Hi Wade. Here are a few more cool style numbers, this time from the range of 101 through 200. 108 - Space background, neon lights, vibrant colors, pop art style, high contrast, detailed digital painting. style of David Uhl and Loish 110 - Watercolor Clipart, White background, ultra detailed, hyper realistic, Light pastel color palette, beige aesthetic 117 - Digital illustration, minimalist background, geometric colorful shadows, backlight, volumetric light, hyper detailed, fine details, high resolution 118 - Flat illustration, light blue background, soft lighting creating gentle shadows. style of James Gilleard, Greg Tocchini 122 - Vector graphic design in the style of a simple clipart with a flat art style and no shadows. 123 - Polaroid photo with color washes, vintage 70s photo collage, magazine cutout, mixed media collage, muted colors, monochromatic, very low contrast/brightness Style of Richard Kern 130 - Beautiful dark gray background, cool color temperature, wide shot, soft lighting, high detail, art style by Greg Rutkowski 133 - Dark photo, shadow play, low-key lighting, dark shadows, muted color palette, shot in the style of David Newton 134 - Blue and Purple, pastels simple illustration, flat vector illustration in the style of risograph with minimilistic, simple shapes, art style by John Holcroft for Amanoguchi. 137 - Cyberpunk style, with high resolution, ultra realistic photography. White background, digital concept, flat lighting, art by Ashley Wood, Dan Brerard. 139 - High-quality photo, ultra detailed, natural lighting, high resolution, HDR, and crisp details in the style of Magali Villeneuve 141 - Lofi photography, muted colors, a grainy texture, a vintage vibe, pastel tones, a beige background, low saturation, nostalgic, natural lighting and natural features. style of Petra Collins. 144 - Ink painting in a monochrome, vintage style with texture. brush strokes and ink blots, high contrast, low saturation, in the style of Frank Miller and Alex Ross. 145 - /Display says it is a photograph and very little else. I think its colored pencils with really great shading. 149 - Gray tones, light silver and dark gold, a concept design sheet, with strong shadows and high contrast. Style of Artgerm. 153 - Light color theme photograph, bright environment, light blue background, light depth of field, minimalist style, simple design, natural lighting, soft tones 160 - Watercolor painting, muted colors, soft pastel colors, dreamy mood, Style of Greg Rutkowski 162 - Photograph with a blue glowing light background element on a dark background, realistic digital art. 171 - High resolution, high quality photography, natural light, professional color grading, soft shadows, rainbow background with a purple to green gradient color tone effect 175 - Vintage magazine cutout advertisement. muted colors, 70s style, halftone effect, black ink on beige paper 176 - Editorial photoshoot, dramatic, capturing every intricate detail of textures and colors, a dark background with dark lighting and a spotlight 181 - Another one where /Display and I disagree. It is a "full-body studio portrait with sharp focus in a hyperrealistic style". a.k.a. nothing special here. I like the color pallete. 196 - Clipart style illustration, isolated on white background. Hyper realistic graphic design in the vector art style of Loish, Amanda Clark, WLOP and Greg Rutkowski Once again there is a lot of good stuff in this photo range and I'm leaving a lot out. I'm having issues with the text portion of my prompting. For the first 100, I used "a full body shot of a 20 year old man". For this second 100, I used "a full body shot of a 20 year old woman". Both prompts amplified certain sref values while overpowering others. The issue is that the values which they amplified or overpowered differed dramatically. Both of these text prompts are obviously highly opinionated and dominant. I need to find a text prompt that is a push over that lets every sref value totally run the show. I believe that these lists would be much longer, more comprehensive and more useful, if I could find that text prompt. Well, I've only burned 1.5 hours tonight, so I'm not going to have enough free time to use up my hours in just treasure hunting. Analyzing and documenting the results takes a lot of time and I am a 1.0 version of computer using human... I don't multitask as well as the younger folk. Thanks again for all your great videos!

