Getting Started With Midjourney: Creating Advanced AI Images

Happy 2026. Hey, how's this for a New Year's resolution? How about you? Master AI marketing? The AI Business Society gives you live training with experts who are getting real results. No hype, no fluff, just proven workflows that work. And you've got access to an incredible library of training from many of the incredible guests that you've heard on this show. So secure your AI transformation by visiting socialmediaexaminer.com AI welcome to the AI Explored podcast, helping you put AI to work.

And now, here's your host, Michael Stelzner.

Hello, hello, hello. Thank you so much for joining me for the AI Explored podcast brought to you by Social Media Examiner. I'm your host, Michael Stelzner, and this is the podcast for marketers, creators, and business owners who want to know how to use AI. Today, I'm excited to be joined by Jeff C. And we're going to explore how to create advanced images utilizing Mid Journey. You probably have seen all these amazing images. They're not the ones that come out of Chat GPT and you've always wondered, how in the world do I do that? Today is the episode for you. Also, if you're new to this podcast, be sure to follow this show so you don't miss any of our future content. Let's transition over to this week's interview with Jeff C. Helping you simplify your AI journey.

Here is this week's expert guide.

Today, I am very excited to be joined by Jeff Z. If you don't know who Jeff is, he is a content strategist who helps creators get more from their existing content. He. He's also host of Creator News Live, a weekly live show. And his forthcoming course is focused on helping creators make great images using AI. Jeff, welcome to the show. How you doing today, man?

Mike, it is so good to talk to you again. I'm so excited to be here. It's been too long.

Yeah, it's great to have you back. Today, Jeff and I are going to explore creating advanced AI images with Mid Journey. Now, Jeff, there's going to be some people from the Social Media examiner crowd who are going to know who you are, but why don't you share a little bit of your journey and how you got into artificial intelligence and start wherever you want to start.

Yeah, so I've been doing visuals and marketing for over 25 years, and I started as a web designer. I did those Flash websites. You can thank me later.

I did some of those, too. Yeah.

Remember those? Then I got into video very, very heavily. And Then social started and I thought, well you know, I'm telling my clients to do this, I better start doing it as well. So I started a blog and it was actually I was driving back from Kansas. My parents used to live in Kansas. We drove, we're back from them. And I heard this lady named Cynthia Sanchez on this little podcast called Social Media Marketing podc. And it was Mike was interviewing her about how Pinterest could drive traffic. I thought, hey, I've got a blog, I need traffic. So I started doing that. It kind of grew from there. I created a post called mainly Pinterest Tips number one where I shared a secret board with my daughter and how all that worked. And it kind of took off and it was in the wild wild west days of Google. I was able to turn that into a live show where we had like four other people who all had way bigger followings than I did. And, and I got to interview people like Guy Kawasaki and the producer of Pretty Woman and Under Siege. And it was just a crazy wild time with live video. Well that led me to being interviewed by Mike on his podcast about Pinterest. And so that led to me being actually working for Social Media examiner running their Pinterest. Then it was Instagram and then it was help producing and host live shows. And I worked for a social media examiner and spoke at the conference room for about five years. Then started Social Media News Live which is going to be Creator News Live and been doing that as well as many other projects, producing courses, consulting and. And the studio side of the company produces podcasts and videos for people like Guy Kawasaki and Agora Pulse. And in the midst of all that process, when I was started speaking at the conference about Pinterest, I always, you know, even the last couple years at Social Media examiner, at the social media marketing world, I kind of started to talk about tools for non designers to create great looking images. We combine different tools to make different effects and all this stuff. So I've always had this interest in helping people who aren't designers create images. So it was so natural for me when these AI tools started emerging to really jump into it. And then of course for guys podcast we started using Descript and that's all AI based way back when we first started over three years ago. So that's how it all kind of came together. And you know, both you and I, Mike, we kind of started way back in the day, you know, like I was using Premiere 2 and Audition like 3.2 or something like that. But now AI has totally changed how I produce content as a creator. So it's so much faster and it's such an exciting time to be in.

Well, and tell us a little bit about what you're doing now, specifically these days with AI.

Yeah. So once again, using it all for my own stuff, repurposing my content, using all the tools and, you know, slicing things up. But really, we've created. All of our graphics now are built in AI. So if you look at social medianewslive.com, you'll see all of those are created with AI. They all fit a certain theme. I'm kind of have like a 50s retro theme on my site. And the same thing goes with all of our social media posts. So it's. And it speeds everything up because we have. And we'll talk about it in the show today, like prompts that we use, the tools that we use. And it really does speed up the workflow. And it's really, really cool how you can optimize now and create things that you've never been able to create before.

So there's a lot of people who are creators and marketers listening to this show or watching this show, but maybe they have not begun really getting into the image side of things. So what do you want to say to them? Why should they consider using AI tools to create images?

