Midjourney Prompt Tips: Codes & Tricks - WHAT STILL WORKS? (Perfect for AI Art Beginners!)

[@dameanvil]

01:20 🖼️ Aspect ratio code (AR) works on every algorithm, allowing customization of image dimensions. 02:29 🎨 Chaos code (chaos) adds variability to generated images; higher values lead to greater diversity. 03:25 🚫 Negative words code (no) de-emphasizes specified words in the prompt, influencing the output. 03:53 🛑 Stop code (stop) controls the percentage of the image that is rendered, providing a way to focus on specific details. 04:34 🌱 Seed code (seed) maintains consistency in generated images by using the same starting point. 05:53 🔁 Same seed code (same seed) generates similar images when using specific seed numbers. 07:03 🎨 Style code (style) offers variations (4A, 4B, 4C) in the output style, depending on the chosen algorithm. 07:58 📊 Image weight code (IW) influences the importance of an image URL in generating the prompt. 09:34 🎥 Video code (video) generates an mp4 file or a link to a video based on the specified prompt. 10:48 📷 Quality code (queue) adjusts the rendering time, impacting the level of detail in the generated image. 11:43 🖌️ Stylize code (stylize) allows for customization in image stylization, with different levels of intensity. 13:21 🌈 Creative code (creative) enhances creativity in output images for test and test P algorithms. 14:04 🧩 Tile code (tile) enables seamless repetition of images, useful for creating patterns and designs. 15:12 🆙 Upscaling provides options to increase image resolution based on the chosen algorithm. 19:50 🔄 Multi-prompt feature allows for separate and weighted inputs within a single prompt to influence the output.

[@LburgVAGuy]

The possibilities are endless! I've spent 1000s of hours in front of a computer editing on photoshop,then this comes along. I almost fear for the art world now 🤣

[@WadeMcMaster]

Haha i hear ya. But when you combine the two, man the possibilities are insane. I still think skilled artists will have a place even with ai

[@LburgVAGuy]

@@WadeMcMaster it's gonna be fun to play with! I only heard of this new tech a few days ago when I found some cool Ghostbusters Clive Barker concept AI art on YT

[@mumumumumumumummm]

@@WadeMcMaster Absolutely! Those who are not familiar with Photoshop, image editing/retouching, and even illustration will be limited to what MJ gives them. Those who are already artists will have the edge.

[@Jeholista]

Before A.I......i spend hours sticking to one specific design of my work...but now i can view it in a much complete version of my style....i have more control over design...and the design that stands out will reign...

[@AISocialProfitsHub]

Your video has given me a better understanding

[@AICharacterMandyAnderson]

I'm always on the lookout for the latest tips and tricks to take my AI art to the next level. Your video helped, especially with the upscaling. My content will be better, thank you.

[@WadeMcMaster]

Glad it was helpful! Thank you :)

[@deronnyz]

I love your content, fast and to the point 👍

[@WadeMcMaster]

Thank you so much! I put a lot into doing it that way, so very good to know that :) much appreciated.

[@deronnyz]

@@WadeMcMaster No, thank you😁

[@DanGeesky]

the best midjourney tutorial ive seen, cheers mate.

[@WadeMcMaster]

Thank you so much!

[@mughalrais5968]

Helpful 👍🏻

[@WadeMcMaster]

Thank you

[@hectorruiz9862]

How did you get your thumbnail??? That’s a dope picture

[@WadeMcMaster]

Thanks man! I cut out the mouse in one picture and popped a ink dripping image behind it. Bith made in AI but put together in Photoshop

[@KayLyda]

Thank you so much, you're awesome!

[@WadeMcMaster]

Thank you!

[@luuk0101]

Awesome job!

[@WadeMcMaster]

Thank you :)

[@brainrussell6811]

The updated Version 4 threw me for a loop... it initially Sucked, and I was burning through my Fast hours for the first couple of days. I usually sign up for the $60 plan, and work 12 - 14 hrs/day... and then spend a couple of weeks reviewing/developing written character synopses, cleaning up via Affinity, etc. The algo(s) seems to respond positively to intense periods of prompt reiterations.

[@WadeMcMaster]

Use --style 4a or --style 4b!

