Prompting In V6
The official announcement says that prompting in V6 differs significantly from how we learned to communicate with previous Midjourney models. Up to a point where you will ”need 'relearn' how to prompt.”
Haute-couture advertising campaign photographed by Erik Madigan Heck. Two models wearing Comme des Garcons Avant-Garde costume. Mongol steppe in background. Northern lights in sky --v 6.0
Haute-couture advertising campaign photographed by Erik Madigan Heck. Two models wearing Comme des Garcons Avant-Garde costume. Mongol steppe in background. Northern lights in sky --v 5.2
Word Order + Optimal Prompts
To start with prompting, let’s experiment with the word order to see its influence on the outcome. For this test, I used four variants of the same prompt with different word/token order. First, I applied the same --seed to each generation to get the results as similar as possible:
Queen Of Valkyries by Aaron Horkey --v 6.0
Queen Of Valkyries by Aaron Horkey --v 5.2
Queen Of Valkyries. Aaron Horkey's illustration --v 6.0
Queen Of Valkyries. Aaron Horkey's illustration --v 5.2
Queen Of Valkyries. Illustration by Aaron Horkey --v 6.0
Queen Of Valkyries. Illustration by Aaron Horkey --v 5.2
Aaron Horkey's illustration depicting Queen Of Valkyries --v 6.0
Aaron Horkey's illustration depicting Queen Of Valkyries --v 5.2
And even though there are slight differences between the four images, the overall result didn’t change that much from variant to variant. However, this happened when I removed the --seed and let each generation be unique:
Queen Of Valkyries by Aaron Horkey --v 6.0
Queen Of Valkyries by Aaron Horkey --v 5.2
Queen Of Valkyries. Aaron Horkey's illustration --v 6.0
Queen Of Valkyries. Aaron Horkey's illustration --v 5.2
Queen Of Valkyries. Illustration by Aaron Horkey --v 6.0
Queen Of Valkyries. Illustration by Aaron Horkey --v 5.2
Aaron Horkey's illustration depicting Queen Of Valkyries --v 6.0
Aaron Horkey's illustration depicting Queen Of Valkyries --v 5.2
But the true importance of the word order (and the benefit of using Optimal Prompts) becomes obvious with less powerful Midjourney styles than that of Aaron Horkey.
Here are just a couple of examples of how the right wording can change the outcome drastically:
cyberpunk character by Sophie Taeuber-Arp --v 6.0
cyberpunk character by Sophie Taeuber-Arp --v 5.2
Sophie Taeuber-Arp's painting depicting cyberpunk character --v 6.0
Sophie Taeuber-Arp's painting depicting cyberpunk character --v 5.2
cute Mainecoon cat by Joost Swarte --v 6.0
cute Mainecoon cat by Joost Swarte --v 5.2
Joost Swarte's illustration depicting cute Mainecoon cat --v 6.0
Joost Swarte's illustration depicting cute Mainecoon cat --v 5.2
Prompt Complexity
When exploring Midjourney's capability to understand your prompt, an important question to answer is: how complex can we make our prompt before MJ ”breaks” (i.e., stops reacting to new additions to the prompt or “loses” its previous parts).
Let's start with a simple prompt, and gradually add new tokens to it, while keeping the --seed paramter fixed for consistency.
redhead boy wearing black hoodie --v 6.0
redhead boy wearing black hoodie --v 5.2
redhead boy wearing black hoodie. Photograph by Loretta Lux --v 6.0
redhead boy wearing black hoodie. Photograph by Loretta Lux --v 5.2
redhead boy wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux --v 6.0
redhead boy wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux --v 5.2
redhead boy wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux shot with green neon light --v 6.0
redhead boy wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux shot with green neon light --v 5.2
redhead boy wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux shot with green neon light. Night city in background --v 6.0
redhead boy wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux shot with green neon light. Night city in background --v 5.2
redhead boy wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux shot with green neon light. Night city in background. Red moon in orange sky --v 6.0
redhead boy wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux shot with green neon light. Night city in background. Red moon in orange sky --v 5.2
redhead boy in sunglasses wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux shot with green neon light. Night city in background. Red moon in orange sky --v 6.0
redhead boy in sunglasses wearing black hoodie with cyan flower pattern. Photograph by Loretta Lux shot with green neon light. Night city in background. Red moon in orange sky --v 5.2
Midjourney V6 made it to an impressive seven steps before parts of the prompt started “disappearing“. While V5.2 started losing tokens one step earlier.
Obviously, results can vary depending on different situations and prompts. However, in most of my experiments, the predecessor model tended to reach its limit at around 6–7 steps before ”breaking“, whereas the new model consistently managed to handle a couple of steps more before encountering similar issues.