So here's an example. I've been speaking at both of our friends Lou Mongello's event. He has one called Momentum every year. So last year I was able to create my entire slide deck with AI. So, you know this Mike. So, like before, we're always trying to look at the stock Photoshop sites and we're trying to kind of pigeonhole something that may or may not work, but it's the only thing we have to make that point. But this was the first time that my entire deck was original images that were mine that I had created. That was the perfect point or the first perfect laugh that I wanted to get. And it all fit together. And that was the only way I've ever been able to do that was with AI. So as a creator, it is super exciting because now you can have the same look and feel with totally custom images. So it is so liberating as a creator. And, you know, I know a lot of people feel like I do that you used to be able to. You would have it in your mind what you were wanting to put in Photoshop or Illustrator, but you could. You didn't have the skills to make it work. Right. Well, now with AI, you're a lot closer to being able to take what's in your mind and put it on the page, the digital page, if you will, than ever before. So your images, Mike, like for your banner graphics, like you have on your website, you have this incredible illustrator who's been doing your stuff for years. And it's your look and feel is perfect. A lot of us don't have the skills like your illustrator does. And AI has been the first time we've been able to do this in my lifetime that I can actually take what I'm thinking and actually make it be an image on the page.

Well, and you were telling me about how you were using these images inside of actual videos themselves, not just for thumbnails. Talk about that for a second.

Yeah, so we can even use it as B roll. So a lot of times the same thing, like with stock image sites, you're trying to find something that would kind of, you're forcing it to fit. Well now you can actually create those for like your B roll that you need for your YouTube videos or for your shorts or anything like that. You can actually create an image like let's say in mid journey and then just do a slow pan in it and a slow pull away. And you could have the exact thing that you're wanting to convey to your audience and not trying to get one of those stock photos that just is like cutting away to a guy shaking hands or whatever you're trying to do for. But you can get really, really specific and really, really put your point across the way you want in that video.

What I love about, and I'll just tell you my perspective on this, I used to own a creative agency and I do art direct illustrators. We have one in house, full time designer and then we have another contractor. What I use AI for is to come up with concepts, right? So I can come up with visual concepts of kind of conceptually where I want to go. And in the past I would have to search the Internet and find images that I thought were cool and give them as concepts. In addition, I have company meetings every other week and I use these images in my decks to kind of conceptualize, actually convey a concept that I'm talking about. And obviously that's the easiest way to do it. We use them in all of our thumbnails as background images. You just brought up the fact that you can use it as like inside of a video using the Ken Burns effect, right, where you're doing that slow pan. So there's so many applications and you do not need to be a designer. I am not a designer, you know, at all, but I know how to art direct. And I think if you know how to art direct, which is providing a little guidance to the AI, you know what I mean? You can do amazing things. So we're here today to talk about Mid Journey, and I want to explain everybody why Mid Journey. I know that a lot of the people I've had on my show are doing really advanced images that I cannot do with the other tools that I'm using, like Dall E3 and Mid Journey, to me, seems to be way more sophisticated. So let's just start by talking about. Because probably everybody who's listening has seen these images that come from Chat GPT that are powered by Dall E3, but maybe they don't understand, like, the. The problems and challenges with those versus Mid Journey. So let's just start by talking about, like, what makes Mid Journey different.

Yeah. So this is my opinion. So a lot of people like Dall E3 and. And there's. There's some really great prompts, and you can make some great images with that. So what I would say is, if you are on a budget and you can only afford one, being that you have image creation built into Chat gbt, I think that's a great deal. So. But when you look at the pros and cons of the image generators out there, in my opinion, Midjourney is a clear winner. And the reasons are, is there's just. There's more options, there's more styles that you can do. You can just tweak it a little better than some of the others. For me, Dall E3 is. That's what's built into ChatGPT3. It's a good option. But it's like, if you remember, Mike, when Canva first came out, you would see images in the feed and go, oh, there's a Canva template. Oh, there's a canva template. And you could just kind of pick it out. I'm feeling the same way when I see some of these Dall E3 images. A lot of it has to do with. They're just not digging deep enough into the prompting. And we'll talk about that a little bit later. But our goal as entrepreneurs and marketers is to stand out. Right. And Midjourney is the one that I would point people to just because of the way that you can tweak it. I just think it has great image quality. They just released 6.1 last week, and that lets you like render even more finer detail like reflections, and some amazing things, like prefer way better than you've been able to do before. So that's the thing, I think that is the reason it stands out. It just, it gives you more options. You can. When we'll talk about some of the prompts you can pull back, you can go to the left or the right, you have more options than some of these other ones you have out there. Here's the downfall, I think, of Mid Journey. I don't think they even have a paid plan anymore. If I. They may have some free credits you get when you test it out, but it's. It's a paid option. The other things that doesn't do very well is text. A lot of the newer options out there can do really great text on images. But a lot of times in Mid Journey, and with this new version, it's better, but it's still not there yet. It'll put up weird words or it'll like skip a letter or do double words sometimes. So it's still not there for. If you're wanting to create like quote graphics or something like that. It's not the best option in my opinion, for that. But for everything else, if you. And there's a lot of options too, and you know, have it create a blank canvas and then put your text on it in some other option like Photoshop. But I mean, it just stands out. It's the best one out there. And I think the biggest reason people don't use it as well is because of discord. And we can talk about that in a little bit too.