[@brainrussell6811]

@@WadeMcMaster It took a few days to feel through - like getting to know/train a new employee - but finally had a breakthrough, and the results are exceptional once again. And yes, "4a" is more my personal style... "photorealistic" was also helpful. Also switched back to V3 which is still useful and can yield great images. So, original comment amended. Thanks!

[@bySterling]

Wow I didn’t know about Stop prompt Great video I didn’t know about Stop either but don’t know why anyone would use that one

[@WadeMcMaster]

No me either haha. But its still fun to play with

[@bargadtoons]

I like your typing speed. Well content and the presentation also. Love from INDIA:) 👍

[@WadeMcMaster]

Thank you! From Australia :)

[@AlbaLoveBuzz]

This is great, thank you for this video! But I am terrified of clowns and at some point was hard to see hahahaha

[@WadeMcMaster]

haha, sorry about that, just close your eyes and listen lol. Thank you for checking out the video :)

[@AmmerSlammer]

Wade excellent video. Thank you. One question, I upload an image of my face and prompt MidJourney to use the image of my face and put it on a superhero for example. I can not get it to work. It always renders the image as just my face size still without ever showing the entire superhero body with my face on it. I have tried all the suggestions of show full body, full body perspective, wide perspective as well as specifying what the shoes and hands look like, but I always only get a face shot at best. Any suggestions from anyone? Thanks!

[@WadeMcMaster]

Yeah i hear what youre saying. It can be tough. Can you photograph yourself with a full body photo and add that with it? Also maybe try word weight, so saying [full body]::3 or something like that?

[@WadeMcMaster]

Also, thank you! :)

[@madastos]

would yo know a way to get the same image to repeat in a different motion for example? instead of reproducing different images of a person each time i prompt the same words?

[@WadeMcMaster]

I personally would try saving an image of that character, then adding it (the url) into your prompts, and describing what it does in text

[@chiloveradiouncut]

Can you get the seed using the Discord on your phone?

[@namitakalita6837]

Hey, I have a question. When you put the algorithm to imagine any prompts and click 'enter', it produces four choices, Right? So my question is - as you know we can create 25 images in the free trial - so, are those four choices also taken into account out of 25? That is, now will I have 21 images left? Or is it the final selection that they take into count? Please answer.

[@WadeMcMaster]

Im confident the 4 images counts as one. But ive been payi g for months so that may have changed (i dont think so tho)

[@namitakalita6837]

@@WadeMcMaster Yo thanks Wade. Actually, I wanted to give a try in Midjourney later this month. So, better be full prepared, right? :) Your videos are like the guiding light for me. Keep it up!

[@WadeMcMaster]

Thank you!

[@raselahmad1144]

I Like your your content.

[@WadeMcMaster]

Thank you Rasel!

[@Northern_Star]

Isn’t image weight active in V4? You can use it by following prompts with :: 1-5

[@WadeMcMaster]

Word weight is active, which is what you've referred to above, but --iw only works with older versions. I'm not sure if you can apply a word weight o an image actually. I'll have to test it out :)

[@Northern_Star]

@@WadeMcMaster image weight isn’t active in V4!

[@WadeMcMaster]

@@Northern_Star Exactly :)

[@powerfulclips101]

love it

[@WadeMcMaster]

Thank you

[@WorldFishingAlliance]

The envelope does not give me "seed" except puts the envelope icon below the images. Is there a trick?

[@WadeMcMaster]

Another way of doing it is to click on the discord message the same way as you would to react, except hit the '...' then when the dropdown appears, hit 'Apps' and 'DM Results'

[@tn7183]

im no artist and am new to all this, but my question is how do you know all the codes to imput into mid journey?. Its pretty confusing if you have never been using stuff like this, it's super interesting but just overwhelming 😂.

[@WadeMcMaster]

Haha i hear ya! For starters people used to twll me about them, but you can also sign into discord and checknout their announcements and whats new section, along with their website documentation.

[@bektassari9430]

How can i add my own text to midjourney?

[@WadeMcMaster]

Ar this stage youd have to do it in another program like photoshop after, ai doesnt really do text well right now, it comes out a bit gibberish.

[@bektassari9430]

@@WadeMcMaster Thank you

[@stevenlangfield9120]

Can someone help me please? If I create a cool image of a character in MidJourney, can I then get the AI to use that same image in different poses or in different scenery?

