Это видео недоступно.
Сожалеем об этом.
Turn ChatGPT into a Powerful Midjourney Prompt Machine
HTML-код
- Опубликовано: 5 сен 2024
- In this quick video, I combine the powers of ChatGPT with the brilliance of MidJourney so that you can reliably create winning formulas that give you consistent results inside of MidJourney. This is a quick power-tip that will forever change how you use MidJourney. Enjoy. :)
Resources Mentioned:
ChatGPT: www.futuretool...
MidJourney: www.futuretool...
🛠️ Explore hundreds of AI Tools: futuretools.io/
📰 Weekly Newsletter: www.futuretool...
😊 Discord Community: futuretools.io...
🐤 Follow me on Twitter: / mreflow
🐺 My personal blog: mattwolfe.com/
🌯 Buy me a burrito: ko-fi.com/matt...
🍭 My Backgrounds: www.futuretool...
Outro music generated by Mubert mubert.com/render
#MidJourney #ChatGPT #aiart
Here is a MidJourney Prompt Formula:
(image we're prompting), (5 descriptive keywords), (camera type), (camera lens type), (time of day), (style of photograph), (type of film)
Please respond with "yes" if you understand the formula
For the Ctrl C/V gang
Yes, I understand the formula.
if you're getting prompts back that are split over multiple lines each, try - write me 5 MidJourney prompts for urban photography. simplify your responses. put them on one line
@@dogbreath226😂
.
Thank you u da man!
Here is a tip that some might find helpful. If you're using different formulas in the same chat, you might have one for an art style and another for photography then give the formula a name and tell Chat GPT the name of the formula (keep it short). So you would type out your formula and then tell Chat GPT that this formula will be called A1 (or whatever you name it) and the another formula will be called B1 (or whatever). Then when you want your prompts just type the name and chat GPT will know which formula you want to use.
Example: 5 prompts of cats. A1. Chat GPT will generate the 5 prompts and will use the formula of A1.
Also you can combine formulas by using two (not sure if you can use more) formula names and Chat GPT will combine the formulas. I have a formula for photography and a style of a famous photographer and another for a cartoon style. If I tell chat GPT to use both these formulas then it will use the style of the cartoon and use the inspiration of the photographer. The result you get from Midjourney may not be what you expected.
Here is an example: Leonardo DiCaprio, portrayed in the style of The Simpsons and inspired by Annie Leibovitz, with introspective, genuine, engaging, charismatic, and expressive characterizing the scene, rehearsing lines for a play at the Springfield theater, shot with a Canon EOS 5D Mark IV, using a 50mm f/1.2 lens, during a late afternoon, in an editorial-style portrait, captured on Kodak Portra 160 film.
I hope all that makes sense
Very good, thanks
Once again, you brilliantly show how these two AI tools bring out the best in each other 🙂. Great video!
here is a midjourney prompt formula {Image we're prompting} {5 descriptive keywords}, {camera type}, {camera lens type} {time of day}, {style of photograph} , {type of film}, Please respond with "yes"if you understand the formula
Bless you sir
How the hell i copy from mobile ?
I am on an iPad Pro and I cannot copy and paste from RUclips either. I did just pause Matt's video and wrote down his Midjourney Prompt formula exactly as he had it. Alas, it didn't work for me this time around. I'll keep trying it because ChatGPT can be finicky. Tomorrow I can use Matt's Midjourney Prompt and it will probably work even though it didn't tonight. I love Matt's videos though. I learn so much from him.
@@FLEXTORGAMINGERA make a screenshot of the text then send to Google lens then copy the text
@@FLEXTORGAMINGERA take a screenshot if you have iPhone you should be able to copy text
Pro-tip: now, teach it to break them into logical phrases, and add individual phrase weighting::2 like this::3. It works great, and the images are extraordinarily creative.
Could you give us something we could copy and paste as an example?
Just did a video on something similar! I find Bing picks things up even quicker, is that what you find too?
Honestly, I haven't tested it on Bing yet. I'll have to play around with it over there as well.
Very informative.
I have been struggling hardcore with prompt generation. This video helped so much! Thanks for this one Matt!
Hi! I think you might like my content about Ai. It's about making money by using Ai and I share helpful tips and ideas. I thought you might find it interesting, so please take a look and let me know your thoughts!
Hey man! I just wanted to say "Thank you" for this vid. I'm a fan of both ChatGPT and Midjourney. You really opened my eyes with this vid on the potential of using ChatGPT for not only MidJourney but other use cases as well. Cheers!
