First AI Wedding Photography Course with Flux AI

1. Intro First AI Wedding Photography: Have you ever dream of stunning vetting photos in exotic locations, but felt limited by budget or time? What if AI could make those dreams a reality? Welcome to the first AI vetting and pre vetting photography course. Hi. My name is Soli art, and in this course, you will learn how to use Flux AI to create breathtaking, personalized images of you and your partner? Placing yourself in any scene you can imagine. To inspire you, I will share some beautiful photos of my wife and me that showcase how easily you can create your own magical wedding moments, no matter the settings. Your dream album is just a course away. Let's make it happen. 2. Quick Start with Flux Ai and Prompt Formatting: Access to Flux Laura Model. First, we have to sign up on Hugging Face. Let's go to Hugging Face dot C on the right side, Tabn sign up. Since I've already logged in, it's just going to let me in. After you sign in, you have two ways to access Flux La model. If you sit on the right side, tab on it, Flux Laura, the Explorer. And remember, always look for the new Flux model here. Sometimes the Flux Laura, the explorer is runtime error. It's getting an error, right So you always work with the new one Flux 01 development version, Flux 01 Chanel, Flex realism, so just open them up and use them easily. So don't forget. Always look for the new version of the flux here. And these models will be updated every day. So for example, here, flex realism laura emigerator. Put your prompt here and generate, or the simplest one, flex 01 chanel, which is giving you the exact same thing, and the settings are here, or flex 01 development version, Given you the better quality for your images. All right. Let's start with one of them, and you're just going to jump in. And also in your PowerPoint a slide show. The link is here, just tab on it and urine. Flux AI is an advanced open source text to mage AI model developed by Black Forestab, is free and it's like met journey. Here in Laura Gallery, we have too many filters. If you're looking for cartoons or animation tab on it, put your prop here, and you're done. But in our case, we only need flux realism. Just tab on it. Make sure here is type on prom for flux realism is appeared. It means we are good to go. Now in prompt section, let's type realistic portrait of a beautiful blonde girl. Tab on generate, and let's see what we get. As you see, it says, it's going to take around 46 second. And this is the maximum. Most of the time, it's just going to be finished around 22nd. All right. It's done. And how beautiful is that? Without even touching the advanced setting, we get this beautiful portrait. On the top right corner of the image, you see the share and download button. Tab on download to download the image. As you see, the format is P with 73 kilobytes, which is very low. And this is the main issue that we have here because it's totally free. We can use so many images here without paying anything, but the quality is very low. How to fix that. Just simply go to hug inface.com slash space, or you just click here in the search spaces type face scale. The first one is Night Ferry, the creator of this free model is the one that we're going to use. Just tab on it, and here you are in mage scaling and restoration model, which is specifically works on Face. Just drag and drop your mage Here, make sure is on version 1.4 and put the re scaling factor on ten to get the best quality image of it and submit. In less than 30 seconds, we get this large upscale image with the best quality. As you see, is 2.4 megabyte, and just simply download the image. And this is how we fix this problem. Let's come back to Flux Laura the Explorer, and at the bottom, you'll see advanced setting, Tabans litter icon here. Here, we have CFG a scale, steps with and height. Before I explain CFG and steps, let's put the highest amount that we have here. Here, we have 15 36, which is the highest. But I will show you a track. You can put any number here, and it works. Let's test it with 15 36 heights and generate it again, and we get something like this, which is beautiful. Let's download it. Again, Dragon drop here, number ten, submit, and we get the highest quality. Let's come back to advanced setting again. This time, let's talk about the CFG. And what's going to happen if I put it on 20. CFG scale is a parameter that controls the flux model and how strict, it should follow the prompt input in me generation. So basically, the higher the number, the more actual generation is going to be look like prompt. But based on my experience, the 3.5 is the best for this model. But you can test it out if I put it on 20 and generate it again, As you see, I will get a more realistic image. Again, we can do the process of upscaling that. This time, I'm just going to put the CFG on 3.5 and steps and 50. What is the steps? The flux parameter is a parameter that controls the number of etration in the generation of images. Atration refers to the repetitive cycle starting from generating random noises based on the text input. When an image is generated with a flux model, it will go through many steps, typically about 20 to 30. That's why, if I put it on 50, it's just going to take more time and GPU to produce