Abstract Concepts
For the concluding test of this chapter, let’s see how both models—and their Raw Modes—interpret more abstract (and even nonsensical!) prompts.
multi-dimensional petrichor --v 6.0
multi-dimensional petrichor --v 5.2
multi-dimensional petrichor --v 6.0 --style raw
multi-dimensional petrichor --v 5.2 --style raw
intergripped vision of multiplied adjustment misturned via trivoxel flatmension --v 6.0
intergripped vision of multiplied adjustment misturned via trivoxel flatmension --v 5.2
intergripped vision of multiplied adjustment misturned via trivoxel flatmension --v 6.0 --style raw
intergripped vision of multiplied adjustment misturned via trivoxel flatmension --v 5.2 --style raw
Despite the dramatic official announcement from the Midjourney team, the prompting strategies I developed with earlier models proved highly effective in the new version. Nonetheless, V6’s advancements—in terms of quality, variability, and the AI's improved understanding of prompts—are unmistakable.
Image Prompting
Image Prompts, allowing the use of existing images as a basis for Midjourney to build upon, have been a part of the platform since V3. With each new model, this functionality has grown more powerful. So, what changes does V6 bring to Image Prompting?
dark and dramatic close-up manga comic-strip illustration by Frank Miller. Red and blue hues --iw 0.75 --stylize 75 --v 6.0
dark and dramatic close-up manga comic-strip illustration by Frank Miller. Red and blue hues --iw 0.75 --stylize 75 --v 5.2
Here to present to you the capabilities of V6—Francis D., my friend, muse, and the ideal subject for Midjourney’s Image Prompts!
A major issue I've found with Image Prompts in earlier Midjourney versions was their limited ability to transform photorealistic images into something else. For instance, using a photographic portrait as a source image to create a non-photographic result often didn't fully work. The outcome would still lean towards photorealism.
Let's see if V6 brings any improvement in this area!
as painting by Richard Edward Miller --v 6.0
as painting by Richard Edward Miller --v 5.2
as illustrated portrait drawn by Kim Jung Gi --v 6.0
as illustrated portrait drawn by Kim Jung Gi --v 5.2
as map of pirate's island --iw 0.75 --stylize 55 --v 6.0
as map of pirate's island --iw 0.75 --stylize 55 --v 5.2
as Op-art by Mark Wallinger --stylize 75 --v 6.0
as Op-art by Mark Wallinger --stylize 75 --v 5.2
thermographic photograph. Portrait --iw 0.75 --stylize 50 --v 6.0
thermographic photograph. Portrait --iw 0.75 --stylize 50 --v 5.2
black-and-white distorted portrait as abstract fragmented composition of fractured pieces by Jannis Kounellis --stylize 75 --v 6.0
black-and-white distorted portrait as abstract fragmented composition of fractured pieces by Jannis Kounellis --stylize 75 --v 5.2
Another notable aspect of Image Prompts in V6 is the model's boldness in completely transforming the source image to meet the prompt's requirements. How about turning Francis D. into a lion? Or an owl? Or Stan Winston’s monster?
king of lions by Simen Johan --v 6.0
king of lions by Simen Johan --v 5.2
as Johanna Basford's illustration. Portrait of old wise owl --stylize 75 --v 6.0
as Johanna Basford's illustration. Portrait of old wise owl --stylize 75 --v 5.2
by Stan Winston --stylize 75 --v 6.0 --style raw
by Stan Winston --stylize 75 --v 5.2 --style raw
The reason why Francis is such a great model for Image Prompts is the combination of his standout facial features and the dramatic lighting that amplifies them. That’s the best recipe to help Midjourney “recognize,” or “read” a face to then reproduce it. But how about less obvious photographs? Or not photographs at all?
Andrei Kovalev — E.T. Mask by Catherine Zolotukhina (2007)
Vincent van Gogh — Self-portrait (1887)
Hilma af Klint — Group IV, No.3, Youth, from The Ten Largest series (1910)
Egon Schiele — Self-portrait (1910)
For this experiment, I began with a text part of the prompt that complemented the contents of the image. Then, I proceeded with prompts designed to transform the source picture into something entirely different:
alien with trigonometrical head --v 6.0
alien with trigonometrical head --v 5.2
photograph by Rachel Maclean --iw 1.2 --stylize 75 --v 6.0
photograph by Rachel Maclean --iw 1.2 --stylize 75 --v 5.2
Google Maps satellite footage of colorful landscape --v 6.0
Google Maps satellite footage of colorful landscape --v 5.2
in style of The Tatami Galaxy anime --iw 1.2 --v 6.0
in style of The Tatami Galaxy anime --iw 1.2 --v 5.2
Throughout my experiments, an intriguing observation emerged regarding the text component of single-picture Image Prompts. When the prompt includes the image of a well-known individual, adding this person's name tends to “weaken” the impact of the other elements in the prompt. It appears that Midjourney's preconceived notion of the individual forces the AI to favor more traditional, real-life representations of that person.
court photograph of Pablo Escobar. Prison mugshot attached to criminal case file --v 6.0
court photograph of Pablo Escobar. Prison mugshot attached to criminal case file --v 5.2
court photograph. Prison mugshot attached to criminal case file --v 6.0
court photograph. Prison mugshot attached to criminal case file --v 5.2
Alberto Korda — Che Guevara, Guerrillero Heroico (1960)
Che Guevara as flower revolutionary by Miles Aldridge --stylize 75 --v 6.0
Che Guevara as flower revolutionary by Miles Aldridge --stylize 75 --v 5.2
flower revolutionary by Miles Aldridge --stylize 75 --v 6.0
flower revolutionary by Miles Aldridge --stylize 75 --v 5.2
Image Weight
Talking about Image Prompts, it's impossible not to mention the Image Weight, or --iw parameter. This setting in V6, ranging from 0.1 to 3 (to 2 in V5.2), dictates how closely Midjourney adheres to your source image in the generation process.
