Midjourney V6 RELEASED! Everything You Need To Know

mid Journey V6 just dropped and people are already creating these amazing images with it here's everything you need to know about the new update and how to get started using it let's Jump Right In we'll start with the announcement David on the mid Journey Discord says we're going to let the community test an alpha version of our V6 model over winter break starting now to turn it on select V6 from the drop down menu under settings or you can just add the d-v version 6 essentially after your prompt what's new with the V6 base model well much more accurate prompt following as well as larger prompts this is interesting because I think this is actually going to stack up favorably to Dolly one of the things I really love about dolly is the fact that you can give it sort of a paragraph of text you can even give it a script to a YouTube video and it's going to coherently create an image based on all of that data so let's see how much better V6 is at that improve coherence and model knowledge improved image prompting and remix and minor text drawing ability that's cool cool because one of the big things from sdxl and from Dolly 3 is that it's getting better and better at creating coherent text on images so we'll test that out as well there are a whole bunch of new arguments you can pass here are some of the bigger changes though that I want to talk about style and prompting for V6 prompting with V6 is significantly bolded different than V5 you'll need to rearn how to prompt that's kind of a key thing so I don't want to just drop in some version five mid Journey prompts into this and expect some new miraculous things to happen I think we're going to have to try a different prompting method and I have an idea for that V6 is much more sensitive to your prompt avoid junk in quotes like award-winning photo realistic 4K 8K this was widely used in prompting up until this point so I'm kind of glad to see that stuff going away be explicit about what you want it may be less vibby but if you're explicit it's now much better at understanding you and if you want something more photographic less opinionated more literal you should probably default to using style raw lower values of style eyes may have better prompt understanding while higher values may have better Aesthetics that's interesting to know all right so what better way to test this out than to go to my bestie Chad GPT and ask it for some prompt ideas give me 10 different prompts to use for Mid journey to test its new image generation abilities and adherence to natural language I think that last piece is the key I want to be able to test this in a way that uses more natural language prompts because V6 has changed drastically over previous versions of mid Journey so we've got 10 different options here we're going to drop these into Dolly 3 and we're going to drop these into mid Journey V6 and compare them side by side and the reason I'm testing it with Dolly 3 and specifically the Bing version of Dolly 3 Bing image Creator is that Bing image Creator is arguably one of the best free image generation tools out there so I think just comparing these side by side is going to be nice it's also really good at understanding long paragraphs of text the first thing we'll do in mid journey of course is enable V6 so we did Slash settings we're going to drop this down and you can see mid Journey model V6 Alpha just select that from the box now that's going to be the default going forward our first test prompt from Chad GPT is a bustling futuristic cityscape at dusk with neon lights and self-driving cars zooming down multi-level highways the skyline is dotted with Sleek high-rise buildings and drones are flying in the sky let's kick this off in both systems all right and here are the results from the first prompt now the first thing you'll notice is that both of these have drones flying through the sky they have a very similar vibe to them both very dark the interesting thing is there's a lot of neon going on in the Bing image Creator version and it looks a little bit more cartoony like an illustration which is pretty standard for being in Dolly 3 the one on the left from mid journey is quite a bit more detailed it's still not as high quality as I would have expected it's not really a photograph you can still see there's a lot of abnormalities around the cars things like that but interesting nonetheless and you can see that both of them had a similar adherence to the actual prompt that we gave them even though it was a sort of complex sentence and not something as descriptive as you'd normally provide to one of these systems let's try another one our second prompt is an advanced 3D printing Workshop filled with state-of-the-art 3D printers of various sizes the room is busy with engineers and designers working on intricate projects from small gadgets to large mechanical Parts GPT knows me really well it knows that I love 3D printers this could be pretty cool and here are the results both systems produced very similar sort of print Farms it looks like a whole bunch of 3D printers in various warehouses I will say the creativity looks to be a little bit higher on Bing but you've got to give the aesthetic and the overall photo realism to version six of mid Journey for sure this looks obviously more cartoony on the Bing image side which again is sort of the aesthetic that Bing always has with all of its images very cool let's check out another one for our third image we've got a modern living room showcasing an AI powered home automation system the room includes voice controlled lighting automated window blinds a robot vacuum and a central Hub with a holographic display that's very specific spefic and here are the results back from both now obviously I can't tell if these are automated blinds or not but you can see on the wall there's some sort of holographic design diagram some sort of interface from the Bing image Creator side these also all seem to have at least from being image Creator a robot vacuum so it followed the prompt very well there now on the version 6 mid Journey side we've also got what appears to be robot Vacuums in at least three of the images I don't see one in this upper left these are again very photorealistic what I don't see in the mid Journey images is any sort of indication of a holographic display that was mentioned in the prompt number four a conceptual design of a futuristic self-driving car the car should have a Sleek aerodynamic shape with transparent glass panels Advanced sensors and wheels that suggest high-speed capabilities I got to say these are pretty cool you can see the mid Journey versions on the left it's got a lot of glass a lot of transparent materials it really followed the prompt well and I would say these definitely look like a Sleek sort of high-speed vehicle on the right hand side with Bing image Creator we've also got some really slick looking Vehicles it looks like they stole the grill straight from a Mercedes but apart from that these all also look like very Advanced self-driving Vehicles Sleek full of glass one of the things that you'll notice with this