an image. But the result is going to be more realistic. Now, what if the CGF scale is on 20 and steps are 50, let's try it out with the same pront. And here it is. The result is absolutely very realistic image. And how cool is that? So by controlling the settings, you can get exactly what you're looking for. But for us, most of the time, we put a CFG scale on 3.5 and steps on 28. The left one is CFG 20 and the right one is CFG 3.5 and steps on 50. And I believe both of them are perfect. So now you know how to control the advanced setting and flux Laura model. I have generated more than 200 images here, and I got this error, and I have to wait for 29 seconds. To let me generate the next image. The simple solution for that is log out and log in with another g mail. And I can run this again for a week without even any problem. And after that, I can go back to my previous e mail. And this way, it's just going to work forever. And let me show you a track. Here in Advanced sitting, you can put your desired height, for example, here, 1920. And you can generate your image to get exactly the same size. And how cool is that? All right. Now you know how to work with flux, Laura, and Hugging Face. Let's understand the importance of well written prompts for generating high quality images. And at the end, I'm just going to give you my prompt formatting. The first is the subject like a woman walking. Then is the film type. If you like, you can put it here. If you don't, don't put it, like Corak gold 400. If you're looking for a specific film type. The next is camera angle. If you want your model or image, be at eye level shots or low angle shot or high angle shot, extreme low angle, Vb cam style, underwater shots, underwater close up. We have so many style and angles here. The next number four is lighting. Are you looking for natural light, like sunny day, snowy, overcast, foggy, hazy, Sunrise, golden hour, or studio light like catch light, split lighting, loop lighting. The number five is ambient or styling, like dark and moody image, bright and vibrant, Fantasy. Historical, the next is location. You can put like New York, Central Park, Egypt, Eiffel Tower. The next. If you're looking for a specific style or photographer, put photographer's name, like in a style of Steve McCurry, Ansel Adams, Mario Testino. The next is cameras and lenses. Put taken with a R Alexa, LF camera with RA signature prime lens, 12 millimeter, 1.8 aperture or red ranger monster. At the end, you can put some fine tuning, like cinematic, photo realistic, atmospheric, filmic, hyper realistic, boke elegant, so on. And here is an example. A couple walking in the straight. A fm type can be ora gold for 100 camerngle, eye level shot, lighting, in natural lighting, Nes stria style in New York, photography by Annie Lebowitz, shot by Ai Alexa, realistic, supernatural, and you're good to go. Now, let's put the prompt like this. Couple embracing on a beach at Sunset, Fuji film Vil Va 50, which is the camera and Photo style, low angle shot, Golden hour lighting, cinematic style, Maldives, which is the country, photography by Peter Lick shot by Canon EOS R five camera with cannon RF 15 35 millimeter 2.8 aperture, SM lens, hyper realistic, dreamy. Now, let's copy this formula, Control A Control C. Go to one of the moro. Just paste it here and run it. I'm just going to run it in different models and generate. Let's see what we get. It's cool. Awesome. Both models is giving us something perfect. But the problem is, most of the time, we don't know how to explain our environment, how to explain the lighting, what photographer to choose, what camera angle to choose. The best way is put our prompt in the formula that is understanding everything like this. First, let's understand the prompt formula. This is what I put in GPT or perplexity or clot. Any time is started with prompt, create a prompt with a word count limit of 500 words for the AI based text to image program using the following parameter, one, two, three, four, five, 6789 parameter. The first paramet number one should be replaced with a user supplied concept. Two should be concise descriptive summary of a subject. Ensure that description is detailed, uses descriptive adjectives and verbs, a diverse vocabulary, and sensory language. Number three should be concise summary of sense of environment, number four, mood of the scene, number five, atmosphere, number six, lighting, number seven, realistic photography style, number eight camera on lenses, number nine, cinematic, eight K realistic. It's important to remember that the description is the prompt should be written together, separated only by commas and spaces and should not contain any line brakes or columns, brackets, and their contents should not be included. And this is the general idea of our prompt. Now, we have this. Let's go to full prong formula. I'm just going to control A, Control C. We can go to cloud AI or CGP t. I'm just going to control V, past it, and inter. It's understand and is giving me an example. Now, I'm just going to come back here. Copy these parts. Control A, Control C. Let's go back to GPT. I'm just going to type prompt, column space Control V. Inter. Now, it's putting my prompt to this cool formula and it's giving me a very detailed and sophisticated prompt. I'm just going to copy this go to one of the models Control A, Control V run it. Let's go to another one, Flax the realism. Control V and generate. Let's see what we get. This is so cool. Let's see. Also this one. This one is also cool, and I can download it. Nice. Now you know how to use the formula. It's time to create our own images. 