Maybe ask chatgpt to improve the formula and include that? I'd like to know this information myself. Good question.
You can add tables for additional prompting inspiration: "create a numbered table with columns of 20 photography styles, angles, distances, lighting options and film types."
Its said: I apologize, but as a language model, I cannot create visual content such as tables. However, I can provide you with a list of 20 photography styles, angles, distances, lighting options, and film types:
Photography Styles:
Portrait
Landscape
Street
Wildlife
Action
Black and White
Still Life
Fashion
Food
Macro
Night
Travel
Documentary
Architectural
Underwater
Aerial
Event
Fine Art
Sports
Candid
Angles:
Low angle
High angle
Bird's eye view
Worm's eye view
Eye level
Dutch angle
Tilted angle
Oblique angle
Overhead angle
Side angle
Distances:
Close-up
Medium shot
Long shot
Extreme close-up
Extreme long shot
Full body shot
Headshot
Three-quarter shot
Waist-up shot
Wide shot
Lighting Options:
Natural light
Studio light
On-camera flash
Off-camera flash
Ring light
Reflector
Softbox
Umbrella
Beauty dish
Spotlights
Film Types:
Color
Black and white
Slide film
Instant film
Negative film
High-speed film
Low-speed film
Infrared film
Cross-processed film
Cinematic film
@@tombradford7035 Hi Tom, we create tables all the time. What is your prompt?
@@tombradford7035 This is what I used in ChatGPT and it worked. It created me a table providing exactly what I asked for: "ChatGPT is an expert at creating numbered tables. Create a numbered table with columns of 10 photography subjects, Cameras, camera lens, photography styles, angles, distances, lighting options, and film types. Do you understand?" It responded with "yes, I understand" and immediately created the table for me. For some reason, I couldn't get Matt's Midjourney Prompt formula to work, which is odd since I learn so much from Matt's videos. But, I've noticed that ChatGPT can be finicky. One day I'll give it instructions and it will follow it exactly and return what I asked. On a different day, I can start a new chat and give the same instructions and it will miss the mark completely in the results returned.
Thanks for your formual. The answer I got with your formula was: I apologize, but as a language model, I am not able to create tables with formatting. However, I can provide you with a numbered list of the requested items: I the get a long list of 1) Photography Styles; 2) Angles; 3) Distances; 4) Lighting options; 5) Film types.
@@MegUSN52 Thanks so much, I've used your input and it worked flawlessly!
Also, i found this Formula I made for creating buildings or rooms very useful,
(subject of image), (6 adjectives),inspired by (name of modern artist), style of motif (style of motif), architecture style (style of architecture), color scheme (color scheme), other decoration ideas (random decoration element).
It may be due to a slight difference but I duplicated your steps to the "T" except I said stable diffusion not midjourney. My results were closer to a anecdote than they were with you. ChatGPT didn't even know what midjourney was. It asked for more information about it. I tried sentence/anecdote, it worked with pretty good results.
Thanks so much for this formula. I have been trying to get realistic images for an ebook I am writing about my young adult life becoming an under-keeper to the estate's Gamekeeper. Try as hard as I may, I never got any results I could use until I used your formulae. Here is a prompt that ChatGPT generated and it is awesome!! Thankyou! Thank you!! Thank You!!!
A portrait of aan old, craggy faced male gamekeeper in front of a hut in the woods, with pheasants running around in a straw cover clearing, capturing her clear facial features and the stunning natural scenery. Shot with a 50mm lens, f/2.8 aperture, in the late afternoon light for a warm, golden hour effect. The style of photograph should be moody and atmospheric, with emphasis on the subject's natural beauty and the dramatic landscape. Film type can be either color or black and white, depending on the desired aesthetic. --ar 16:9
I hope one day, we can input photos into chatgpt and it can generate prompts based on those photos, like a reverse of the AI art generator. I think that'd help people figure out how to write more effective prompts and and it'd also show how AI interprets certain keywords
how about using /describe to get the description of an image. then editing that text to what you want?
ur my new favorite channel dude! thanks for making these
Thank you!
You’re amazing! I’m watching you everyday, visit your website daily and read your newsletter weekly!!! Keep going…I need you! x
The thing with ChatGPT is it will always provide better results when trained. This is a great prompt idea! Thank you
This was Super useful thank you!!, what i found interesting playing around with these formulas is that ChatGPT often takes the initial subject matter, and then recognises what or who it is, and then uses it as inspiration for what it generates for the rest of the prompt formula guidelines. very handy!