is it seems like both systems are pretty good at understanding these prompts and creating an image it's I'm almost shocked at how similar the images are that are created with both Dolly and mid Journey not from an aesthetic point of view but just from an overall following the prompt perspective now if you look solely at the quality of the images obviously mid Journey V6 is far superior you see the lighting the reflection it looks like a photo versus more of again that illustration style so depends on what you're trying to get out of these systems but both of them seem to follow the prompts really well and this is something cool to see out of mid Journey let's jump on and check the next one so for image number five we have a depiction of everyday life enhanced by artificial intelligence show various AI applications like a smart fridge suggesting recipes a virtual assistant planning a day and an intelligent security system monitoring a house that's a lot going on in that image again we have kind of shockingly similar results from both systems it sort of picked up that you have multiple things going on I think Dolly really did a good job here illustrating that there are three or four different asks in the prompt and sort of highlighting those in separate scenes if you will within a single image I did this in it looks like all four of the images now a little bit different from mid Journey but it looks like it sort of did the same thing in the first image and then the second third and fourth are just sort of depicting this sort of futuristic home it looks like still very cool images and I would say both have a very similar coherence and adherence to The Prompt number six 3D printing and space exploration a scene to depicting the use of 3D printing and space exploration astronauts are using a 3D printer to create tools and parts on a space station with a view of Earth in the background these are really cool prompts and I know it took you can even see down here below it says these prompts are designed to explore the various aspects of technology and Innovation reflecting your interests in 3D printing home automation self-driving cars and artificial intelligence Chad GPT you know me too well again very similar results photo realistic on the left with mid Journey but the overall View and look and feel of these is pretty similar I will say that it looks a lot more like a 3D printer in all the images that Bing created and I don't really get a sense of a 3D printer in any of the images that were created with mid Journey they're all based in space they're all seemingly exploring something although with that hatch open in the back there I think these guys would all be dead but beyond that point I think Bing win's out here for following the prompt a little bit better even if the image quality isn't as good number seven Smart City powered by an AI automation panoramic view of a smart City powered by Ai and automation include features like automated public transport energy efficient buildings and interactive digital Billboards again a lot going on here on the mid Journey side we've got some beautiful futuristic looking cities that came back I love this one with all the trees running through the roadway I hope the future cities actually look like that on the Bing side it looks like what they did is sort of again highlighting different elements within the picture showing these circles these arrows that sort of indicate here's the automation features here energy efficient buildings here digital Billboards it's more of something you'd have in almost a brochure for example number eight Next Generation home office setup a Next Generation home office setup with advanced technology include a multiscreen computer setup a 3D printer creating office supplies and a virtual reality meeting space I don't know about you but I'm interested in any of these home office setups I would take these in a second again I think being image Creator does a better job at depicting 3D printers you can see there's a couple of what look to be 3D printers in every single one of these scenes but man the images from mid Journey are just stunning you can see the multim monitor set up it looks like a futuristic office it doesn't appear to have a 3D printer and any of these images but still really really beautiful number nine Aid driven YouTube cont content creation Studio it's like they know my future here an AI driven YouTube content creation Studio showing a content creator working with AI tools to edit videos manage social media and analyze viewer data is that R2-D2 in this image I think it is that's really interesting I think you might have some copyright issues there Bing looking at these other images pretty cool they do a lot more to kind of depict the fact that this is an artificial intelligence YouTube Studio I think it does a better job at getting that across now on the left- hand side with mid journey man these are stunning but I don't really get the sense that this is anything other than a YouTube Studio that you'd have today someone editing video it doesn't look like it's generative AI powered or really futuristic in any way but man it looks good and for our 10th prompt personal transport drone a futuristic personal transport drone large enough to carry a person the design should be sleek and modern with with vertical takeoff and Landing capabilities and a comfortable interior who doesn't want a comfortable interior and here's what we got back from both systems man again those mid Journey V6 images just the lighting and the reflections the overall look and feel of these is just amazing no one else can touch mid Journey right now as far as realism goes that's for sure and the level of realism and detail coming back from these mid Journey V6 images is second to none I think mid journey is still leading the charge there I will say that Dolly seems to follow prompts a little bit better it has all of the details that you ask for and specify in the prompt but given that mid Journey V6 is still just an alpha product we're going to see a lot of iterative changes over the coming days and weeks it's still missing some key features that other mid Journey models have like pan and zoom they're also building an entirely new describe model describe as a name suggestion and if you haven't used it in the past it's able to take an image as input and then give you back a prompt the reason I think they need to build an entirely new model for V6 is because it has a new prompting style and technique and so being able to describe older mid Journey images in the V6 prompting style could be really interesting something to check out for sure as they release these new features and if you want to see it here first hit that like And subscribe Button as always I'm Brian love it this is all your Tech AI thank you so much we'll check you next time

Channel: All Your Tech AI

Views: 7,802

Keywords: ai art, midjourney version 6, midjourney v6, midjourney news, midjourney tutorial, midjourney ai, midjourney v6 features, midjourney v6 rumors, midjourney v6 new features

Id: YELxABTX0T8

Length: 13min 26sec (806 seconds)

Published: Thu Dec 21 2023