3. Train Your Photos For Ai Photography Update: Now it's time to train your face into the new AI model. And you're going to love it. Replicate is a platform designed for running and fine tuning open source machine learning models, allowing users to deploy custom models at a scale with minimum effort. This platform is very, very low cost. You can train your face or any model and get 100 images for less than ten bucks. And, trust me, it's worth it. You will get amazing and beautiful images that you always wanted in your life. Alright, let's get into it. Let's go to replicate.com and log in to the website. Search for Luka Taco. Scroll down and Enter. Now tab on Luka Taco profile name. So we are here or just simply go to replicate.com slash Luka Taco Scroll down and look for AI Toolkit. Here is where we are going to train our model. Now here, Tab and train at T Destination, select a model and create a new model. By default, is public. Now, let's give it a name. I'm just going to call it couple. Scroll down. Here is the part that you have to add your images as a zip file, and we have to rename our model exactly like this, a photo of, I'm just going to call a couple, a love with capital letters. It can be PNG, JPEG, et cetera. And it's better to be PNG, but it doesn't really matter. Most of the time, it works perfectly. Here are some of my images of me and my wife. These are Selfi photos, and they are not really high quality. Since most of the people have some images like this, let's try and train our model with selfiPhotos. And I rename them as a underline photo underline of underline couple. The couple capto make sure to select them all and right click and zip them. Now, dragon drop the zip file exactly here and let it go. And make sure you follow the rename formatting. For Hugging fake token, let's go to Hugging Face. Tab on your profile name setting access tokens, create new token, token name. I'm just going to just put couple, give it some access, repositories, inferences, webhook collections, and create token. Copy this Come back to replicate and paste it here. Scroll down. Don't change anything at the end, repo ID, which is going to tab undone. Again, profile name on hugging face, new model. The honor here, the first part is Mr. Mobile and model name. I'm just going to put couple. Make sure it's public and create the model. And come back to po ID. I'm just going to type Mr. Mobile couple. And sometimes if I do create training, I'm just going to get an error and sometimes it doesn't. If you do get an error, just come back and try again. Like this, I'm getting an error. Alright, this is an update. If the previous Loco took training doesn't work, we are going to use Astris. This is the new training model that it works perfectly. So if you go to this link, put it on file projects and you can have access to it. Just go to replicate AstrisFlxva Trainers train. And it's the same process. You can click on new model. I'm just going to put this, for example, my model. Input image is the same process. The trigger word just put a trigger like put a trigger word that you don't use so much. So like W Girl, something like that, something that is not really in the text. The next is this is the hugging face po ID, which is the same, and also hugging face token, put these two here, and rest is totally the same. So just create training and it's going to work perfectly. You just got to wait for around 30 to 40 minutes and then come back. Alright. Let's finished now. The status is successful. It took around 37 minutes. Let's go to Hugging Face, and we're just going to refresh this page. And you see the model is created here. Now, let's tap on Luka Taco again. Again, Luka Taco. I'm looking for flux De Laura tab on it. The first part, you put a prompt. The second part is aspect ratio. I'm just going to come back on Mr. Mobile. I'm just going to copy this part. Go to Hugging fakes Laura and just paste it here. Let's change the aspect ratio to nine by 16, which is the portrait one. Let's scroll down, show hidden in plut and let's change the format to JPEG. And let's put a simple prompt. We can put the trigger name, which I don't think it's needed anymore, but I'm just going to do it. Couple, then a couple walking in the street and looking at the camera. Let's run it. As you see, it's not what I want. Let's try again. Alright, now this time, they are looking at a camera, but it's not us. She's not looking my wife and he's not looking like me. So what's the problem? Let's try something else. Let's go to full prom formula. This time, let's go to Cloud, paste the prom formatting here. Inter. Right, going to give me an example. Let's go back here. I'm just going to copy this a couple walking distant look at the camera. Let's come back to clots AI, paste it. Before that, I'm just going to type prompt, column, space, and Enter. All right, cool. Copy this. Just going to paste it here and let's try this. Again, this is not bad, actually. It's