[@WadeMcMaster]

Some people have had success creating a couple of image references to use in prompts (adding the image url in for MJ to refer to). Id play with that and the prompt text after to adjust. So, If you can genetate one or two images with character looking similar, you can use those as the reference. But i still think youll find some mimor differences.

[@stevenlangfield9120]

@@WadeMcMaster yeah that's my mission today. Sometimes you get something great and wish to play using same character in different scenarios. Thanks for taking the time to reply

[@MusaMusa-nm1ti]

But how did he do the rat on the keyboard

[@WadeMcMaster]

I had to catch it first, then train it whilst creating tiny clothes and goggles for it to wear. But that was too difficult to so I used this prompt: (2 images added here) cyberpunk mouse computer hacker --ar 3:2 The two images I added were: The rat: cheeky rat computer hacker, cyberpunk --ar 3:2 The neon paint: ink dripping neon abstract effect --ar 3:2 So I used the process shared in this video: kzbin.info/www/bejne/bKSUhIikfZZ9jdk Keep in mind I added a few more paint drops from the reference image in photoshop afterward.

[@MusaMusa-nm1ti]

Hey thank u for the fast response great man ❤

[@WadeMcMaster]

@@MusaMusa-nm1ti No worries mate!

[@WatchMyShorts303]

Where do i get all this midjourney commands like "chaos"?

[@WadeMcMaster]

Try docs.midjourney.com/docs/en/parameter-list

[@МарянСалабай]

Hi Can I sell images generated by Midjourney AI with Basic Plan an stocks?

[@WadeMcMaster]

I beleive you can, id upsize them though. Topaz is awesome for that: kzbin.info/www/bejne/fqe3e4OdereXftU

[@WatchMyShorts303]

And another totally noob question i think 😅 How i can find my prompt/result in discord? sometimes there are so many messages after mine and i cant find my result. Is there an option to jump to my results?

[@DaTruthakaMack]

Not sure if this is the "pro" answer or not, but I create a new thread in the least busy newbie channel. Hope this helps my boy.

[@WadeMcMaster]

You shoild be able to do a search for your handle and find your message that way. Look for a magnifying glass icon on your phone (similar on desktop, cant remember and im not near a computer) But id do this insyead to cutout the noise, if you can: kzbin.info/www/bejne/jZ7FZaOZYr-kqtk Or message the bot directly- itndies depend on your plan though.

[@WadeMcMaster]

Awesome advice man, thank you!

[@teenat7044]

You can just invite the bot to your own private server. That way you can make categories and text channels and just use the bot in an organized manner

[@teolita]

OK we are now at Version 5, one month after 😳

[@WadeMcMaster]

Alawys the way, soend a heap of time on a video and the software changes just after lol

[@mihleo6391]

Is this still working on v5?

[@WadeMcMaster]

These codes are mostly available in v5, but the idea of th video is to link up which coses work with which version. This video covers v5s codes briefly: kzbin.info/www/bejne/rJiVo5xpdp2ie5I

[@mihleo6391]

@@WadeMcMaster thank you. Gonna check this out

[@dstarks360]

I cannot get Midjourney to change the aspect ratio whatsoever...am I inputting something wrong?

[@WadeMcMaster]

Thats odd, are.you using --ar (space) number:number ? So --ar 16:9 ? Its pretty consistent so im surprised if that wont work

[@bobparsonsartist564]

Lost me on the "envelope" i cant even get that window.

[@WadeMcMaster]

This video covers the envelope in more detail: kzbin.info/www/bejne/l5bSo3ZuYqunhK8 thia feature doesnt work with default / version 4 though, so you may need to switch to v3 or a test model in / settings

[@legendary_stories]

Wow, very interesting video. Like if you like this video

[@Steve_Fid]

talk about papercut style "midjrny-ppcut"

[@WadeMcMaster]

I had a quick look, awesome model by the looks of it.

[@Steve_Fid]

@@WadeMcMaster thanks

[@parasgoyal2819]

How a girl come in hot dog ??

[@WadeMcMaster]

Im going to theorize.... the word 'hot' was emphasized in that prompt. Its a word often used to describe them online. So the AI may be misinterpreting that lol.

[@Jirodyne]

@@WadeMcMaster I've got a question about that tho. Even if the word Dog is lower emphasized, it's still part of the prompt. But no where in the images is a dog at all. Is the setting 1, basically the same as --no ? or did the AI just completely ignored the word when it wasn't meant to?