Well, and I want to add, I mean, as a Guy who uses ChatGPT to generate images almost every day, mostly for our thumbnails, and I'm going to say this, and this is going to resonate with some people, it's infuriating sometimes. Like yesterday I, I said, okay, show me a picture of a person pulling a slice of pizza with the cheese spreading out, but place the person on the left. It kept putting the person on the right. It didn't matter whether I placed it on the left. A few weeks ago, I said, show a picture of a person taking a selfie. For whatever reason, the selfie was always horizontal, never vertical. No matter what I tried, it wouldn't put it vertical. And when it did put it vertical, it. It flipped it so that it wasn't the selfie, it was the opposite of a selfie. And like, these are the kind of things that kind of are infuriating to us. Who are trying to create images. I'll say, put no text in it, and it'll put text in it. And then I'll say, you put text in it and it'll say, no, I didn't. You know, and these are the kind of things that just make you want to pull your hair out of your head when it comes to Chat GPT, I would imagine you have more granular control over these kinds of things with Mid Journey. Is that true?

Yes, for the most time. But it's, you know, AI is frustration. Sometimes you can still have it screw up and you're like, no, that's not what I meant. Like, it's getting smarter and smarter so you don't have to argue with it anymore. Midjourney now uses a lot better natural language, and we'll talk about it a little bit with prompting. But it can get frustrating. And one of the things I tell people is that's why I push for people who are serious about this to use a paid plan, because you are never going to tell it what to do. And it will give you the right image on the first time. Like, we call it rolling. Like, because you're rolling the dice and you continue to roll and roll and roll until you get what you want and then you constantly refine that. So it's more for me than like Dall E3. I just think it works better. But there are times, like, you're like, no, that's not what I wanted. And you just have to either. When that happens, I usually take what I was trying to prompt into, like ChatGPT and say, rewrite this in a different way. Because it's usually my fault that I'm not explaining something correct that I. I have in my mind, but it's not doing. And so I tried different ways to actually write the prompt. And that usually fixes it in the long run.

Okay, let's get into kind of what we need to know about starting with Mid Journey. I want to start with pricing. As of this recording, ChatGPT is about 20 bucks a month. It's a flat rate. You can hit a limit on when it pause you, and you won't be able to create any more images. How does Mid Journey pricing work as of today? And then let's get into discords and all that other fun stuff.

So for me, the. The reason I picked Mid Journey is and I play for the higher plan. So I think they have a $30 plan that gives you the commercial license. So all of the paid plans give you the commercial license, which is very, very important to me because I use it for myself, I use it for, for clients. And so a paid plan covers me for that commercial license. So the other thing that's really kind of maybe turns people off about midjourney is that it's, it's run in Discord and you have to go in and set up a server. They walk you through it. It's really easy. They have directions on their site how to set it up. And so you, you create a server, invite the Mid Journey bot to your own server. And, and the best way I can talk about Discord is I say it's like slack for gamers, right? It's like built for gaming. But it's a good skill to learn anyway because a lot of communities are based on there. My friends over at ecamm, they have a great Discord server. But so if this old bearded geek can figure this out, you can too.

So real quick on the pricing. So right now it's about $30 a month for the premium plan, which will allow you to basically use these images in a commercial way. Is that what I'm hearing you right?

Yes. Yeah.

And is there a limit on how many images you can generate over which they start charging you?

So the one I have, which is the $60 plan, which is the one that gives me, it's the pro plan. So I've got what they call 30 hours of fast generations, which allows you to get them really quickly, but then you have unlimited relaxed generations. So that's what I like. I have mindset to be always relaxed. I can wait a little bit like because we're rolling a lot, we're seeing things, I'm fine with doing relaxed generations. And I love that unlimited plan because the pro plan, also that $60 plan, that's where it starts. And then it goes up to their, like, I think their enterprise plan is, it gives you stealth image generation. And this is what's important to me is because a lot of people don't know this, but if you're creating stuff on Mid Journey and you're not on the stealth plan and actually set it to be on stealth, everybody can go and look at your images. They can look at what you're prompting and how you're doing that. And I'm pretty an open guy, but sometimes I'm doing stuff that I'm experimenting with or I want to try something. I, I just don't like people looking over my shoulder. So I pay for that specifically so I can have that stealth image generation.

Love it. Okay, so that's really good. So what I'm hearing you say, unlike ChatGPT, where nobody sees your prompts at all, with Mid Journey, it's public record, it's out there. People could figure out how to find it. And if you've developed some really cool prompting that you kind of want to be your proprietary prompt, it's worth paying the 60 bucks a month. Discord. So for those who are not familiar with Discord, you're right, it is a lot like Slack. And my understanding is you're interacting with a bot, right? So it's kind of like the equivalent of messenger probably is the best way to explain to our audience, right? You're like, you have this message thread and you're interacting with a bot and you're giving it a prompt, and then it results in an image that you can download right there. Is that correct?

Yes, that is correct. So what you do is once you invite it to your server now you can actually go inside of their rooms and create stuff as well. But I just like it all on my own server where it's all me, it's all clean and I just have it there and it's all, all in stealth.

