Advanced Midjourney V6 Guide (Pushing Boundaries of Lifelike Cinematic AI Photography)

in this video I'm bringing you the most complete guide to create stunning photorealistic cinematic AI photos with recently launched mid Journey version 6 their most advanced model you will find a detailed look at the differences between mid Journey V6 and 52 what's new what's improved and how it impacts your AI photography we will unlock the doors of Storytelling with AI utilizing the simplified prompting structure for V6 also in this video I will show you a visual style guide with visual comparison of different keywords and how they're rendered in mid Journey V6 for photography types film stocks camera angles lighting motion fashion materials hand and face expressions emotion and writing text on the images with mid Journey first if you want to try V6 you need to turn it on using slash settings command on Discord V6 is going to be the default model in January but until then we have to turn it on manually I would like to have a structured and empirical comparison between mid Journey version 6 and 52 for AI photography Aesthetics I will analyze and give scores of low medium high improvements for each of these metrics natural language understanding coherence photo realism accuracy of details image prompting for all challenges I use diverse parameters of style roll styliz as well as diverse aspect ratios to minimize biases of certain parameters as much as I can let's start with natural language understanding this challenge is about how well can mid Journey understand my prompt the better the language comprehension the better it becomes at combining and separating things I started with this problem prompt because mid Journey team mentioned that there was some improvements with multiple character rendering that's why I picked this prompt where two characters sit in a cafe with very different looks I can see V6 was able to render my first character with leather jacket where other character has some kind of suit or at least a jacket whereas V5 actually rendered leather jackets for both characters still in V6 both characters look very similar to each other but at least it was able to understand Nuance in the prompt next challenge is about unorthodox or unusual semantics I think here V6 did an okayish job at least portraying whale in a perfect way and somehow adding a dragon looking entity to the image B5 on the other hand morphed whales and dragons to each other and created giant sharks next I picked a prompt with challenging semantics and long word CL clusters with Rich detailed descriptions to see how new model survives this first of all I like that couple is actually not looking at the camera here it gives a natural realistic feeling to the image generally V6 images look much more realistic with prompt understanding I think here both model did a good job except V5 has this stock image Vibe it's too brushed and here I think V5 did better job with Messy background whereas V6 is going into more realistic Direction in the next prompt I throw a bunch of crowded word clusters at the model and see what it does with them I think both model did a good job however V6 was able to understand and reflect body covered with diamonds part better whereas V5 ignored that both model did a good job with tattoos and braids and both models struggled with implants next challenge is about World Knowledge I picked a well-known character from Demon Slayer universe and tried to create a realistic cyberpunk version of him I think both model did a good job but I feel like V6 Tango's face much more similar to original character but also really like red hair on V5 image after this point I decided to go back to more basic prompts with nice twists both models struggled with this prompt it wasn't straightforward for models to understand that I actually want a horse on top of man's shoulders I decided to be more clear with my intention so I constructed a new prompt but result was pretty much same I tried a similar approach with this prompt just for fun and surprisingly V5 did a better job in the first run but with more context and raw mode things got better overall version 6 looks much much better at picking up a lot more nuances in natural language and reflecting those nuances on the photo there is still room for improvement for natural language understanding my Improvement score is high challenge two photorealism how realistic output looks like for this challenge I choose prompts for maximizing photorealism I tried three different prompts and compared next to V5 images and old bo V6 did an amazing job images look much more realistic and detailed especially skin realism is heavily improved airbrushed feeling of V5 is diminished V6 feels much less stock image like and much more natural and lifelike photo realism Improvement score of V6 is high next challenge is accuracy of details how well model renders images with greater details and these are more likely to be accurate and AI defect free for example depiction of hands images of people holding objects guns katanas balls and arrows umbrellas a witch on a broom cigarettes details of Faces in crowded scenes image inside image Etc I started with a basic but challenging prompt correct depiction of a witch on a broom is always challenge for AI I'm really surprised with the Improvement level V6 brings in this challenge I thought this is really promising then I was simply shocked with how V6 renders both umbrella and cigarette so accurately of course it's not perfect but in comparison to V5 it's very significant Improvement here is another challenge with model needs to render art inside art people and details of surroundings V6 didn't really made a better job with accuracy of details but also correctly reflected my wish to have both sitting and standing people in the scene however volleyball scenes are still a big Challenge and final boss for models even for V6 overall my accuracy of details Improvement score for V6 is high last challenge is about image to image