For this test, I used some photographs from my own archive:
“Faces of A Picolo“ series (2012)
Marco Casolanetti for Simple Wine News (2016)
Using these portraits as source pictures, let's explore how changing the --iw value affects Image Prompts in V6 in comparison to V5.2:
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.1 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.1 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.2 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.2 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.3 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.3 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.4 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.4 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.5 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.5 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.6 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.6 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.7 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.7 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.8 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.8 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.9 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 0.9 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.0 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.0 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.1 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.1 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.2 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.2 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.3 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.3 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.4 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.4 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.5 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.5 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.6 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.6 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.7 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.7 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.8 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.8 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.9 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 1.9 --v 5.2
girl with whirlwind of threads and wires around her head by Bert Stern --iw 2.0 --v 6.0
girl with whirlwind of threads and wires around her head by Bert Stern --iw 2.0 --v 5.2
Tsutomu Nihei's illustration --iw 0.1 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.1 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 0.2 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.2 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 0.3 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.3 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 0.4 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.4 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 0.5 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.5 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 0.6 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.6 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 0.7 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.7 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 0.8 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.8 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 0.9 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 0.9 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.0 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.0 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.1 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.1 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.2 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.2 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.3 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.3 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.4 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.4 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.5 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.5 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.6 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.6 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.7 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.7 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.8 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.8 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 1.9 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 1.9 --v 5.2 --style raw
Tsutomu Nihei's illustration --iw 2.0 --v 6.0 --style raw
Tsutomu Nihei's illustration --iw 2.0 --v 5.2 --style raw
portrait of wizard of flowers with green leaves in background --iw 0.1 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.1 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 0.2 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.2 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 0.3 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.3 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 0.4 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.4 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 0.5 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.5 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 0.6 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.6 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 0.7 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.7 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 0.8 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.8 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 0.9 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 0.9 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.0 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.0 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.1 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.1 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.2 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.2 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.3 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.3 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.4 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.4 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.5 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.5 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.6 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.6 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.7 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.7 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.8 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.8 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 1.9 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 1.9 --v 5.2
portrait of wizard of flowers with green leaves in background --iw 2.0 --v 6.0
portrait of wizard of flowers with green leaves in background --iw 2.0 --v 5.2
Considering that V5.2's maximum Image Weight is limited to 2, let's explore what the increased maximum for this parameter in V6 can achieve with the same images and prompts (just V6 samples for this one).
girl with whirlwind of threads and wires around her head by Bert Stern --iw 3 --v 6.0
Tsutomu Nihei's illustration --iw 3 --v 6.0 --style raw
portrait of wizard of flowers with green leaves in background --iw 3 --v 6.0
Blending Pictures
Midjourney’s /blend command is a fascinating capability of merging multiple existing images to create a new, unique result.
Titian — Portrait of Pope Paul III (c.1543)
Amanda Jones, Rajkumar Ranadeva Singh — Andrei Kovalev (2012) (yep, that's me 12 yrs ago ಠ_o)
1540s Portrait of Pope. Oil on canvas painting --v 6.0
1540s Portrait of Pope. Oil on canvas painting --v 5.2
Unlike single-picture Image Prompts, blends don't need any text input. This means you can just put two or more images into Midjourney’s blender and let it do its magic—based entirely on the visual elements of the pictures.
Jacques Le Moyne de Morgues — A Young Daughter of the Picts (c.1585)
Unknown photographer — Portrait of a girl (1880s)
To explore the nuances of the “blending behavior” of Midjourney V6, I again turned to my photographic archive.
Paul Cézanne — Madame Cézanne In A Yellow Chair (1893)
Georgian writer, screenwriter and movie director Nana Ekvtimishvili for Die Zeit (2018)
painted portrait of woman with flowers in her hair by Paul Cézanne --iw 1.1 --v 6.0
painted portrait of woman with flowers in her hair by Paul Cézanne --iw 1.1 --v 5.2
Okay, it works truly well with portraits, but how about more abstract things, more difficult blends—using visually polar images with conflicting aesthetics?
Andrei Kovalev — Red (2013)
satellite footage of rainbow-hued mountains in Zhangye National Geopark. Top view from space --stylize 75 --v 6.0
Walt Disney, Ub Iwerks — Steamboat Willie (1928)
Edvard Munch — The Scream (1883)
Louis Wain — A Cat In "Gothic" Style (1925/1939)
John Comes Again, Lakota (1899)
Both, Image Prompts in general and Blending specifically open a whole new dimension of creative possibilities in Midjourney. With V6’s sensitivity and ability to recognize a source image’s main features to later reproduce them in an unusual mix, this universe of possibilities has just become much broader.
In the next chapter, we will learn to control our prompts and increase the predictability of Midjourney's generations with parameters. Continue reading →