And for people that are freaking out about server, let's be clear, you don't have to really set up a real server, right?

No, it's. It walks you through it. It's a free account for Discord. And they just call it a server. Yeah, they just call that Discord Server because that's where your stuff lives. And then you can join other servers. Like I'm part of a Descripts Discord server. I'm part of eCamm's, where I can go in and do the message boards kind of stuff. But my own private one is where I run Mid journey from. And so it's really easy. You do forward slash, imagine, and then you do your prompt, and that's how you start generating images. When you do a prompt, midjourney brings up four images and it's going to be square by default. And you can look at those. It gives you four different types of images and you can actually go through. And there's a, underneath each image, there's a U1, U2, U3, U4, and then they have for. That stands for upscale. And then there's a V1, V2, V3, V4, and that's a variation. So if you like what it has, you can upscale it right there and you can actually go and. And like we talked about, you can pull out, you can, you Know, make it a square image. You can do some other. You can inpaint with it if you want, which just means going in and fixing some flaws that you see in there. But you can also, if you like an image, have it do variations on it. So it keeps it in that kind of thinking that it was doing and gives you kind of based on that original image that you selected that you kind of like. But let me see some other options so you can drill in really nicely. And that's one of the things I like about midjourney is being able to, if you see something you like, is to upscale it right away. You can then download it and use it, whatever you want. Or you can say, give me some variations on that. And then you can once again go through that whole process, like, oh, I like that image image. I want to do a variation on that one. And they even have now with version 6.1, you can have a creative upscale. You can have, you know, a subtle upscale. So there's a lot of things that you can do inside of Mid Journey that's very, very cool. One of the things that I would also tell people to start learning is at the end of the prompt, they have things where you can change the parameters, like dash, dash, ar, which stands for aspects ratio. And then you can put 16 by 9 or 9 by 16 or one by one. Whatever. Remember, it defaults to square, so you need to put those. If you're wanting to make an image for, let's say, you know, your Instagram, you want to be 9 by 16, you got to make sure you put that parameter in there. There's other parameters that are good to know, like dash, dash, style, raw, which actually gives you a more photography, like, it's not so much beautified. It doesn't have any automatic beautification applied to it.

It's like no compression, really.

Right.

Isn't that.

Yeah. And it looks like more like a realistic photo, so that's fun to mess with. Another great one, I think Mike Russell talked about it when he was talking about his thumbnails on your show is there's one called cref, which actually you can upload an image and use that as a character reference, which gets close, it's not going to be on. And, you know, if I uploaded a picture of me and then it would give me a guy with a beard kind of my style, but it wouldn't be me. And that's what Mike does. And he puts the face swapping stuff on that you talked about.

All right, cool. A Couple quick questions on this. First of all, for those that are not familiar with Discord, Discord is something you can have on your phone or on your computer as a standalone app, or you can work through a browser. So that means you could generate prompts when you're out and about and you get inspired. And I love the fact that you can have a character reference, because you could use your phone, potentially take a photograph of something and upload it as a character reference, could you not?

Right. And you can also use characters that you create on mid journey as well. So let's say Mike, that you went in and you played around, and you got scout as a 3D, really cool character. Right. That you have for your mascot. Well, then you could put him. You could use him as a character reference and put him in different scenarios inside of different things that you wanted to showcase him.

So character reference has to not do with the background. It has to do with. Subject is.

Yeah. The person. Yeah. So it's actually the subject that you upload the photo and say, this is the character I want you to create more like that.

Oh, so you could do yourself too.

Right? Yes. It's close, but not always. So that's why with Mike, he can get a guy who looks similar to him, and then what he's able to do is face swap that, because rest of it already matches really closely. And that just puts it, you know, that little twist on it that makes it, like for his YouTube thumbnails.

Okay.

So you could do that same kind of thing with Scout if you wanted to.

A couple other questions on the variation. Let's say I found an image I really love, and I like everything about it, except for, like, one thing. With the variation, is it going to completely redraw the entire thing, or is it smart enough to know just to focus in on that one thing that I want to change?

There's two, like, weights to it. There's one you can go into a variation where, you know, you have the variation at the bottom, where you'll give you four. Four more images in that variation, and you can tweak that a little bit. But they also have. Once you upscale an image, you can say heavy variation or really subtle, where it'll maybe make some more details on the. On the face or change some things. Like, I was doing one with a robot painting on a canvas, and it kept most of the robot. It looked the same for the robot, but the canvas artwork kind of changed subtly. And so you can do things like that for something. Like if you don't like something totally, but you love the background. I would use something and take it into like in painting inside of there and see if you can fix it that way. Or you could take that original image, download it and put it to like Adobe Photoshop and do that generative fill and fix it that way.

