Generative artificial intelligences have gained popularity in a very short time. ChatGPT is largely to blame, but this project only works, for now, to generate text according to our needs.
[Las IA para crear imágenes han roto el mercado y ahora les llueven las demandas]
If what we want is to create quality images using artificial intelligence, one of the best options is Midjourney. This project has been going on for quite some time and has already been used even in commercial projects, such as the first manga created by Artificial Intelligence.
New Midjourney V5
With its V5 version, Midjourney has the potential to be even better. This was announced on the project’s official Discord, and among the list of news we find obvious improvements in some of the weaker parts of the AI so far.
The best example we have in hands. It’s one of the most difficult parts of the human body for an AI to recreate, so much so that it’s been easy to discern between real and fake images just by counting the number of fingers. It is common for characters created by Midjourney to have six or seven fingers, or even for the hands to be shapeless “tentacles”. With the new version, this should change.
Just a warning – Midjourney AI can now do hands correctly. Be very critical of any political imagery (especially photography) you see online that attempts to elicit a reaction. pic.twitter.com/ebEagrQAQq
— Del Walker (@TheCartelDel) March 16, 2023
The new version of the AI is also able to create higher quality images, with twice the resolution of the current one; this was one of the main drawbacks, especially if we wanted to use the images for other projects. AI now also has a much wider stylistic rangealthough at the same time its creators claim that it will no longer force your style (many images created with Midjourney are easy to recognize by this).
In short, the V5 version of Midjourney seeks to achieve higher quality results and that it is not so noticeable that they were created by an AI. And now we can try it for free.
How to use the new Midjourney
Midjourney V5 is available the same as the previous version. Therefore, we must enter the official page of the project and log in with our Discord account. For now, the AI can only be accessed through the official Discord group.
Once we have logged in, we must enter one of the rooms reserved for “beginners”, called “beginners”. To create an image, enter the command “/imagine” (without quotes) and write in English what you want to create; don’t forget to add “–v 5” to force the use of AI version 5, by default V4 is always used because it is the most stable and the best tested.
At the moment the option to use Midjourney V5 is limited to paying users so we can’t use it if we are free (trial) users.
Other new options we have are “–tile” to create transparent tile-based images, “–ar” to set an aspect ratio other than the default 2:1, and “–iw ” to set the weight the image and text should have for the AI. All of these features are optional and experimental.
You may be interested
Follow the topics that interest you