Midjourney V6 FULL BREAKDOWN (INCREDIBLE, Text, Light Rays + More)

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
mid Journey version 6 is here and it's raised the bar once again for the quality of AI art images everything from the rendering of light to the cence of fine details like singular strains of hair are now vastly improved and what's very exciting is the ability to render text directly inside of mid Journey opening up a whole new realm of possibilities so let's dive in and take a look at everything that's now possible in mid Journey version six so in this video we're going to break down exactly what's new in mid Journey version 6 how you can use it we'll approach the limitations that exist then we'll take it to the test and run a number of prompts through version 5 and version six to see exactly what changes are here we'll take a look at what we've got to look forward to in 2024 and that is where things are going to really change change the game so these are some of the images that we are getting out of version 6 and as you can see we've got an incredible increase in detail coherence and complexity in the announcements chat of Discord on Mid Journey we've got a full breakdown of everything that's new with the V6 base model so let's run through what's here we have much more accurate prompts following as well as longer prompts improved coherence and model knowledge which means that the V6 base model is more closely following the words that we put into mid journey and it's handling much longer prompts more effectively beyond that it's improving its coherence which means how well the entire image works together how coherent is the image how much does the entire image make sense and the model has a much better knowledge of the real world so this relates to improved understanding of people events and other cultural references now one of the most exciting features that is coming in version 6 is text so now there is the ability to render text which opens up a whole host of abilities for everything from creating logos to creating captions and dynamic quotations so what they've said is there is a minor text drawing ability and to use this you have to input your text in quotations and it works best with a style raw or using lower styliz values for example you can put in/ imagine a photo of the text quotation hello world written with a marker on a sticky note and this sort of prompt would get you out something like this so let's have a look at some of the text images coming out here you can see that someone is rendered V6 is the best and this is coming from the mid Journey V6 showcase channel on Discord which is a great place where you can browse through a lot of different different images from V6 and if you also want to take a look just at what is coming out of V6 without having to dive into it yourself there is an option to help rate the images from V6 and this is going to help fine tune and bring V6 to its final form its final evolution like Charizard and you can simply click on the left or the right image depending on which one you think is more beautiful is the prompts that they offer you I would say oh it's a tough one they're pretty close I would go with two then you get another one come up oh I'm blocking it with my little face wow I would say two again two is winning it ooh H the classic cyber Punk dog in glasses nothing says mid Journey more than a dog in sunglasses but I have to say that's a particularly realistic looking pug so let's get back to reviewing the exact improvements another style that is vastly improving in V6 is painted or Illustrated images you can see that the realism of these is just Next Level individual brush Strokes are so well refined these these added drops you can almost see every single stroke that's been inserted to the image and you can see that a user Natalie is also reflecting with her common V6 does so well with paintings every brush stroke is so realistic there are also improved upscales with both subtle and creative modes improving and increasing the resolution by two times you'll see these buttons for these under your images after clicking upscale 1 2 or 3 4 so if you come to an image you've upscaled it then you will have the opportunity to upscale subtle or upscale creative and I'll show you what these look like for this little robot chappie here so another major improvement with mid Journey version 6 is its ability to understand relations between objects for example this user has pointed out that it's impressed that it finally is coherent enough to understand the ball Cube test prompts a small red sphere next to a large Blue pyramid on top of a larger green Cube and red sphere on top of a green cube next to a blue cylinder so what it's trying to achieve here is understand three different objects and how they are situated next to each other and as this user mentioned only darly 3 has been capable of this kind of subject separation before version five could not do it but version six is now capable of really placing objects characters and environments in very specific ways however another user was questioning exactly how far I had come on this subject they suggested I wonder why MJ is still so bad in matter of prompt coherence compared to darly example photo of a cap on top of a table and a dog lying beneath that table well I would say that the prompt here is actually not particularly articulate or precise and I would suggest rephrasing it so that it's very clear what is the relationship of the cat and the dog so I would say photo of a cat on top of a table and dog underneath so let's jump in and actually get to grips with using mid Journey version 6 this is where things get really exciting so simply to use mid Journey version 6 you just have to come into Discord type in for/ settings and from here you'll get a drop- Down menu where you'll be able to select which version of the mid Journey algorithm you would like to use and you'll see that now we have the mid Journey model V6 Alpha available you can simply click on that and then you're good to go now simply type in Imagine and go ahead and enter your prompt but this is where we come to styling and prompting for version six and prompting