Quick question. How many podcast episodes have you recorded? How many webinars have you hosted? How many client calls or team meetings have happened that are now just sitting there in your files? Here's the big unlock. Those aren't just old recordings. They're full of insights about your audience, your brand voice and your customers actual objections. You just need to know how to extract them. That's exactly what AI educator Erica Stanley is teaching AI Business Society members on January 8th. By the end of this training, you'll have a repeatable workflow that turns every past recording into into a content and data powerhouse, one that keeps paying dividends long after the mic's been turned off. This training is exclusive to AI Business Society members, but here's the good news. You can join us right now by going to socialmediaexaminer.com AI and start transforming your old content into new marketing gold. SocialMediaExaminer.com AI to join right now. We're going to get to inpainting in a little bit, so let's talk about prompting, because I think the key to this entire thing is getting these prompts Right. Right.

Yeah.

So talk to me a little bit about like what do we need to know about prompting? We've already talked about U and we've already talked about V. U is upscale, V is variation. We already talked about dash dash, which is the minus sign if I'm not mistaken, AR aspect ratio. And we've also talked about style, raw and cref, which is character reference. But let's just talk about like general prompting tips on how to use mid journey because it seems to be a little more strict obviously than ChatGPT. Dalle3 Right.

Yeah. So they've changed some things now so they've actually started using more natural language inside of the prompting. And it used to be a lot of people with keyword stuff in Min Journey because they thought it would be that would be the best way to do things, but you don't do that anymore. So you don't use keywords like 4K or ultra realistic. You just really don't do that. And one of the best ways I've found, and this is, I can't remember who said it on the interwebs, but they Said it's a. They'd like to use a 4W 1H and that stands for who, where, when and how. And so what that means is so like who is the main subject? Right. It's the theme of the. The image. Where is like, where's it the scene and where's the settings of the scene? When, like the time or the lighting of the day? How is kind of like the style and rendering detail. So let me give you an example of one that, that I use. So I like Alex Ross. He's a great comic book artist. He has a very distinct style. Really like it. So a prompt for me would be a comic book image in the style of Alex Ross of a bearded man sitting in a recliner smoking a curved pipe in the 1940s. High contrast. So that's kind of using that prompt together.

What does high contrast mean?

So it's going to give me like almost like black and white, but not really black and white. So it'll be really like sharp. Sometimes it'll interpret that as even some ink drawings, like pen.

So that high contrast is kind of the style you're telling it to draw.

The thing in at the end. Yeah. And mid journey is now good enough that you can kind of mix those up. You don't have to have it exactly in that order that I found, but it's good to have that kind of pattern when you're thinking about it. Here's the thing. Many of us don't know artists or styles. What you can do is study those prompts in the feed. You can learn what styles you like, what artists you like. I have a whole notepad entry that I just keep copying prompts down and, and things that I like and, and have that kind of running in the background as you start to learn what you like and what kind of styles you're wanting to apply. So example I had, I wanted to do something in the style of the Jetsons. Remember that old cart? And I was like, I want it in the style of the Jetsons. And I found out that's called Googie architecture. Not Google, but Googie. G O O G I E. Okay. And so then I was able to apply that to the style that I'm trying to get. And it made it. It made it the way I wanted. So that's one of the things I tell people to do is go and look. And you probably don't know what to call it. You have it in your mind, but research it and figure it out and then you have it and then you can Use it for all your images down the road. And I actually use ChatGPT to help me come up with abstract. A lot of times. We'll have something like, okay, here's I. I had an episode of. It was called the Power of Sound. I had this voice, voice artist Jody Krangle come on my show and talk about the importance of sonic branding. And I couldn't think of, like, how. How do I portray that as an image? Right. Like, it was just weird. And So I asked ChatGPT and it was able to come up with. Try it with. And it wrote me a prompt and I was able to modify it a little bit with my style and I got what I needed.

Is it not possible that you could have snagged in a screenshot of the Jetsons and popped it into Chat GPT and asked it to describe the style of the art for mid journey? Could you have done something like that, hypothetically?

Yes, that. I do that a lot. Midjourney has a great. It's a. Instead of going forward slash, imagine you would do forward slash describe. And you can upload an image and it will tell you. Now you have to pick and choose kind of what you want. And you're like, okay, because it'll. It'll tell you everything. It'll give you a long list of what it sees. And so that is very, very helpful as well, is when you're trying to learn your prompting style and how to prompt.

Talk to me about this free flow AI tool that you were sharing with me when we were prepping for this and how it can help people if they don't know what they're looking for.

Yeah. Think of it as a visual dictionary of prompts. Right. So if you go in and you type, I want to find out, you know, retro images or steampunk, like, it will. You can do a search for it and it'll give you all these different styles and it'll give you the prompt to use. It'll actually now do the character reference on a lot of these things. Where you can enter those, there's specific numbers, like a code. If you remember canva filters, you can copy and paste those. It's kind of the same way with this AI stuff. And it has that as well. It's all broken down, it's all visual. You can see the styles that you want and it even walks you kind of through how to, you know, craft your style. And you can copy and paste it and put your own stuff in there. It's amazing.

So it's free Flo AI now there's also some plugins that you use for chatgpt or custom GPTs. We should talk about this one. I'll try to pronounce it glibitree. G L I B A T R E E Art Designer. What does that do?