maybe 60%, me and my wife. Let's try again. Not at all. This is not me at all and not my wife. So it means we have to train our model differently. To make it 100% look like us, we have to do this. So basically, you need every angle possible with you and your partner. So, for example, the front view and side view, side view, looking at the camera, totally side view. A bit have a distance, and all the other photos, I try the background be white. So basically, you have to do something like this. And this is the best way you can train your model. So let's go to Photoshop. I'm just going to show you a simple trick how we can do this. Just drag and drop one of the photos that you are just going to drag it and put it in Photoshop here. Just drag and drop. Now press S on your keyboard to choose the crop tool and just make it a bit larger press inter. Here, you see something called quick selection tool, tab on it. If you tab on it on the top, you see something called select subject. Tab on it, and simply just going to select you and tap on this little log to make it disappear. Now, press Command and Control J. You just make it like this. And you can just put it this way, go back to home. I'm just going to bring my wife Again, tap on here, select subject, Command or Control J, and let's select it and dragon drop it here. Control T to make it a bit larger. And we have the side one. Now, File exports as format, it's PNG and export. You need 12 photo like this, or you can make it like mine, front side, side, and dressed like this. Make it 12 photo. This time, I'm just going to name it underline photo under line of love, which is capital, and it's 12 photo, which is the minimum. Now let's go back and create the marvel again. 4. Recreate the Models in Replicate: Again, let's go to replicate, Luka Taco, slash AI tool kit, Tab on train, select model, create new model. This time, and we're just going to call it Love. For images, make sure to make all your image a Zip file. This time AO of love. Now, drag and drop it here. For Hugging phase token. Let's come back to hugging phase, tab on your profile picture, setting access token. Create new token. I'm just going to make it love, give it some access again, create token. Copy is done, come back to replicate and paste it here. Scroll down for rep ID, t's come back to Hugging phase, tab on profile picture, and create new model. I'm just going to call it Love Mr. Mobile slash Love Create model. Wait for it around 22nd. Copy this, go back to repo ID and paste it here. Now you can create training. Again, failed, let's go back and try again. This time is working. This time took 5 minutes, and the status is successful. Let's go back to Luca taco. Flux Dev Laura. I'm just going to type couple Selfie aspect ratio nine by 16. I'm just going to run it. All right. This is cool. This is exactly me and my wife. Now it's time to create we're looking for. 5. Best Prompt Formula for AI Wedding Photography: All right. Take a look at this formula. Couple looking into the camera plus creative couple pose plus description of lighting setup or effects plus mood or atmosphere plus couple positions or intractions, plus description of overall composition or visual style. I'm just going to ask CGPT based on this formula. Give me two example of creative poses and lighting. Let's go back here. Copy is past between two sem column and k. The C GPT is giving me the whole description. I'm just going to say, make it for text of mage prompts. Let me copy the first one. Let's go back to replicate and paste it here. Let's run it. Cool. I love it. It's totally us. Everything looks perfect. The hands, the face, shadows. Let's download it. Let's run it again. Let's see what you get. This is super cool. I'm a bit fat, but still it's me. Let's try again. Awesome. I love it. Here, I asked CGP to make it shorter. I thought if I make it a bit shorter, maybe I can get a better result. I gave it an example I make it shorter like this example. A couple looks at the camera back to back on the warm rim lighting, their hands gently touching, creating a serene atmosphere. It's giving me some examples. Let me try this. Let me copy, paste it here, run it. This is not bad. This is very cinematic lighting like studio lighting. Let's go back to the long version and copy the number two, paste it here, run it. All right. This is super cool. Now let's try one of the short versions. Paste it here, run it again. Nice. Most of the time, Man I make it shorter, I'm getting a better result. It's totally us. It's very simple, like taking a selfie or something. It's very close up, very natural. Run it again. Cool. I love these photos. The shorter version, it's much better than the longer version. Try to make it shorter like this example exactly to get a better result. If you didn't get what you wanted. Let's try one more. A couple looks at a camera in a playful puse ber and gel lighting adds energy as one leans forward, let's see. Just going to paste it here, around it again. Wow. This is fantastic. Oh, my God. I love them. I'm just going to download it. I don't want to lose it. I'm just going to ask GPT to give me ten more examples with different landmarks in the background. Let's come back and add the first one. This time, the background would be statue of liberty shining