with version six is significantly different than version five you'll actually need to relearn how to prompt entirely but fear not I'm here to take you through what are the major changes and that is that V6 is much more sensitive to your prompt so you can avoid using junk words like award-winning photorealistic 4K 8K 16k 32k 64k unsplashed dribble beads and that is because it actually focuses much more on uh usable instructive words rather than trying to bring out purely stylistic words it also needs you to be explicit about what you want it may be less vibby but if you are explicit it's now much better at understanding you so that means it's slightly less opinionated in the styles that it's presenting it's giving less of its own Vibe and relying on you to Define exactly the type of style that you want in your images is now if you want something more photographic less opinionated more literal you should probably default to using style raw which brings it back to a very realistic photographic style and if you're using stylized and you're using a lower value then you're going to get better prompt understanding now on the flip side if you use a higher value with V6 you're going to get much better Aesthetics so this image is with the prompt using low stylize value and this is an image using high stylize value they are the main changes to prompting in V6 that you have to be more specific more explicit you can cut out all of the junk additional words that you throw in at the end of The Prompt and you can experiment with whether you're looking for a a more stylistic image or if you're looking for something more realistic you can uh tune that using either raw or styliz parameters it's also useful to jump into the prompt chat channel on Discord to see what other people are saying on this and here you could see people experimenting with how to get particular effects for example how to get some pixel art video game style or looking at Watercolor style images and it's a great resource for asking any questions that you have about specific types of images and we can take a look at the difference of the upscales this image is the creative upscale and this one is the subtle upscale and you can see that there is pretty small differences between the two of them say uh the this one has a lot of very fine detailed glimpse of Light added in which is the cre the subtle upscale So currently the following features arguments are supported at the launch of version six which includes AR which is aspect ratio chaos which is the chaos Factor weird is how weird they are tile means repeating patterns sty styze is the amount of stylization or opinion that mid Journey applies style raw brings it back down to a more raw photorealistic minimal approach very subtle gives you variations on your image generation very strong gives you very strong variations remix allows you to change a prompts that you've put in blend allows you to blend images together and describe allows you to describe an image that you input so you can get out different prompts which are all very useful tools so you can see here in some of these images the the realism the detail is taken to the next level you can almost see the individual hairs included on this person and the imperfections like this subtle acne scarring is absolutely Sensational but it's important to talk about what are the limitations of version six this has just been released and there are still a few issues this is an alpha test and things will change frequently without notice so make sure that you're not expecting this to stay the same forever because it is likely that it will be completely different very quickly so initially speed image quality coherence prompt following and text accuracy should improve over the next few weeks as they gain more data from people rating and responding to the quality of images coming out it's also important to recogniz that V6 is much more powerful and because of that it's more expensive than version 5 but they will get faster it will optimize and relax mode is already supported and we're also going to get some of the other features from version 5 coming out in the next month which include panning zooming varying region tuning describing which means that we don't have in painting yet but it will come soon this model can create much more realistic imagery than anything ever released on Mid Journey before and it is a huge leap forward yet again and I believe that it is the number one AI image generator out there creating more realistic more detailed and better quality images than any other art generator for a little bit more background about version 6 it is the third model trained from scratch on their AI super cluster it's been in the works for nine months and versin six isn't the final step but we hope you all feel the progression of something profound that deeply intertwines with the powers of our Collective imaginations so let's dive in and actually compare some of the same prompts in version five and version six it's important to note that these prompts are the same and we are not reflecting the updated effect of prompting inside of mid Journey but we are looking at how different styles compare in version five and version six so the first image that we'll take a look at is this fantasy style image of a woman and in version five you can see there is a very good coherence these are aesthetic but they're lacking some some life some opinion and in version six you can see that the the details have increased that the expressions are more realistic more lifelike now let's take a look at another example this is a glowing portrait of a woman in version five and in version six you can see that the specific renderings of small details is vastly improved you would say that this is a very effective image but yet you can see where it is lacking detail in this area on the frontal part of the face and then comparing it to the new version where the the highlights are so mesmerizing they they're simply outstanding beyond that you can see that the proportionality the coherence of the entire image is it's more structured it's more reflective or the true nature of a face where this is a very very oval circular