10 Secret MIDJOURNEY V5 Tips And Tricks (How To Use MIDJOURNEY V5)

The Midjourney V5 has just been announced, well, just been released. And I'm going to be getting into every single new feature that there is and some tips and tricks that you should be doing, especially since Midjourney is evolving pretty quickly and you don't want to be left behind. So a couple of hours ago on the Mid Journey Discord, they actually posted this message and we're going to get into this right now because it is very, very interesting and very, very awesome. So you can see right here that V5 is currently active. So if you are a mid journey subscriber, V5 is currently active and you can use it right now. So some of the things that they did add with was, as you can see right here, they said a much wider stylistic range and much more responsive to prompting. So essentially what this means is that if you write in a prompt now, you don't need to add as many details, such as 8k octane render. When I was in the Discord, listening to them talk about, you know, the kinds of prompts that you want to use, you don't want to add 8k render, all of that realistic stuff, because Midjourney now actually does realistic images and we're going to get into that. So they said much higher image quality, 2 times resolution increase, which is very, very true. But more, more, more detailed images. Very, very true. Supports tiles. So previously, you know, for game designers, for people who design textures for. Maybe you're designing a game, maybe you're designing a pattern for a wallpaper or something, it supports the tile for seamless tiles, although it might not work all the time because they do say in brackets that it is experimental. They export aspect ratios greater than 2 to 1, which is also experimental. And they support the waiting and image prompts versus text prompts. So that's also good because it allows you to specify exactly what you want. So what I'm going to do now is I'm going to show you some of the examples as to how different it is. And I do think that right now the reason Midjourney is really good is because I think what, what I will be doing is I will be using version 5 and version 4 together, because both of them give very different outputs based on the same prompt. Version five being more realistic, version four being more artistic. Okay, I'm gonna show you guys exactly what I'm talking about that. So let's get into some prompts right now. So before I go into some of my prompts and exactly how these specific examples, I want to show you this thread quickly. So this thread by story by Phil is actually really cool, but this one basically shows you mid Journey the best. Okay, so it says street photo in New York, medium short, natural lighting, short on a Fujifilm. Detailed realistic environment, cinematic aspect ratio 2 to 1. You can see that on version 4. This still kind of doesn't look realistic. I know that you might think that this does look very realistic, but there are small nuances in this image that shows me that it's not realistic. Okay, maybe it's the skin underneath here. Maybe it's just, you know, the subtle details in this that just show me that it kind of looks like just kind of fake. Okay, I know it does kind of look realistic for an image generator, which is absolutely mind blowing. But when we switch To MidJourney version 5, you can see that this genuinely looks like a real image. The textures, the, the details, the roughness in the skin, the small, just. Just the way that the light bounce in the fog. The simple, the way that the light is, you know, the way that the Gaussian blur is in the back, the way that the depth of field is added, just the blur behind the glasses, the eyes, just the detail in the beard, the little, you know, flickers there. Just everything about version 5 is way more realistic, just magnitudes more realistic. So if you're someone who wants realistic imagery, then, then version five is going to be your best bet because this is exactly what you want out of an image generator if you're someone who wants it. Wants it to be more realistic. So I'm going to show you exactly what version 5 is good at. And so here you can see, okay, the prompt that I, I gave in to Mid Journey was. In fact, let me just open this in another tab. So the prompt I gave me Journey was man holding up hand. Man holding up five fingers in the middle of a street. And this was the prompt I gave mid Journey. Now, you can see here that there's many different ones, but I would argue that a lot of these are very artistic. You can see that this one is black and white. This one has a design in the middle. This one right here, it has, you know, someone on the street. This one right here has a man in a hand. I don't know what he's doing, but I did this for several reasons, because number one, I wanted to test the fingers, and number two, I wanted to, you know, just show you the exact prompts. Now I'm going to show you the exact same prompt, but in version five. Okay, so let's Take a look at version five. So with version five, you can all see that this is ten times more realistic. Now, that doesn't mean to say that version four wasn't that good. Of course, it's definitely really, really realistic. But you can all see that right here. What we have is insanely realistic. Look at the detail in the hand. Look at the little details in, you know, in the veins in the hand. Just look at that detail. Look at the details in the knuckles, in the lighting. Look at the details in the back. The Gaussian blur. For example, this man, the way how his hair is a little bit, you know, not lopsided, but just. Just the way it's stroked. Just, you know, his beard, everything just has an extreme ounce of detail. For example, his hand right here, you can see it stretching and moving. It's definitely a lot more realistic. And the reason I say that is because with. With the hand prompt, you can see that here it messed with the colors, it messed with the composition, the designs. In the middle of the hand, there's, you know, a person walking through, there's all these perspective changes. But