Midjourney has released V6 of its already impressive AI image-generating model as an Alpha release on its Discord server.
V6 comes with a number of significant improvements but the biggest upgrade is that Midjourney can now add text to your images. Adding text to images with correct spelling has been a challenge for all AI image generators. DALL-E added the feature in its last release and Midjourney users’ wait for similar functionality is finally over.
The text placement, spelling, and conformance to style prompts isn’t 100% but it’s still pretty good.
Midjourney V6 …. We have TEXT!
It can be hit or miss but still learning how it works.
These 4 were all from one generation.
Maybe got lucky 🤷♂️
Prompt in image 1 ALT#midjourneyV6 #MJV6 #AIart #aiartcommunity #digitalart #midjouney #MidjourneyAI #AIArtwork pic.twitter.com/BIJwr49489— Orcton (@OrctonAI) December 21, 2023
The level of detail and realism in generated images has had a big upgrade too. In his announcement of the release of V6, Midjourney founder David Holz said, “This model can generate much more realistic imagery than anything we’ve released before.”
Some of the images shared by the community of Midjourney users look amazing.
Midjourney V6 is looking mighty tasty!
Huge increase in detail at the same resolution.
These are NOT final model images and are not upscaled.#midjouney #AIArtCommuity pic.twitter.com/SJYK6v7LyC— GifCo (@giffboake) December 20, 2023
As with any image generator, the quality of the output is highly dependant on the creativity of the user. Because V6 is a lot better at following the nuances of prompts, the final images are better aligned with the idea that the creator had in mind.
Chris Perna, who was the art director at Epic Games for 22 years, put Midjourney V6 through its paces and came up with some really good images.
View this post on Instagram
New prompting methods
Writing a prompt to get previous versions of Midjourney to deliver the image you wanted involved a lot of prompt trickery. Users would include camera models, lens types, image resolution, and a lot of other phrases to get the right look.
With V6 that has changed a lot. “Prompting with V6 is significantly different than V5. You will need to ‘relearn’ how to prompt. V6 is much more sensitive to your prompt. Avoid ‘junk’ like ‘award winning, photorealistic, 4k, 8k’” Holz explained.
Holz says users will need to be more explicit in their prompts and V6 now allows for longer, more detailed prompts so that images more closely match the user’s vision.
The remixing of images as well as image upscalers have also been improved in V6. Because this is an Alpha release there are a few features that aren’t supported yet. Pan, zoom, and vary region, are available in V5.2 but we’ll have to wait a while for these to be added to V6.
It’s amazing to see how AI image generators have advanced. It’s easy to forget how excited we were when they first came on the scene even though they generated pretty bad images. This post from Marco Nedermeijer is a great example of how Midjourney has improved since July 2022.
Midjourney v1 until v6, same prompt”
white background, closeup portrait of a very old mean man, 92 years old, wrinkles, realistic skin, studio lighting,, canon f/4#midjourneyV6 #midjouney #aiartcommunity pic.twitter.com/g8wAALAbH3
— Marco Nedermeijer (@MNedermeijer) December 21, 2023
Choosing between the different AI image generators is becoming more difficult as their capabilities converge. Midjourney users are mostly stuck with the Discord interface although users who have generated more than 10,000 images are being offered access to a web platform.
The big advantage that Midjourney has over other options like DALL-E is that it will allow you to generate images of real people. This probably makes it the go-to tool for people trying to create parody content or disinformation.
What’s next for Midjourney? There are several good image animators or text-to-video tools coming online lately. Maybe Midjourney will let us add movement to our images in their next release.