I have run into limitations though, mostly cuz of CHatGPT's PG filters. I was getting it to create prompts for prison cell designs and every prompt was something like "prison cell with a forus on comfort and privacy" or "prison cell focused on wellness and calming" LOL, when i tried to get it to create prompts for a not so nice prison cell design it refused to do so, and proceeded to Scold me about human rights :)
But overall this training guide is awesome and helped immensely with my Midjourney projects!. thank you!
Amazing thank you!!! Any ideas how to create formulas for non-photo image prompts? Aka coloring book pages, cartoons, graphics, etc.
omg! short and full of nutrition tutorial! Thank you so much Mr Wolfe!
You change my life with the music AI bro thank you so much
That's amazing Matt nice one.
Matt you never miss💯
U are a genius.. after i watched your last video i ask my self it will be tuff to remember all that things on the prompt and i forgot about Chatgpt 😂😂😂.. life getting easy
Thanks for the video, Matt! Your tips on combining ChatGPT with MidJourney are game-changing. Can't wait to apply this formula to my own projects. Keep it up! 👍
Chad Jippety prompt formulas are the bomb!!! Thanks, Matt, you're the goodest.
When I am trying to do this with the formula: Here is a MidJourney Prompt Formula:
(image we're prompting), (5 descriptive keywords), (camera type), (camera lens type), (time of day), (style of photograph), (type of film)
reply with a yes if you understand the formula
I am getting this results :
Here are five mid-journey prompts for urban photography:
(Image we're prompting): A bustling street with neon lights and skyscrapers in the background.
(5 descriptive keywords): Vibrant, electric, dynamic, modern, exciting.
(Camera type): Mirrorless camera.
(Camera lens type): Wide-angle lens.
(Time of day): Night.
(Style of photograph): Street.
(Type of film): Color.
Prompt: Capture the vibrant and electric energy of a bustling street with neon lights and skyscrapers in the background during the night. Use a wide-angle lens on your mirrorless camera to capture the dynamic and modern cityscape. Choose a street style of photography to showcase the exciting and lively atmosphere. Use color film to highlight the bold and bright colors of the scene.
Love this new prompt 👏
Thank you Matt for always sharing with us!
I really enjoy every single one of your videos! Your work is truly impressive. 👍👍👍👍👍👍
I used the same exact wording as you did with the exception that I changed out "Urban photography" for "Landscape photography". I even asked if it understood the prompt and it replied with "yes, I understand". Unfortunately, the 5 prompts it returned did not work for me. This is an example of one of the prompts it returned: "Capture the vastness and grandeur of the landscape around you by using a specific camera type with a specific lens type to capture the unique features of the scenery during a specific time of day in a specific style of photograph. Experiment with different film types to create unique visual effects that showcase the beauty of the landscape." How do I get it to tell me the camera type, lens type, etc.? Again, I wrote the Midjourney Prompt formula exactly as you did even using parenthesis.
same thing happened to me :(
Edit: i figured out, remember to put in "Here is a MidJourney prompt formula:" before you put in the formula and it should work fine.
Hola, puedes decirme si lograste solucionarlo, estoy en la misma situaciónte lo agradezco
@@shankerm3959 On behalf of Meg A, I would like to say "thank you, Shanker M"! : )
GPT 3.5 or 4? GPT-4 is much smarter.
thank you bro, i cant believe this AI stuff and im hoping to get in on it asap with my own business. the world is really changing and we have to change with it and be a step ahead
This is awesome! And I presume that you could save the chat on the left side and not have to retrain it the next time you go back into it
Actually you need to delete a chat, it is saved by default.
I am incredibly grateful I found your channel. You are officially my new BEST FRIEND!
Thank you Matt! You are helping me to understand this new world, from Portugal! God bless you!
Thank you Matt, this is going to be really useful!
thanks matt for putting your energy in each of yours videos .
Would you mind sharing what video making software you are using to create that cool circle that you can move around with you talking in it? It allows you to record your screen at the same time right? Very well done video, and very informative. Thanks.
Definitely I will use this tips to generate my images! Thank you!
You are giving my secrets. I was the only one in the world who thought about that and it bothers me everyone doing now 🤣
Great work! Didn't expect such good results! ;)
I'm about to binge your entire channel. See you in last months comment section
not sure if ChatGPT use info from other conversation with other people, but I didn't have to ask without the parenthese as it gave me the good formula on first try (even though I write the formula with parentheses)
Thanks for this tip, it's a great time saver :)
I was about to ask Matt about the parentheses, with or without them, what would happen? I thought they made no difference, and you proved it!