prompt performance I started with an original image and copy pasting the image linked to M Journey as an image prompt to understand how loyal new generations are to the original image and quality of resemblance I used the image of Lucia from GTA 6 and I was hoping to have a realistic version of character but I was slightly disappointed I think V5 did much better job here but I didn't have so much time to play with it yet so hopefully I will give you more updates on this in upcoming days let's continue with ultimate hyper realistic cinematic AI photography prompt structure V6 brings some changes to prompting you don't need to use junk words like award-winning 4K 8K anymore since model is much more capable an improved natural language understanding of V6 gives us a massive creative opportunity to be able to Define our desired art in great details we will organize components of our prompt according to three l ERS equals organism molecules and atoms which together creates a fully defined artwork of course we don't have to add every element in every level but one by one we can allow mid journey to fully understand our creative Direction in V6 I will start by defining my art with a simple prompt that only consists of foundational level decisions that looks good but can we Define the artwork more precisely let's add secondary level details like shot type Weather film stock camera and emotion here is the result with this image looks different now still we can do more let's take our character and create a version where ice meets Fire by adding Sunset to our prompt and different camera angle what if we go deeper and make some Atomic decisions for example give our character a futuristic armor or maybe he has a metallic Knight armor from 17th century Victorian age maybe we need to learn his backstory and what changed him with V6 you have more creative control additionally for building your Cinematic Universe in mid Journey quickly you will need shots of landscape people and architecture from your AI film Universe this prompt structure with permutations will run three different jobs with same style and different combinations for Words landscape people and architecture then I can upscale one of these elements and then use strong variation feature of mid journey to create different landscape shots for the same town with same style now I will Deep dive to every element in atomic prompting one by one starting with photography type let's explore what mid Journey V6 brings for wide spectrum of Photography types portrait photography is a style of Photography that primarily focuses on capturing the personality and unique expressions of an individual or group through effective use of lighting backdrops and poses street photography is a candid genre that captures everyday life and Society in public spaces often offering a spontaneous unique perspective on The Human Condition product photography involves taking photographs of products to highlight their features for commercial or advertising purposes typically emphasizing detail Clarity and appeal you can combine this type with various decorating items compositions lighting or with different angles fashion photography specializes in showcasing clothing and other fashion items often involving elaborate staging and lighting to enhance the aesthetic appeal of the designs you can combine this type with magazine photography for young expressive eye-catching photos documentary photography refers to images that a company text in Publications like magazines newspapers or websites intended to illustrate and enhance the story or article that's perfect to create realistic AI photography with raw unfiltered human emotion and moments that communicate the realities of events and in this sense it's very similar to documentary photography and photojournalism Architectural photography focuses on capturing buildings structures and other man-made designs in a way that highlights their beauty symmetry or historical significance we can create both indoor and outdoor architectural photos you can combine your prompts with various color schemes weather types locations Styles movements and you can use G mavic 3 as keyword for drone shots this is a genre focused on capturing expansive natural scenes often emphasizing the aesthetic Beauty and Grandeur of the environment you can combine your landscape photos with various locations objects like Lighthouse house Caravan color schemes weather types you can add G mavic 3 for drone shots or use various keywords like soft dreamy Landscapes surreal architectural landscap capes National Geographic photo for elevating your artwork food photography specializes in capturing food in an appetizing and aesthetically pleasing manner often used in cookbooks magazines advertisements and blogs Still Life photography is a genre focused on capturing inanimate subjects often a group of objects with creative Arrangement lighting and composition to evoke a sense of Simplicity Elegance or Intrigue low-fi photography is a style that Embraces imperfections using lower Quality Equipment or techniques such as film grain blur vignetting and light leaks to create a nostalgic atmospheric and often abstract aesthetic fantasy photography creates theal imaginative scenes often involving elaborate costumes props and digital manipulations to evoke a sense of magic and wonder macro photography involves close-up shots that capture small subjects or details at a high magnification often revealing features unseen by the naked eye Whimsical photography creates playful quirky and imaginative images that evoke a sense of fun magic and wonder you can combine these photos with stylized weird Duo and as location Peak Urban exploration location for truly Rising outof the Box results highkey photography uses unusually bright lighting to reduce or completely remove dark shadows in the image this technique results in photos that are predominantly filled with light tones and white creating a mood that's often associated with positivity innocence or