in the background. Let's see what it get. Wow. I love it. Oh my God. I'm just going to download it. I'm going to lose it. Let's copy number three, for example. Let's come back here. Paste it, run it. Oh my God. This is the Eiffel Tower version, and I love it. Just take a look how realistic is the photos are. This is exactly 100%. It's me and my wife. I love it. Try to make it shorter if it doesn't work, and if you like something, try to change the background, and that's it. Oh my God. And just make sure to download it, if you don't want to lose it. I re run this again, and it's like my wife taking a selfie in Eiffel Tower. I love it. So cool. Now, let's try to change the outfit. I ask a j to describe the outfit, girl with red dress, and guy with suit. In the same prompts, a couple looks at a camera in playful pose, vibrant el lighting, adds energy. Let's copy this. Try this. It's a bit long, that's what it get. All right, not bad at all. The selfie one. It's really realistic. But sometimes it's like you hired a photographer in Eiffel Tower and is taking a photo of you guys, and it's perfect. I love it. Let me try again. Now we are talking. I love this one. This is so cool. It's like a poster of Indian movies. Now let's ask Cha ji to make it a selfie photo. Copy this. All right. This is cool. It's like someone taking a photo was when we are taking a selfie. Alright, let's try again. Now we are talking. It's super cool. I love it. How classy glamour with red dress and suits. I love it. So most of the time, make sure you put your dress outfit and style. So to get what you want it. All right. I'm just going to download this one as well. L et's rerun it again, the same thing. Now this time, let's make them like two models. I'm just going to change the prompt with number eight and change the girl with the dress and here the outfit like two models. Let's see what we get. Let's try number nine, like outfit wearing ski suits. Let's try ski suits. Let's change it to ski suits. Let's see what we get. I'm just going to add snowing, close up and copy the prompt. Let's paste it here. Is what to get. Because I love a ski. Oh, my God, I love this one. This is exactly what I'm talking about, guys. A close up in a ski, and I'm just going to describe the background, GPT. Oh, my God, I love this one. Just going to make sure to download all of this. 6. Final AI Wedding Photography Formula and Formatting: All right. To create something like this, we have to follow this structure. Create a portrait of couple, the girl is wearing a specific dress style. You can put flowing Maxi dress, tailored blazer, or so on, accessorized with specific accessories like statement necklace, elegant heels, check handbag, and the guy is wearing a specific dresser style, suit jeans, vintage, captured in the style of photography style or technique in setting or background. As Chagp also Cloud AI to give me 20 fashion couple photoshoot ideas with this format, and I got this. Let me show you how to do it. Just copy this part. Let's go here. Cloud AI or Cage P doesn't matter. Based on this formula, give me ten couple photography ideas in different location. And semicolon, I'm just going to paste the formula inside and press inter. Let me copy this, paste them here, and let's try the first one. Copy. Paste in. How car is that? Oh, my God. I love it. Let me try again. Oh, my God. I love it. I love this one. Now, to make it more interesting, I'm just going to make it a couple looking at the camera. We can also change the formula to this. Let me try this. It can come back here. And also create a portrait of couple. We can do something like this. A couple looking at a camera. Let's come back here. And here we go. Don't you love it? It's amazing how cool is that. This is exactly me, and she's my wife. I love it, and you can be anywhere. Now let's try another prompt. Let's try one of these. Number four, maybe copy paste. In? Oh, my God. I love this style. Let's try again. Nice one. So cool. Let's try this one. Give me a couple having fun in different landmark around the world. Let's copy this ten, spaced it here, let's try the first one. Copy pasted here. The location is Barcelona. Oh, my God. How is that? Looking at the camera in trigon Still they are not looking at a camera. Close up of a couple looking at a camera. L trig. All right, cool. Now they are looking at a camera. So sometimes by changing some wordings, you can get exactly what you're looking for. But the cool part is it's always you. But this is kind of a fat version of me, but it's steal. It's me. Now let's try another one. Copy, Paste it here. Is cool. I love to see that. Oh, my God. This is so funny. I love it. I love the idea. Now, let's try this. Copy this the full prong formula. Put it here. Inter. Cool. Let's come back here. I want to get something like this lovely couple in the wedding day. Let me type prompt column space, give me five different poses and locations. Five different prompt. Now it's giving me five different prompt. It's going to copy this. Now let's try this one. The first prompt. T Oh, my God, I love it. Let's see one more time, what's going to happen. All right, not bad this time. Let's try again. Exactly like this one. I love it. Super cool. Now, let's try this and put it on a GPD this time. Photo idea of a