Chat GPT 4 outputs the formula using Carriage Returns and as (Variable) Prompt. Looks like if you say, please rewrite that without Variables, it works.
Thanks Matt, so so useful tips!!! Support from Australia!
I just discovered that if you give ChatGPT a list of prompt style settings at the end of the formula & you tell ChatGPT their variables it switches them on & off as needed.
Super Detailed Infocus: Yes
Photorealistic: Yes
Hyperrealistic: No
Detailed Infocus: Yes
Dynamic-Colours: Yes
Full Focus: Yes
Hyper-realistic: No
Infocus: Yes
Lumen-Reflections: No
ProPhoto-RGB: No
Realistic Detail: Yes
Realistic Materials: Yes
Smooth Infocus Details: Yes
Smooth Sharp Focus: Yes
Super-Resolution: No
Symmetrical: No
Ultra High Definition: No
i've been trying to do this, had some success! chatGPT is so darn powerful! it is really the future!
This is great, thank you, Matt!!
Just Brilliant! Thank you so much!
Also, you can ask chatgpt to give you other descriptives that will make these even better
Incredibly helpful. Thank you. You make so many valuable aspects of this very accessible. Appreciate it.
Great. Have to try it myself
your video are always top notch, but I am mind blown by this one! thank you
this is awesome !! i knew this is about time to happen
Exactly what I was looking for. Thanks!
Thanks Matt 👍
Very impressive results
Thank you for great tips.
Amazing content on the quick tips!
Thanks for the info will definitely apply this to my prompts and workflow.
Great stuff! PLEASE keep it coming! You're doing the heavy lifting for the rest of us.
Awesome tips. I would’ve never come up worth anything like this. It amazes me how smart people are lol
Great video Matt! Thank you.
Thank you so much for these prompts.
Your giving me more and more ideas. Great contents. Thank you.
Great video, thank you so much Matt! :)
We are luvin your tips. Well done!!😃
wow thanks Mr Wolf Dis great!
Great tips Matt 🤩 thank you
This formula approach is simply amazing. Do you have any advice on applying to many other dynamics in ChatGPT? How / why does approach work so well in ChatGPT?
Great video bro! Keep them coming… 💪🏼
Great Idea. Could you do something like this for the current stable diffusion models too, like deliberate 2 or realistic vision 1.3? I feel there is a much bigger need because of the more specific prompt you have to use to get good images.
Absolutely. You can train ChatGPT to create any type of formula you want.
Awesome vid man, thank you for sharing!
It will be great if i can implement with these with Blue Willow. Can I use this for Blue Willow also?
cool thank you , definitely helped me
Excellent. Thanks for sharing.
it so help me, thanks man, it is very helpful
You can also use chatgpt to create a prefer option set.
This video made me subscribe. 🙂 Thank you for sharing!
I love this! Thank you!
Great video. Keep them coming.
So helpful, thank you!
Matt is doing the lord’s work. I was struggling this weekend trying to make things happen in Midjourney. Joined the newsletter. Thank you!
Thank you man ❤
That was really helpful 🔥👊🏽
amazingly helpful. thank you kindly
Hey please tell me how to create 3d inflated puffy effect design using MidJourney 🙏
Real good, Matt - thanks!
Thank you very helpful
These tips leave me... i don't know... all this is awesome.
What a great video! Thanks!
Subscribed. Thanks for the interesting video.
Great video, liked & subscribed
Thanks for a nice work, Matt :-)
Liked and subscribed your awesome thank you!
MidJourney Prompt Formula:
(image we're prompting), (5 descriptive keywords), (camera type), (camera lens type), (time of day), (style of photography), (type of film)
Great job at usual
Great tip
that works pretty cool, thanks
Hi from a big fan of you! Could you tell me what video editing software you are using to make youtube videos? Your videos are always cool.
Can you Show this style of prompting for other imagestyles as well not only for fotographs maybe, thanks btw edit: sry didnt watch it until the end 😅
Hey Matt, love you work you're doing a great job
I'm wondering if something is up with Chatgpt because I can not get this to work... even tried typing in exactly as you did following step by step in the vid. I don't get prompts i get full descriptions. I even went so far as to ask Chat to read me back the formula, which it did , then went right on giving me what I don't want... tis frustrating but not your fault (obviously) just wondered if anyone else is having the same issue
Awesome stuff, thanks for that saved me a lot of time
More ChatGPT training videos please Matt!