according to your own needs   we'll touch more on how he does this later in  the video now this all sounds really fun but   there is still one issue that that mid Journey  hasn't solved yet which is copyright mid Journey   has been facing criticism for using images of  other people without their consent when the AI   generates images it sometimes uses copyrighted  works as inputs or creates images that closely   resemble existing copyrighted material so until  the problem is solved as creators there are some   things we can do to use this in an ethically and  legally correct way first just give credit to   words do if you are using mid-journey and you're  using prompts that reference a specific artist or   Creator say that you are and tag the artist in  your work second try to use your own projects   and images as inspiration you want to learn how  let's take a look so one of the features you can   use for mid-journey is the blend command where you  can blend two or more images to create a new one   let me show you how to use it so first of all I'm  going to open my Discord Channel which is where   mid-journey does all of its magic now I'm going  to look for the images is that I want to blend in   together so first of all I want to open this image  which if you're a faithful follower of the channel   you know that we created it last year I'm going to  take this image and I'm going to drag it inside of   our chat box now I'm going to click Send and hit  enter now I'm going to search for a second image   let's say this is going to be our second image  right here they're both pretty similar in Styles   and I'm going to drag this one inside as well  now I'm going to hit enter so first I can type   in Imagine and then I can just drag my first image  hit space and drag my second image in and now I   can just click on enter here I'm going to wait  for Mid journey to make its own interpretation   of both of these images another option that I  can also use to blend these two images is type   slash click on blend and upload these two images  separately and hit enter now if we take a look our   first grid of images has already been generated  which is this one right here as you can see it is   a faithful mix of both images which we have a red  house you know some chairs in the foreground and   we also have some trees in the foreground but we  have a more night type of now what happens if you   try to blend three images so I'm going to upload  another image that I did which is this image right   here which again is from the same video that we  did last time and I'm going to upload it right   here and I'm going to drag in this first image hit  space drag into the second one hit space and drag   in the third one now this is the other image that  mid Journey generated for us which looks pretty   cool it looks very interesting if I was creating  sort of a mood board for my next architecture   project and I know that most of my my projects  are in a rural area and it's only a cabin house   type of things this would be a cool first step to  start my next idea so for example you know any of   these houses maybe I don't know I like all of them  a lot but let's say this third one right here is   an option that I would literally choose so from  here on what I can do is just click on upscale the   third image and in a matter of seconds mid journey  is going to enlarge this image make it bigger and   if we open this in the browser we can see that  this is a really nice image it doesn't look very   realistic but as a concept sketch initial design  phase it looks really interesting another feature   you can also use is the describe command where you  can upload an image and mid-journey will give you   four example prompts from what it detects you  could now work off of these prompts to create   a variation of or a new image let me show you  how so first I'm going to type slash describe   and I'm going to upload the image that I want  my journey to describe for me now as soon as I   have my image loaded I'm going to click on  enter and upload it now as you can see mid   Journey gave me four different prompts and it's  just trying to describe what they're seeing in   the image so the first one is for example a 3D  rendering of landscape design in the style of a   post-apocalyptic Landscapes orange and gray Oscar  Niemeyer 32k UHD missed environmental portraiture   so what this feature allows us is to grab specific  words that mid-journey found interesting inside of   the images and use this prompt to create new  images out of this first prompt right here I   want mid-journey to create certain very Asians  and I'm going to generate variations for these   four images to see which one gets closer to the  original image okay so now our images have been   generated and let's take a look at what mid  Journey gave us the first one was this one the   one where I changed the parameters from Orange  to red and as you can see although they look   pretty different from the original prompt it looks  very gray and red so this one maybe is not going   to be my first option now for the second one we  have different interesting options but it but it   interpreted the original design as train tracks  that is which again is not part of the original   design it's more of a pathway so this is not going  to be of my style either now for the third option   we have a more mystical post-post-apocalyptic uh  kind of image which obviously is very far off and   for the last one we have this one which again goes  back to the trains now you can say you know this   doesn't really work but what I can do is generate  variations and modify specific parameters so for   example I'm going to take the second prompt  and click on generate variations now instead   of train I'm going to type in pathway an orange  pathway on a track in the forest in the style of   ethereal sculptures and I'm going to erase this  and I'm going to click on submit and see what   are the results these were the results that I got  it didn't really resemble the original image but   I really like the different options that it's  giving me so for example this first image looks   very realistic it looks like an architecture  rendering and for the other ones I think they   just they're just different from what I'm looking  for now I want to take you on a step-by-step guide   on how to use mid-journey in two impactful ways  the first one is using only text prompts using an   additive prompting technique that the Creator Nick  floats has detailed in his Twitter account the   first thing we have to do is type inside chat GPT  the following just create a table that breaks down   an interior architecture photograph composition  into the following key elements where each of   these key elements is a column like composition  camera angle style room High focal point textures   detail color palette brand lighting location time  of day mood and architecture now this is how Nick   categorized an interior render nonetheless you  can add the categories you consider necessary or   eliminate the ones you don't want for example I  don't really want to use the brand architecture   or detail instead I'm going to write reference  Architects and typologies now I'm going to tell   it to fill the table with 10 rows of data now you  can copy and paste this data into mid journey to   create a specific prompt how cool is this so  these were the results that we got for our   first option which to be honest they look pretty  cool they don't look very Frank gary-ish which is   our reference architect but they still look pretty  interesting and we can you know repeat this prompt   and maybe change a few words here and there so  they can be a little bit more important what I'm   going to do is I'm going to generate a variation  on the first one but what I'm going to do now is   click on Frank Gary architect separate these with  two dots after the two dots I'm going to type   in 10 which is going to indicate mid-journey the  importance that I'm giving to this specific word   or this specific type of words and I'm going to  click on submit to see the the results the results   are in and this was the grid that it generated of  course here it tried to actually put Frank Gary   inside of the image but I really like the woodwork  that it did around it here we have some different   options I think you know these all look pretty  interesting but overall I'm really happy with   the results you can also do this for exterior  renders just adjust the original prompt in chat   DBT and direct it towards external parameters  with this method you are basically testing words   and how AI interprets them that way when you're  working on a specific project you can use these   set of words in the initial stages of design to  sketch out different ideas and start from there   now let's say you want to be more specific with  your images and get specific reference images   well here I recommend the second option which  is missing image and text prompts for example   you can start with a base image and start adding  parameters to change specific details this will   take some trial and error but again this is the  exciting thing about AI if the results were too   specific we wouldn't have any crazy inspiration  now after many tasks with different methods these   are some of the images I have created with mid  Journey some of them are far from realistic and   others are incredibly precise and uplifting I  recommend testing these tools as it starts to   get familiar with them I'm sure you don't want  to sound like you're old teachers when they said   rendering is for the devil we only use the Old  Reliable pencil and paper of course this doesn't   mean that you have to leave your own tools behind  but learning new ones won't hurt and can prepare   you for the coming years where you will work hand  in hand with AI to create architecture designs are   you using mid Journey yet does this all sound  scary and gloomy to you or are you excited for   what's to come let's talk in the comments section  down below I would love to hear your thoughts and   ideas also go give some love to the creators  and artists mentioned in this video there will   be links to their websites down below now  if you want to learn more about AI how to   communicate your ideas as Architects and general  visualization tips and tools subscribe to the   channel and watch our other videos thank you guys  for watching and I will see you in the next video foreign

Channel: Show It Better

Views: 272,558

Keywords: show it better, architecture tutorials, photoshop for architects, architecture videos, tutoriales de arquitectura

Id: UVNP-4U8jxY

Length: 12min 37sec (757 seconds)

Published: Wed Apr 19 2023