So it's really cool and it's great when you once again like you're trying to figure out how to do an image. Like a lot of you can easily do like I want a cat riding a unicycle, okay, that's gonna give you an image. But when you're trying to do something like the power of sound or something like kind of abstract, using a tool like this, Glibtry Art Designer allows you to type in what you're kind of thinking and then it will expand on it and will like give you five images in five different styles. And the cool thing is it will actually give you a mid journey prompt at the top of it where you can copy and paste it so you can see what Dall e thinks. But then you can also take that, it makes a specific midjourney prompt and you can take it over to midjourney and enter that and you get a whole nother option that you can look at for your image and your design.

Some of the stuff that we were talking about is things like a lens. Talk about a little bit about like why lenses matter so much and how maybe ChatGPT can help us with some of that.

So I mean I've, I've played in video and some with you Mike, for years and we, we know some stuff but we at least I don't know specific lenses and how they act and all that stuff. And it's easy. You can go into ChatGPT say hey, I want, you know, I don't even know it's called Bokeh or Bokeh that behind me. I just want it fuzzy in the background. And ChatGPT will help you write a prompt. Here's a great prompt that I actually use quite a bit because I think it's great to combine tools like ChatGPT and MidJourney. So you can say, act as a professional photography prompt generator that accepts visual description and outputs a detailed paragraph that I can copy into my diffusion model. Include a variety of photography related terminology including the description of the exact Zeiss lens you use and most importantly a description of the volumetric lighting. Now give me a clear and concise natural language visual description of the following and then you type in what you're trying to portray. That does a great job of giving you those really photorealistic images.

Well, and you said diffusion. Right. But you could just substitute that for Mid Journey.

Right.

Diffusion is just another tool. Is that what I'm hearing you say?

Yeah, yeah, you can submit Journey Diffusion. It will give you a great prompt, even if you use that one. I just said, okay.

So in summary, what I'm hearing you say is that the power of getting great output in Mid Journey really comes from the prompts, and it's understanding. And you. It sounds like the more you prompt, the more you begin to understand, like, the little nuances that matter, like the aspect ratio, the lens, and the other descriptions and stuff like that. And the best way to get started, obviously, is to go look. And I think you mentioned that Mid Journey has their own library. Right. And you can look at their images as well.

Right.

And see what the prompts were behind those. Is that correct?

Yeah, I usually go once a week and look at the. Kind of have a feed of what the best ones are. And usually there's something that you go, I like that. And, you know, this works well. And so I like to go in there and kind of see and. Because that's how I learned artists. I don't know names of artists. I never studied, really art. And so looking at names of artists where you can say, make me an image in the style of blank, that makes a huge deal. And then you can further refine. Like, you can combine two artists, like, you could have two different comic book artists and kind of have a fusion of something brand new. So that's really cool to do as well.

And then this slash describe function inside of Mid Journey, it sounds like I can upload an image that was generated from an AI system or a photograph, presumably. Right. That I took of a piece of art or a screenshot or something. And when we do slash describe, that's where Midjourney will essentially create a prompt for you out of it. Or what is it giving you?

It's telling you what it sees. And so what that's helpful for. For is when you're writing prompts using some of those words in your own prompt.

I see. And can you do the Same thing with ChatGPT? Can you upload the image and ask, have you tried this? And just ask it to describe what you see in a way that I could use inside of as a prompt for Mid Journey? Do you think that's possible?

I believe you can do that, because I know with the desktop app you're able to upload your own images and do that. So you could probably go break it down. What I find really interesting is taking something that's created, like from Dolly and taking it over to Mid Journey and say, make it better. Or like, you know, see. Tell me what you see and then tweak it some more.

Yeah, let's get into that. Let's talk about that. Because so many times, especially those of us who have not dabbled in Mid Journey, like myself, I've not dabbled in it. I get something that I think conceptually is really cool. But what I don't like about it is all the random stuff in the background. Like there'll be these random drawings in the background that seem to have no logical sense. And they're. They're not fully drawn and they're kind of messed up and they're like, something's wrong with it. You know what I mean? But I love a certain part of it. So let's talk about, like, whether it's coming out of Mid Journey or Dall? E3 or some other tool. How do we improve our images and how do we expand or tweak them in any kind of way?

So both Dall? E and Mid Journey have what we call inpainting, which allows you to go back in and tweak section. Like you can almost draw a square and then type in a new prompt. It's kind of like what happens inside of Photoshop. Generative fill. And so this is great for especially. They're doing a way better job now. But I had a image where I was trying to explain it was for a slide, for a presentation. Was like anybody having questions. I had these 1950s looking school kids and they were raising their hands. Well, if you have any idea with AI and generation, you know, image generation, they mangle hands like freakishly horror. They're horrible. And I think I had one hand that was good. I was able to go back into inpainting and fix every single one of those. So those are the types of things.

You can do with Mid Journey.

Yeah, with Inside of Mid Journeys in painting. Right.

Just explain how that works for people who have never experienced that before.

So you. You do your upset. You have the image that you're happy with, except for the funky hands. Right. And then you click on. It's nice. Not in painting. It's called very region is what it's called Inside of Mid Journey. Click on that button and then it gives you kind of a window with your image in it. And then you can just draw a square or you can use your mouse and drag out Kind of the handles to what you want to change. Right. Like for mine it was the hands and so I had to go in one at a time and do that because they weren't all together. But I was able to go back and say fix hands or have open palm or something. Sometimes it takes a little bit of experimentation, but that was able to fix the images.