form where here we have real evidence of bone structure as well as this wonderful depth of field where you can see the focus is completely on this first eye and even the the second eye is dropping out of focus and it's these subtle details that are what sets version six apart now let's take a look at some text uh version five could not handle text at all this was from the prompt for getting a logo of V6 and you can see that it didn't even get anything resembling those two forms together however with V6 you can see that it's handled this incredibly well and you get real forms that are readable almost every single time now let's take a look at a couple of other prompts this is a quirky Google writing robot colorful cinematic and you can see the original V5 has yeah it starts to lack some coherence inside of the details where things get a little bit BL Le a little bit confused where they are imitating what something actually looks like whereas in version six you have much clearer coherence in details in each individual knob in each individual button and it certainly is an improvement again with this cyborg there is a very mid Journey esque image from version five this is some of the sort of classic style that you expect from version five whereas version six has really again taken it to the next level the realism in the highlights and Reflections on metallic surfaces particularly astounding as well as the coherence inside of the small details if we look at something a little bit more abstract this was the prompt obviously I've still got ridiculous words in there like 16k and this was the version five image you get a very beautiful glowing aesthetic gradient background and then with version six it just takes it to another level there is something about it that's warming my heart it is the depth of these images is one thing that is really improving and not just the depth the depth of field the focal points within each image and once again you can see that illuminated with this Pixar style image this was the prompt this was version five and this was version six and you can see once more there is it's simply at the same standard as a Pixar film animation is all I can say and it really feels like she the character is looking at something off screen they are in genuinely interested in something whereas this one it feels that their expression is more vacant that the eyes aren't quite aligned in a natural way and beyond that you can see things that are very important such as individual strains of hair start to show up in version six you can see this this one Strain coming off of her Bob to the left and these ones coming down from her bangs whereas here in version five you can see the hair is just one solid block of meshed together brown paint now here is a a background style image of a interior and version five put out something pretty decent there is something that still belies its quality there is it it's lacking some some unification whereas version six uh brings in once again more a more realistic depth you can see how how much flatter version five feels than version 6 as well as the the way that light falls on different surfaces for me these are some of the biggest improvements the subtle treatment of details how light falls effectively on different surfaces and depth of field now looking at something a little bit more graphic this graphic tattoo design once again using the same prompt version five creates a a very reasonable Vector style tattoo but version six you know it just absolutely smashes out of the park really it's like version five was a very good designer that you might hire or fiver and you'd be pretty pleased with their work but you're like hey it's it's not world class but then version six like wow this is like a pro this is serious Pro level it's like this was a a recent graduate from a relatively good University and this is someone who's been honing their craft for the last 10 years that's how I see the differences in these so that's what we have in the comparison you can see the to summarize the main differences are the ability to use text better detail increased realism with light as well as depth of field let's take a look at what we can expect next and we can have a look at a few of the micro Poes that mid Journey has been posting in the community and this will give us an idea of what it's likely to prioritize as the new features coming up in the next few months and you can see that it's very likely that we're going to get in painting coming soon based on this poll which gave the options of Zoom pan in painting and ups scalers and style tuner and overwhelmingly people want in painting and again there's another similar poll and the results were exactly the same now another thing we can look forward to in 2024 is mid Journey video mid Journey has mentioned that they've acquired a incredibly good data source of video to train their own video generators and with other AI video generators like P Labs Runway and kyber who have all released major updates in the last week and taken their video AIS to the new level of Excellence we've got a lot to be excited about I mean have you seen some of these AI videos like [Music] this one simple and interesting thing to do is actually to take mid Journey images put them into run away ML and use the motion brush tool where you can animate small parts of the image so you can take an image like this and add subtle animations to give an effect like this so I am blown away by the quality of mid joury version 6 I'm very curious what do you think do let me know in the comments and thank you very much for watching I hope to see you again in another video If you do there's another one here that you can watch and most of all have a delightful day
Info
Channel: Samson - Delightful Design
Views: 50,853
Rating: undefined out of 5
Keywords: ai, samson vowles, midjourney v6, midjourney version 6, midjourney version six, midjourney, version 6, v6, v6 tutorial, midjourney tutorial, midjourney v6 features, midjourney v6 rumors, midjourney v6 release
Id: OqjeLUZDyqI
Channel Id: undefined
Length: 20min 27sec (1227 seconds)
Published: Fri Dec 22 2023
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.