with here, it's literally exactly what we want. So I would say that, you know, with this new area of prompting that we have, it definitely kind of reminds me of stable diffusion, but I think it's a little bit better because it just works with so many different styles and there's so many different ways that you can alter this. So this was a very basic prompt. There weren't a lot of details. All I said was a man holding up his hand in the middle of the street. And it does that pretty well because this is the middle of the street, as you can see by the road markings. Middle of the street. This guy's in the middle of the street. This guy's in the middle of the street and he's holding up his hands. Okay. And with five fingers. So it's definitely very, very accurate. And I'm going to show you some more prompts to show you exactly more differences. Now, before I show you some more prompts, one of the really cool things about version 5 is that you can. Is that you can actually get these images straight away. So let's say, for example, I wanted this top right images or the top bottom image, and I wanted this guy straight away. I just type in U3 to upscale it, and then I type in U3 and then I get that image straight away. Okay? It's upscaled instantly. Now, if I go to version four, now this is version four. And if I Want to upscale one of these images? Let's say I want to upscale number one. It actually takes some time. You can see that right here it says upscaling image number one. So if you're upscaling an image in version four, it will take some time, okay, because it says waiting to start, it is now processing that image and upscaling the image. But with version five, okay, what's good about it is if you want an image upscaled immediately, you can just simply click that button and you can simply get that image immediately. So if you, if we go up to this image right here, you can see that I can upscale the second one and instantly, boom, I've got that image right there. And it's a very, very high quality image. So with version 5, you have speed that's going to allow you to instantly gain access to your high quality image, which is just going to save you a bit of time on your workflow because this one is still being upscaled. I mean, it might take an extra 30 seconds or so. But I just wanted to note that down because that's a small change that I don't think a lot of people did notice. If you want to use version 5, just type in v5.5 after your prompt or go into your settings and select v5 from the slash settings. Now, me personally, I wouldn't switch to V5 yet because depending on what you want to use mid journey for, there are some cases in which V5 would be better and which V4 would actually be better. And I'm going to get into all of that in today's video. So another example that I wanted to do was some automotive shots. Now, with better prompting, you can actually get higher quality images. But what I wanted to do was just to show you, okay, that with smaller prompts, you're still able to generate a very high quality images. Now, what I've done right here is I've typed in the brand of the supercar, the make of the supercar, and I just typed in driving. Okay, you can see that we got four very decently decent images here, okay? But there are small artifacts in this image that, you know, someone could potentially identify as it not being real. For example, the wheels do look a bit weird on the body. There is just, just a little bit of inconsistencies. For example, the wheels there as well. Just small, very, very minor things. Don't get me wrong, I'm not hating on this at all. I'm saying that there are very, very small things that you could probably realize that this might not be real. For example, the spoiler back there does merge a bit weirdly and the body at the top here. Now let's go on to version five and see exactly what that looks like. Now, version five, you can see that instantly. Okay. Not only do we have higher resolution images, which we can instantly click in, you can see that these images look magnitudes better. They look almost, I would say, 99% realistic. This image on the left here, I wouldn't be able to tell if this was realistic or not, except for the license plate, because it does look a bit blurred. But even then, sometimes in automotive shots, you do have the scenario where they do blur license plates for privacy reasons. So, like I said, this was generated with a very, very short prompt. Small, small prompts, guys. And the reason I've done this is because of mid journey v5, you don't need to add 8k realistic, all of that. If you actually do add 8k realistic, all of that, it actually does the opposite of increasing the quality of your image. It actually does the opposite. So don't do that in version 5, because they explicitly stated that they've already increased the resolution in these images. And you can see right Here in version 4, this is exactly what it looked like. Okay. And this is before upgrades does definitely still look good, but as you can see, it doesn't look as realistic. And with here, all of these are pretty much instant guys. So it's definitely really, really effective. Okay. And I think they may have trained it on more images because a lot of these images that they have just simply look way more smoother. And in order to be able to identify certain vehicles and certain things, it seems like they've gotten this really, really well. Now, another one that I wanted to show you here. And honestly, thanks to this thread, I'll actually leave a link to this thread in the description. The guys at istorybyphil, you can see right here, he just simply used the emoji as a prompt. And this is why I say version four is definitely more artistic and version five is way more photorealistic and way more. I would, I guess you would say, if you're trying to develop photorealism, just use a version 5 because it's really, really something. It's honestly just incredible. So for version 4, all he did was use the prompt of some stars and a leaf. And you can see right here, this is what it created. Definitely. Very awesome. Very