How long does it take to generate the new image? Is it pretty fast?

Yeah, it's the same as creating the other images that you do in mid journey. Yeah, it doesn't take any longer. It's actually faster now with version 6.1.

Yeah. Because for those that are not aware, you can do something like this inside of ChatGPT. The problem is it's very slow.

Slow.

So you have to select it and then you have to wait for it to generate it. And it's like, I don't know, I feel like it's at least 30 to 60 seconds. Do you feel like it's a lot faster in mid journey or is it.

Comparable even with mine in the free, slower mode that I keep mine set on is that it doesn't take long. I'm never, I'm never sitting there going, oh, go on. It's. It's normal images. Yeah.

Cool. So talk about Adobe a little bit, because people don't realize what you can do with Photoshop, right?

Yeah, it's amazing. So if it doesn't work in mid journey or you're struggling with it, or there's too much, I download it and then upload it to Photoshop and then the generative fill is just second to none. I think in Photoshop I do a lot of stuff with Amazon reviews for products and whatever. And a lot of times companies will send me their image and it's a square image and I'm like, I can't use that for a thumbnail. Right. I had a bunk bed that I was doing a video for. My son was moving apartment, he had one. And so we did a video for it and they sent me a square image, but it won't work as a thumbnail. But I was able to go in and extend the whole room and it made a AI generated room that fit perfectly with the. The product shot that they gave me. So that kind of stuff can save you so much time when you're trying to get something to work for like a thumbnail or something like that. I've had it before where I was trying to have a lady unscrew a light bulb and the light bulb was just looking funky and So I was able to just draw around inside of Photoshop and type in Generative Fill, and it gives you three different options. And if you don't like those, you can hit Render again and it gives you three more, and then you can toggle through whichever one you want. So it is a very powerful tool. Don't be scared to use different tools together, because it just makes your stuff look so much better. If you're struggling with getting an image to work the way you want it to.

Earlier in the interview, Jeff mentioned face swapping. And some of you are like, whoa, what is that? So episode 10 with Mike Russell is what you want to go to on that. Some of the stuff you talked about that Mid Journey can do is like, zoom out and move left and right. Talk to people about that. Because oftentimes, at least with Dall E3, I find that the subject is always in the center of the screen. There's not a lot of space above the head. And, man, if I could have the lens or the perspective change, that would be huge. I don't think I can do that with ChatGPT. So what can you do specifically with Mid Journey on that front?

Yeah, so when you have the image with those four images, after you upscale the one you like, you're able to zoom out two times, you're able to zoom out 1.5 times, or you can do a custom zoom, right? And you also have arrows underneath that where you can zoom out to the left, right up or down. So I always try to generate an image when I'm done creating one, even if it's like, perfect, if it's like 9 by 16, that's going to work. I always do one out because you never know when you're going to need one and move something over or the text is covering it somewhere. If you type a text, you know, box on it somehow. So I like to have an image like that that I can use later. And the cool thing is it actually generates that background around it so it fits perfectly in the scene. Like, I had a cartoon astronaut sitting on a lawn chair on the moon, right? And it was too tight on the astronaut. I was able to zoom out and it gave me the whole vista of the landscape with the astronaut there in the corner where I wanted it to. So that kind of stuff with Mid Journey is really, really cool. You could potentially, if you had something that you Created with Dall E3, take it into Photoshop and, you know, do generative fill on half the image and kind of keep moving it over. But Doing it inside of Mid Journey is really, really a big time saver.

And it doesn't mess with the subject at all. Right. It just fills.

No, it's, it's keeps it the same way. Yeah.

What about changing the angle of the shot? Is that an option or is that not. Do you understand what I mean by that? Like if you have a dead on shot, you want to move the camera up so it's facing more down or move the camera down so it's facing more up. Can it do that kind of stuff or is that a little more tricky?

That's a little more tricky. But what you could do is once again use character reference. If you have those reference you like and going in and there's. We're getting really nerdy here. There's like seeds that you can keep it at a certain seed and then you can go through and like have that same character and then you could tell it say like I want it from a side shot down. And then you might be able to create that same character using those character references and then having that seed stay really tight. You can do all that kind of stuff. But that's the Journey 3.0. So. Yeah.

And up sampling, does it get the images high enough that there's print quality or are we just talking about like chat? GPT has a decent size resolution right out of the gate. Is it higher resolution than that?

I do not know the numbers to that, but they just went and redid their up sampling, which I was saying like before had. You can see, I mean you see reflection, you can see stitching on like a sword hilt. So it's just really, really cool. I don't know how far you can push it. I don't know what the print resolution is. I'd have to go look in their help because that, that just kind of changed. But it's, it's high enough for most people to use it for what they need to.

Awesome. I know that you use other tools outside of mid journey and chatgpt/dall e3. So what other cool tools are you using and why do you like them?