Hi Wade. I've stopped in my current efforts at treasure hunting in the style numbers. The more I reviewed the 400 quartet image sets, the more I felt I was leaving a lot of value undiscovered. The prompts that I was using were highly opinionated (young woman, young man) and were made even more overpowering because I added further specifics ("20 year old woman", "20 year old man"). It was a miracle that any of the sref numbers I was testing were even a little visible, given the specificity of the text portion of the prompt. My two lists below were just the values which were powerful enough to survive the bullying of the text prompts. So tonight, tried something different. I grabbed a block of 50 sref numeric values ranging from 201 to 250.. For each value, I queried three different prompts... An Object -- this was an attempt to provide nothing useful in the prompt, leaving the sref to express itself without having to fight against the text prompt. This created a lot of featureless orbs but yielded a little clearer view of what the style was capable of. X -- this was another attempt, this time to make the prompt contribution distinct and easily identifiable. This worked a lot better. Every image has an x in it which can be easily ignored, but everything else in the image comes from the sref. Style -- this was the best option I found. This text prompt seems to contribute nothing. Almost all of the quartet images which were generated show consistent styles, colors and image subjects. /imagine style --sref 201 ...is my favorite prompt for future style treasure hunting. So far, that was the good news. The bad news (which is really even better news) is that when I treasure hunt using the third option above, almost every value becomes amazingly useful. Now that I know what they contribute, I can see easy ways to use them. No list of the best values, even with /describe sourced annotations, is going to extract all the values out of these srefs. Maybe a catalog, or a monthly section in the Midjourney magazine. Either way, it is more than I can do alone at this time. So I'm going spend the rest of the week exploring tonight's 50 styles and retire my treasure hunting efforts, ...for now. Feel free to any or all of what I've posted here, either in your own investigations or your youtube video productions. I really appreciate all that your videos have taught me, and it is good to get a chance to give something back.

Hi Wade, Great video and great idea about collecting a number of cool style numbers. I've got about 17 hours to burn through in the next week, so I'm going to run through a few hundred sequential style numbers each night (easy to do with permutations). I'll post any interesting numbers that I find here so that you and others can enjoy. As I lack any real artistic vocabulary, I am feeding my favorites to /Describe and choosing a few important words for each style number. Here are some interesting finds from last night's work... Style Numbers 1 through 100 15 - Blue light shines from above against an orange-red sky. Fog fills the background in a matte painting style. Dark gray/indigo tones dominate a minimalist design and symmetrical composition 20 - White background, hyperrealism, octane render, 3D rendering, hyper detailed face, digital art in the style of Artgerm, Tom Bagshaw, Greg Rutkowski, WLOP, Frank Cho, Krenz Cushart 24 - Greenish blue and red, standing in a hell world background with lighting and fog, ultra realistic, hyper detailed 35 - Simple, flat illustration of silhouette, muted gold and dark green tones. warm colors. poster 48 - Vector art on a white background, flat design, high resolution, high detail, and high contrast. Cinematic lighting. 49 - In the woods at night, sparks flying from his hands, eerie glow, realistic, detailed, in the style of Zdzisław Beksiński. 55 - White background, realistic black and white photography using an Olympus OMD EM5 Mark III digital camera with an ultra wide-angle lens at an f/4 aperture setting 70 - Dark art, simple, flat illustration with flat color, a vector line drawing, high contrast, minimalism with low details, smooth straight lines, Lomography 72 - Blue clouds illuminated by lightning, Dark anime scene illuminated by cinematic lighting 98 - Green and orange, color pencil illustration, with a retro 90s aesthetic and vintage aesthetic, as a 2D flat illustration, detailed, vector art There is a lot of good stuff, even in this short series of style numbers. My choices above are just my opinions. I left a lot of great stuff off this list. Wade, please give me (and future style number treasure hunters) some advice on what we should be looking for. My limited taste in art will definitely not be adequate to this task. Thanks again for a great video!

Fasciminatin' stuff! Wow! I didn't know you could use style codes from v5.2 in v6. Thanks for the pearls.

Fantastic highlights and just when didn’t think my luv for Style Reference could get any bigger Wade bro! And thank for tip how to get the reference #s via job IDs on the site 🙌🏻✨💯

You can refresh the webpage to see a sref random code

I watch your KZfaq videos very carefully and listen and it was very educational..How can I contact you to ask a specific style code prompt for Midjourney please. I’m having a hard time. Thanks

Awesome video!!! Thanks for helping me prompt better. I seem to be able to simply click the "--sref random" shown under a result on the right and the number code pops up in the imagine box. Seems easier than copying the job id into Discord. But maybe I missed something.

Thank you for this tutorial definitely helpful and I'll play around. Do you know if there's a possibility to create a reference seed number ... from a sref using the url. I currently use the url .. but I think a seed number would be much more accurate. Thoughts.

Now its showing on the website the sref seed from the start. So there was an update on this fortunately. But I must correct you when it was not showing from the start - If you refresh the page in the browser it shows the sref seed, so you dont need to use Discord

So is this how we apply styles now? I used to refer artists styles, are we no longer doing that?

I like the vary region trick, BUT it doesn’t blend

Hello, I have a low-resolution image and I would like to ask an AI to create the same image in better quality. But my main question is, can we submit a photograph and ask to recreate the same or almost the same one?

Hi, I sent you a link to the 300-351 styles in my email, use them if you want.

Can't I add --ar to this particular prompt? --sref random

I can't use the website, only the discord, why?