nice. Yeah, you know, everything does look really immaculate, but it does look like a piece of art. Okay. It does look like an artwork that someone potentially did paint at a very high resolution. However, if we look over to this, you can see that this looks really real. This looks insanely real and it looks ungodly real. It just, it is unimaginable that this has come from an AI. And it is really, really incredible. Like, just looking at that right now, I could literally save that as a wallpaper for my phone because it is honestly just so peaceful. Like, for example, the lighting that we have, the volumetric lighting, just, just the lighting, the shadows, the way how it's casting everything all on the rocks, the shine, the moisture, it's. It's. It's just incredible. Like, I don't even know how this is done. But yeah, definitely, truly, truly weird. Truly, truly strange. But I would say that this is the biggest difference in mid journey version 4 to version 5 is the fact that version 5 is a realism machine and version 4 is an artistic machine. And I don't want to say that there are certain instances in which V4 is going to be better. And I would say that, don't be confused, it does come down to the prompting. For example, in this one, okay, the prompt here is on the right hand side. It says calm medium shot of an elegant bedroom with a large window, 8K shot from a movie, cinema composition shot. And the only thing I would say that this thread did wrong was the fact that using the same prompt, you know, for different image for different. Now the only thing I would say that is wrong with this is that one prompt for version 4 is not gonna work as well as version 5. Because I would argue that with a version 5 you need less details because it's already going to go for that realism effect. So if you're thinking, you know, which is going to be the best for me, I. For version 4, you definitely need to add as many details as possible. But version five, you don't need to add as many details because it does kind of reduce the quality. I know that might sound weird, but more is less with version 5. So you can see right here, this one does look really good. It's like a nice beach thing. But Honestly, for version 4, this one actually does look a little bit better in terms of how realistic everything does look. But like I said, guys, it does depend on what you want to use this for. I think that for many of those seeking the artistic root, I would say that version four is very, very good. But for realism, version five is truly insane. And I would say for realistic people, which is definitely going to be something that people do want to use. I would say that version 5 is really, really incredible. So one thing I do want to let you all know, this is an alpha test, meaning that this is not the full version of version 5, meaning it has been released. But of course it's released in its developmental stage, meaning that things may change, we may need to add more details in the future. It may stop being as realistic you. You may be able to change whether or not it's artistic or realistic. But this is just the current state that it is in now. I would recommend playing around with this to see exactly how it works and exactly what you're going to use it for. But it definitely is much, much better than version four. And I would say that it's definitely really, really effective at what it does. And I would say that I'm not the best person when it comes to midjourney. There are some people who are able to get these levels of details, which is truly breathtaking. And I'm pretty sure that if you go on the subreddit it's. And if you go on the mid journey, if there are any prompts in the comment section below, I will try my best to pin them if they generate some amazing high quality prompts. Because as certain prompts become available, for example, in V4 there was a sign still prompt which generated very, very high quality images. I'm pretty sure that there are going to be a lot more floating around on the Internet. Now, something that I did want to also add, okay, and something that I may have forgotten to add in the video was that a new feature in version five is that Midjourney has, I guess you could say infinite prompts as infinite any aspect ratio. And this is honestly breathtaking because you can see right here that this user on Twitter actually tried an aspect ratio of 100 to 1. And you can see that it is a skyline of, I think it's America somewhere in America. You can see right here that it did the entire skyline, which is honestly truly insane. So I would say that this is the biggest feature is that there's definitely a lot more flexibility. Definitely something that you can try out. So I'm not exactly sure what this prompt is going to generate because I did just typed in a McLaren 57S racing with an aspect ratio of 100:1 it in version 5. But I guess we're gonna have to see exactly how this works. Now another thing to remember is that version 5 images do take a lot more time to generate. I'm guessing just it's because there's a lot more detail in there. So I would say that definitely understand that these images take a lot more time to generate, whilst the four images are a lot quicker. But you can see right here that these are the images that are generated for me. And honestly I would say that these either look like they're from a game or from a real life shoot or something. So it's definitely very interesting to play with different aspect ratios. I don't think Midjourney is accustomed to certain aspect ratios, but maybe if you wanted maybe a YouTube banner, maybe a Twitter banner, maybe you wanted a profile picture, I'm guessing that you could experiment with these new aspect ratios. You could pretty much try any number and I'm pretty sure that it would work. So definitely understand that Midjourney version 5 is absolutely insane. Now if there's anything that was missed, leave a comment down below and I'll be sure to see if I can get to it. But it just goes to show the many different capabilities that this does have. So definitely check it out. It.