Simplicity lowkey photography emphasizes contrast by using strong shadows and minimal lighting to create dramatic moodier images with a more limited tonal range this genre aims to capture images of animals in their natural habitats often emphasizing the beauty power and diversity of the Natural World Sports Photography involves capturing Dynamic and high energy Moments in Sports showcasing the drama intensity and athleticism of competitors you can combine this type of Photography prompts with keywords like Phantom high-speed camera motion blur Dynamic movement aerial photography captures images from an elevated perspective often from a drone or aircraft providing a unique bird's eye view of Landscapes cities or events you can also mention keywords gy Phantom 4 Pro drone infrared photography captures light beyond the visible spectrum it's used for Unique surreal landscape or artistic phot photography unfortunately mentioning infrared photography is not enough to achieve these results you need to add Kodak aeroc Chrome as camera keyword to your prompts for excellent infrared results night photography involves capturing images in low light conditions often using long exposures to reveal details invisible to the naked eye such as stars or City Lights this genre involves capturing images beneath the surface of the water often revealing the beauty diversity and mystery of marine life this genre emphasizes Simplicity using a minimal number of elements to create a clean and uncluttered image that relies on Simplicity of composition color and theme historical photography focuses on capturing images of historical events artifacts or locations often aiming to preserve or document past eras next element in our prompt structure is subject and action it's impossible to think cinematic AI photos without a subject and character's action or motion that's why it's in the fundamental level for us for example our prompt can be as simple as this and we would already get a good results however if we continue to molecular level then we can have more creative control for example examp Le we can add motion keywords cinematic film stocks for great action scenes and results change quite a lot these keywords make an impact notice how we combined subject and action with motion keywords and a suitable photography type for photography type you can choose a variety of words like panning photo long exposure photo high-speed photography or you can go to more of a cinematic Direction with these words cinematic action scene when it comes to action and motion I collected all effective keywords I like to use in my prompts in this visual style guide you can pause the video for a deeper look to see visual comparison of motion keywords and how they impact the image my personal favorites are motion blur motion Trail and Phantom high-speed camera what's yours let me know in the comments down below you can of of course combine your prompts with style Touch of action film directors who are masters of capturing motion I combined visual catalog of famous action directors for you I really liked how adding Ridley Scott and Zack Snider to my prompt completely changed Vibe of the image lastly long exposure is great to indicate Motion in the photography especially when combined with light for example you can also use this pront structure now I will show you how to control your camera mid Journey version 6 is much more capable of distinguishing keywords for different camera angles and shot types this gives us an incredible creative control for our AI photos all you have to do is adding name of the shot type to your prompt let me show you Consolidated keywords for camera controls keep in mind if you struggle with any of these words are not being reflected on the image try applying a higher weight for your shot type keywords within your prompts I level low angle or photo from below high angle or photo from above extreme low angle extreme high angle wide angle or wide shot next on the list side profile shot from behind or back view closeup full shot or full body low angle extreme closeup or macro aerial shot or Birds I view underwater shot over the shoulder shot establishing shot and remember you can always start with close up of your subject and Achieve to full body shot with panning and zoom out features once they're also available for V6 Master shot touch angle of Center rule of thirds candid shot silhouette shot remember you can use pan feature to create off center shots additionally it may be also useful to control composition with visual balance in mid Journey asymmetrical symmetrical if you prefer to go to a more retro Direction there are few ways to do it you can simply add year decade era or analog photography as keyword to the beginning of your prompt and that will alone bring that retro look you're looking for you can always combine this with cinematic effect by adding cinematic keyword to the beginning of your prompt also try the 40s 50s 7s or any era that you love now let's look at how we can advance retro look of our AI photos by adding more details in molecular level in my tests I realized that camera makes a slight difference in final images but real impact comes from film stocks with retro features here you can see visual comparison of different film stocks with strong retro Vibes feel free to pause and take a look longer mour Community loves Kodak ocome Kodak portra or ector as well as AA Vista additional to those I also like ctil film stocks because they have great cinematic visual language I discovered some new favorites like Fujifilm we continue with camera element in the molecular level lot of people are confused about using cameras in their prompts main concern is they don't know if it makes any impact on the final image I compared photography cameras with each other and put to cinematic ones to a different comparison I was positively surprised by Panasonic Lumix output whereas my other favorites like Sony Alpha 3 and Hasselblad always deliver great results now let's take a look at