Yeah, so you know, I mentioned that Mid Journey can really struggle with text sometimes. Even though it's getting better even with 6:1 it's getting there, but it's still not all the way. I love Ideogram AI now. It does an excellent job. Ideogram or id ideogram I D E O G R A M AI Okay. It does an excellent job with text. So we're rebranding social media News Live to creator News Live. And it actually can help you create logos at least idea. I mean, I wouldn't use it as my logo because I want it to be a vector graphic and whatever, but to get it to a point where you can give it to a graphic designer, it's really, really cool. If you need it to create graphic images, I mean like quote graphics or something like that you're going to post on Instagram. Excellent, excellent job. Very, very custom, unique images. So I would check that out if you're needing that kind of use for it. It also has got a thing called Magic Prompt which is very, very helpful for ideation. So like you can put in something and they'll then actually will expand on it almost like ChatGPT automatically in the prompt. And that's a cool thing to turn on to do some, some cool stuff.

So just to be clear, Ideogram is really good at text, but we should be clear. It's not like it's going to do what Mid Journey does. Right. Is this more for logo design or what are we talking about?

It actually does a great pret with images too. So if I would say if you're like bouncing between which one to get, I would look at and once again, they have a commercial license version and also one that you have private on so no one else can see your stuff. I would. If you're going to use a lot of text, you're going to do a lot of quote graphics and stuff. It will cover images. It does a pretty good job of images as well. I don't use it that way just because I use Mid Journey. That's.

But it never screws up the text is what you're saying, right? It never gets the words wrong.

Yeah, that's for text for me. That's what I use it for.

And then what about that animation one we were talking about?

Yeah, but I want to do. I'm a little surprised when I hear it. Yeah, yeah. So it just came out last week and a lot of the YouTubers are calling it a Mid Journey killer. It's called Flux and it's free open source. And so if you are on a budget, I would look at this one. It doesn't do four images, but it does really close Mid Journey ones. Now it's open source, so you're going to need a GitHub account to access it. I run it through a thing called File F A L A I and then I can connect my GitHub account and it kind of runs through that. It does A really, really good job.

How do you find it, by the way? Because there might be a lot of people named Flux or software companies or whatever. Right?

Yeah. So if you do a search for Flux AI, there's a great. I think it's.

You'll find something.

Yeah, check it out. I mean, I don't think it's a Mid Journey killer because I don't see all the tweaking that you can do. I think it makes those quality of images and you can change the aspect ratio and whatever and it's free, which is a big thing. At least free for now. But I still like the way I can tweak things on as a business and as somebody who's creating images on Mid Journey a little bit better. But it is an option, it's brand new, it's worth checking out.

Let's talk about this animation one also.

Yeah. Clingai.com so cling. Like cling on with a K. Well, I showed my nerd right there, didn't I?

K L I N G AI. Yep. Keep going.

So, but it's an excellent image to text generator. So what I've been able to do do is upload some of my Mid Journey images and then animate them. Like it's amazing.

Oh, you said image to text. You meant image to animation, right?

Yeah, image to animation or video is what it does. It is spectacular and I think it gives you a bunch of free tokens. I think I got like a hundred to start and I'll probably end up paying for it because it does such a great job. Not the quickest thing. So don't like sit there and expect, you know, to just kind of crank it out. Like you're used to a Mid journey because it's got to render all those extra frames. But it does a really good job. And you can also, if you have the paid version, actually have it have a starting and ending frame, which is cool. So if you want to have like kind of a loopable thing for reels or something like that, you could do. But you upload an image and you say, I think it's five. And if you have the pro plan, you can actually get 10 seconds of animation. So I was showing Mike earlier, I had a photo for that sound episode that I was talking about and it was a lady who was singing into a microphone and I hit it and she actually sang and her curls bounced on her hair and I mean like it blew me away. So if you're interested in that, go play with it because it's free.

Wow Jeff, this has been really eye opening. I know a lot of people have found a lot of value out of this. If people want to connect with you on the socials, what's your preferred social platform? If they want to learn more about you and the things you've got going on, where do you want to send them?

Yeah, jeffc.com is the best place place to find me Jeff. And SC is spelled S and Sam I E H I before E, especially in C. We do the Friday show at 10am Central every Friday. I'd love to see you there. And also jeffc.com Aiartistry will get you access to a free email based course that will take you through those big three tools that I use, Dall, E3mid, Journey and Ideogram.

Jeff, thank you so much for coming on the show today and sharing your insights with us.

Thanks for thanks Mike for having me.

Hey, if you missed anything, we took all the notes for you over@socialmediaexaminer.com a19 and be sure to follow this show on your favorite podcast app. And would you share this show with a friend? You can tag me if you want. I'm Elsner on Facebook and Ike Stelzner on Accent. Do check out our other shows, the Social Media Marketing Podcast and the Social Media Marketing Talk show. This brings us to the end of the AI I Explored Podcast. I'm your host Michael Stelzner. I'll be back with you next week. I hope you make the best out of your day and may AI help you become more successful. The AI Explored podcast is a production.

Of social media examiner.

AI Business World, early bird pricing ends on January 16th. You can save $300 on your tickets. Register at AI Business World Live today.