Post by tonima5 on Jan 18, 2024 5:51:17 GMT -5
In general, there are dozens of functional platforms capable of generating quite digestible images based on descriptions (prompts). There are even those that can “redraw” your simple sketch and create a real work of art from it. We can name a few of the most popular: Midjourney; Stable Diffusion; Crayon; Starryai; GauGAN2; Dream. Many neural networks generate images in the cloud. And this is a priority option for most, because the operation of a neural network requires large computing power. Creating a 512x512 pixel image can be an extremely time-consuming task, even for relatively new and fairly powerful computers. It is recommended to use macOS devices with M1 or M2 chips, which cope well with such tasks. Hence the advice: if you do not have a high-performance computer, use cloud neural networks.
Otherwise, you will have to spend 10-20 minutes Email Marketing List generating even a small image. It is unlikely that this process will bring you pleasure. But let's move on from theory to practice! Generating images using Midjourney We chose this particular neural network for three reasons: Midjourney is rightfully one of the most advanced technologies at the moment. The neural network “understands” really well what they want from it. With it you can generate up to 15 images for free, and resourceful users will certainly find a way to get around this limitation. The only thing that makes using this neural network somewhat inconvenient is the need to register in Discord in order to interact with the Midjourney bot and send it requests/prompts. It's not difficult, but it does require a little time on your part. We will not describe the process of registering and connecting a neural network bot, since there is simply a ton of this information on the Internet. You will find it without much effort.
Let's get straight to the point! How to query Midjourney to generate images Everything is simple here. In the Discord chat, after connecting the Midjourney bot, just enter /imagine and your request - what exactly the neural network should generate. The prompt must be in English! Let's start with something simple! Let's say we are writing a text about website promotion and we need an image of a man in a suit sitting at a computer in the office. Actually, we enter such a request in the text field: Press Enter and get several options: Note! By default, Midjourney generates four image options at once with ONE image and with an aspect ratio of 1:1. But this can be changed. Details below. Agree, it turned out quite well. But! Personally, we immediately have two wishes: We want the images to be more realistic rather than hand-drawn. We need to get a picture with an aspect ratio not 1:1, but 16:9. It wouldn't hurt to rejuvenate the man. And yes, all this can be done! But we will have to expand and complicate our query somewhat: At this point, we recommend stopping and reading our prompt more carefully to understand what exactly we indicated there.
Otherwise, you will have to spend 10-20 minutes Email Marketing List generating even a small image. It is unlikely that this process will bring you pleasure. But let's move on from theory to practice! Generating images using Midjourney We chose this particular neural network for three reasons: Midjourney is rightfully one of the most advanced technologies at the moment. The neural network “understands” really well what they want from it. With it you can generate up to 15 images for free, and resourceful users will certainly find a way to get around this limitation. The only thing that makes using this neural network somewhat inconvenient is the need to register in Discord in order to interact with the Midjourney bot and send it requests/prompts. It's not difficult, but it does require a little time on your part. We will not describe the process of registering and connecting a neural network bot, since there is simply a ton of this information on the Internet. You will find it without much effort.
Let's get straight to the point! How to query Midjourney to generate images Everything is simple here. In the Discord chat, after connecting the Midjourney bot, just enter /imagine and your request - what exactly the neural network should generate. The prompt must be in English! Let's start with something simple! Let's say we are writing a text about website promotion and we need an image of a man in a suit sitting at a computer in the office. Actually, we enter such a request in the text field: Press Enter and get several options: Note! By default, Midjourney generates four image options at once with ONE image and with an aspect ratio of 1:1. But this can be changed. Details below. Agree, it turned out quite well. But! Personally, we immediately have two wishes: We want the images to be more realistic rather than hand-drawn. We need to get a picture with an aspect ratio not 1:1, but 16:9. It wouldn't hurt to rejuvenate the man. And yes, all this can be done! But we will have to expand and complicate our query somewhat: At this point, we recommend stopping and reading our prompt more carefully to understand what exactly we indicated there.