the Cinematic cameras as you can see change on the final image depending on camera choice is subtle but there is definitely an impact on color and style depending on preferred camera type I really like cinematic tones of a real Alexa a riflex as well as black magic and a Max MSM in the end there is not a big difference between each other light and photography serves as the fundamental medium that illuminates subjects determining exposure mood and the overall visual aesthetic of an image without it capturing visual information would be impossible and it certainly changes to emotional impact of an AI photo I want to start with few basic tips and keywords you can use in your prompts you can simply improve lighting of your images selecting Shadow Photography as photography type you can use Ray pr90 as choice of your camera because it does a great job capturing Shadow sharp contrast and Vivid colors and I can recommend few extra keywords like CHX toos of light and Shadow or simply Shadow play to improve mood on your photos now let's Deep dive into Atomic lighting methods in mid Journey mid Journey version 6 bring brings significant updates to natural language understanding and allows us to control relationship between light source and subject so that we can create very specific compositions now let's observe how keywords for natural light sources being rendered in mid Journey V6 starting with time of the day here is a visual comparison for Sunrise morning afternoon golden hour blue eyes hour Sunset evening or night Moonlight next is weather conditions Sunny partly cloudy rainy snowy overcast foggy hazy clear sky snowstorm dark clouds drizzle Heavy Rain hail storm crepuscular Rays God Rays now let's take a look at keyword comparison of the artificial lighting sources stay medium lights candle light torch bioluminescent lighting neon light or neon tubes fireworks Studio lighting hallogen bulbs fluorescent tubes here we will briefly look at the directional lighting techniques and how mid Journey V6 can apply them to final images lowkey lighting highkey lighting split lighting butterfly lighting top light side lights chyos cural lighting Rim lighting Rim brand lighting by the way if you're enjoying this video so far please subscribe to the channel for my upcoming mid Journey tutorials V6 takes fashion and style to the next level and allows us to describe specific combinations of outfit like top with blouse shirt t-shirt sweater or bottoms like pants skirt shorts jeans or dresses Footwear or accessories you can describe Brands colors physical shapes Styles or designers of these individual garments and make a look out of it fashion element can be added to your prompts simply by mentioning a specific garment style or can be break down into smaller components according to Atomic prompting look how version five struggles with dog patterns versus version 6 is able to understand we want a yellow jacket with dog patterns on it now let's take a look at some of distinctive fashion styles we can replicate on Mid journey to do this you need to include name of the style with words outfit fashion or Aesthetics boho Punk golf steampunk Kawai Harajuku preppy flaer hippie minimalist eclectic tewar aveng guard hot Couture High fashion Bohemian hip hop Urban athletic street wear when photographing fashion the visual distinctiveness of Brands and styles plays a crucial role in making a statement here's a list of some iconic fashion brands and styles that are renowned for their unique visual characteristics Gucci Chanel Louis Vuitton Burberry Versace Christian loton Jeremy Scott for mosino Jean Paul Golder Alexander McQueen and here we will compare some keyword comparison for materials reflective materials translucent fur cotton bamboo velvet lace silk wool linen polyester leather denim saton nylon spandex t neoprene improved language understanding brings significant Improvement for face expressions certain emotions were not really well reflected on version 52 but now there is Improvement there are two ways you can add emotion to your photos first you can describe your character's mood while describing your subject second you can Infuse emotion into the entire photo by simply incorporating it as a separate element this approach has the advantage of creating very dark compositions and injecting a mood into the entire image not just the subject let me also give you comparison catalog of the emotion keywords with variety of face expressions yes hands can tell stories they can unveil your emotional state amplifies emotion hands can add life to your photos making them more expressive a simple prompting tip to integrate hands in your photos is combining emotion element together with hands if you want to be more expressive this is indeed possible you can be more specific with position of your subject's hands try this template and see examples one of the latest features coming with V6 is being able to write text on the images this will be especially helpful for people who make lot of product photography stuff with mid journey to achieve this only thing you need to do is to add your choice of text in quotation marks to your prompts hopefully this video was truly helpful for you to understand impact of different keywords in your mid Journey images in version 6 if so give it a thumbs up and subscribe to my channel for more awesome tutorials if you want to learn more about creating AI photos and AI films click here

Channel: Cyberjungle

Views: 134,877

Keywords: midjourney, midjourney v6, midjourney tutorial, midjourney prompts, midjourney version 6, midjourney ai tutorial, magnific ai, krea ai, generative art, prompt engineering, how to use midjourney, midjourney face swap, midjourney consistent character, midjourney product photography, runwayml motion brush, midjourney v5.2 zoom out, midjourney 5.2 zoom, stable diffusion xl, leonardo ai, pika labs, midjourney 6, topaz photo ai, runway gen 2, pika ai, elevenlabs

Id: Zl_4V0ks7CE

Length: 30min 0sec (1800 seconds)

Published: Sat Dec 23 2023