foreword
What is AI painting
AI painting, as the name implies, is painting using artificial intelligence, which is an application scenario of artificial intelligence generated content (AIGC). Its main principle is to collect a large amount of data of existing works, analyze them through algorithms, and finally generate new works, and the algorithm is the core of AI painting. Get a lot of paying users.
What is Midjourney?
Midjourney is an artificial intelligence program developed by Midjourney Research Labs that generates images from text and is currently hosted on a Discord channel. It will enter the public beta stage on July 12, 2022. Users can operate through Discord's robot instructions, and can create a lot of image works, which is jaw-dropping.
Application Scenarios of AI Painting
AI drawing can be applied in many fields, such as design, architecture, art, medical care, education, etc., to provide people with more convenient and efficient drawing solutions.
Let me give you a few simple examples: such as illustration production, art, self-media image production, graphic designers and other application scenarios are very rich. For example, AI painting can provide designers with more design inspiration and creative elements. AI painting can imitate the artistic style and style elements of human painting, providing designers with more reference and inspiration, making it easier for designers to complete their works.
Second, AI painting can improve design efficiency and quality. Compared with traditional painting methods, AI painting can quickly generate high-quality design works through automation and algorithm optimization, saving designers time and energy.
Copyright issue
Of course, the copyright issue of AI paintings has not yet been resolved. After the copyright is regulated in the future, the future market potential of AI is beyond imagination.
Introduction to Midjourney
About Midjourney
Midjourney uses Discord, and Discord is said to be an online real-time calling software and digital distribution platform with 250 million registered users in 2019. It is also a wise move of Midjourney to generate social robot instructions through Discord, so it has become the only profitable company among several "text-generated images".
paid price
Midjourney has set a total of 3 prices (the picture below shows the annual payment price, and monthly payment users are slightly more expensive):
1. $8: 200 maps can be generated per month, there is no fast mode, and the map cannot be generated without queuing
2. $24: You can generate unlimited pictures per month, with 15 hours of time in fast mode, and you can skip the queue to generate pictures
3. $48: Compared with $30, it mainly has a privacy mode. Others cannot see your pictures and keywords, which is suitable for companies and copyright owners
Ordinary users can use the $10 package for learning, but obviously 200 pictures can't meet the actual test and effect needs, so the minimum for general users is the $30 package.
Here I recommend everyone to use the co-rental platform: Galaxy Video Bureau https://iweec.com/818.html The standard monthly payment membership is only ¥42, and you can save 2.1 yuan by using the invitation code vip, each of which can use 3 devices The number of concurrent tasks does not limit the number of plots, the Pro version is 80 yuan/month, supports 12 concurrent tasks and does not limit the number of plots. Support Alipay, no need for overseas credit cards;
If you really mind sharing, be sure to use your Depay credit card to pay the monthly fee.
Midjourney Tutorial
Midjourney Landing
First of all, we enter the official website: www.midjourney.com , click Sign in, and then go directly to Discord and authorize the login; if you don’t have a Discord account, you can register directly at this step, and you can browse the user’s wonderful pictures after the authorization is successful; of course, click Join the Bate directly You can enter the Discord channel;
There are some sub-areas in Midjourney, for example, enter: newbise to view other people's works; some sub-areas only support chatting;
generate your own work
The first method: we enter any newbise sub-area, click + in the chat window, choose to use the app, /imagine prompt: write English keywords after it; or directly enter the /imagine command can also be used;
The second method: connect the Midjourney robot to your own server
Specific operation process:
1. Create your own server;
2. Add the Midjourney Bot in the Midjourney channel to your own server; if you can see that the robot is in our member list, it means success;
3. You can also generate works on your own server;
PS: U means to enlarge the picture, U1/U2/U3/U4 respectively refer to enlarge one of the four pictures
The meaning of V is to use the composition form of the picture to regenerate a group of similar pictures. The order of V1/V2/V3/V4 is the same as that of U
Midjourney Advanced Creation
1. A quick overview of command collections
/imagine: generate an image
/settings: view Midjourney Bot settings
/ask: get the answer to the question
/relax: switch to relax mode
/fast: switch to fast mode
/blend: switch to two image blend mode
/prefer remix: switch to remix mode
/help: View help information
/info: View basic information, such as subscription status, working mode, etc.
/stealth: switch to stealth mode
/public: switch to public mode
/subscribe: manage subscriptions
/prefer option set: create custom variables
/prefer suffix: Specifies the suffix to add to the end of each prompt
/prefer option list: list all variables set before
/show: combine task ID to generate original picture
2. Image detail processing
Handle these backgrounds and details as well as possible, which can be described from the following aspects:
Theme: people, animals, characters, places, objects, etc.
Medium: photographs, paintings, illustrations, sculptures, graffiti, tapestries, etc.
Environment: indoor, outdoor, on the moon, Narnia, underwater, Emerald City, etc.
Lighting: soft, ambient, cloudy, neon, studio lights, etc.
Colors: Vibrant, Pastel, Bright, Monochrome, Color, Black & White, Pastel, etc.
Mood: Steady, calm, noisy, energetic, etc.
Composition: portraits, headshots, close-ups, aerial views, etc.
Corresponding English:
Subject: person, animal, character, location, object, etc.
Medium: photo, painting, illustration, sculpture, doodle, tapestry, etc.
Environment: indoors, outdoors, on the moon, in Narnia, underwater, the Emerald City, etc.
Lighting: soft, ambient, overcast, neon, studio lights, etc
Color: vibrant, muted, bright, monochromatic, colorful, black and white, pastel, etc.
Mood: Sedate, calm, raucous, energetic, etc.
Composition: Portrait, headshot, closeup, birds-eye view, etc.
3. Image aspect ratio
Use the aspect ratio parameter
Add --aspect:, or --ar: to the end of the prompt. For example: --ar 1:2 (according to Midjourney version, can be set to 2:1, 3:2, 2:3, 5:2, 2:5)
4. Picture quality
Sets the quality parameters used for the job. semi-quality = --q .5, base-quality = --q 1, high-quality = --q 2
in conclusion
There are really many Midjourneys, but there are often no problems in using them. The problem is that the generated images are not satisfactory. You can discuss them here!
Official instruction tutorial address: https://docs.midjourney.com/
Disclaimer: All illustrations in this article are created by Midjourney, and some of the content is referenced from Wikipedia, Zhihu and official documents.
Original Author: Meta Walker