couple in the vetting day based on this formula. I'm just going to past it inside two columns inte. Make sure they are vetting outfit. Let's copy this. Come here. And paste it. Let's see what to get. This is stunning. I'm just going to download it, so cool. Keep everything the same. Just change the location. Let's copy this past it. I love it. This is so cool. Download this one again. The only problem is the hair. Let's try again. All right. This is super stunning. I love it. Let's see them side by side. What are the problems? The dress is changed. The guy is okay, but the girl, the hair, the dress is changed. How to make it more consistent. To do that, I'm just going to tell something like this. It's good. Just explain the exact outfit details and both her color and style. Make it. One, prompt. Everything is okay. The bride, she's wearing this, let me copy this. Come here, and paste it. All right. I love this one. Let's download it. We have this. The only part that we have to change is the scene. The rest must be the same. This is the part that we have to change. Change this to five different location. Let me copy this. Now, I'm just going to duplicate this. Come back here. Just copy this part, the same part, come here and paste it here. Now, Control A, Control C, now let's try this one. Basically, I just changed the location, not the outfit. Let me just download this. Drag and drop it here. And here we go. It's almost the same. We don't see it here, but it's almost the same. Let me try this and run it again. I believe this is exactly like the first one, Let me download this. Let's come back here. I'm going to maybe delete this one, and copy this. The more be specific about the dress, the more we get the same look. I said, be more specific about the dress. Every parts need to be explained. I'm just going to paste this prompt again here. Let's see what we get. I asked GPP to be more specific. Maybe we get the exact same thing for a character consistency. Let me copy this. I love this one again, and just paste it here and run it. I love the first one. I'm just going to download this and put it here. Let's change the location. I'm just going to copy one of the locations here and duplicate this pasted here. Control A, control C. Let's see what we get this time. Okay. Let me check. This one, it's almost the same. If I ask GPT to be more specific, run this again, and let's watch this side by side. It's the guys. If someone take a look at these two photos, they think they are the same dress. 7. AI Pre Wedding Photography Posing Locations Style: All right. Let's go for the best formulas that I've invented for the best results. We have three to four formulas, and this is the first one. Now let's try to see what we get with the first one. Couple looking into the camera, description of creative couple pose like models, normal feet size, and body shape. Is normal. Because sometimes I get a huge, like a giant person. That's why I put these words to make sure they are normal people, plus description of der or environment. So we have the veer description, now plus description of outfits based on the veer or environment, plus mood, or atmosphere plus description of landmark or background. Each time, we can change, for example, the atmosphere or the landmark in each formula. I put this formula to gipty, and I got this pront. A couple looks into the camera, the guy leaning against the guy's chest as they both laugh. As you see the snowflakes fall gently around, covering the ground in white. She wears a cosy fur lined parka while is bundled up in a thick winter jacket and a scarf. As you can see, it follows the instruction. Ask for another prompt and I got this. A couple looks into the camera. The girl sitting on the guy's lap, as you see, it's doing that, both smiling warmly as they relax in the moment. Sonny with a gentle breeze, she's wearing a fluffy white sundress, and he's a casual t shirt and shorts. The mood is peaceful and loving with the eiffel tower in the background. As you see, it's perfect. I ran this again, and I got this too. Now, let's go for the second formula. Couple looking into the camera, both with normal body size and average fit in a creative pose plus description of the weather and environment, including time of the day, if relevant, plus detailed description of outfits, appropriate for the weather or environment, plus description of the mood or atmosphere, plus detailed description of the landmark or background. And the GPT is giving me four prompts. I tried the first 13 times and I get these beautiful images. A couple looks into the camera, both with normal body sizes and average feet, the girl Hugging the guy from behind as they smile together. As you see, these two on the right are perfect, and it's Sydney Opera House in the background. I love it. And you can create this simply. Just follow my instruction. This is my formula for creating consistent character details with varying poses. Based on the formula, everything is the same. Just the posing is different. Look, she's wearing a floby, white sundress and he's in a casual T shirt and shorts. Keep everything the same skip the poses and let's do what we get. Different poses with the same outfit. Now, here, I changed she's wearing a fluffy white sun dress with a fitted bodies and a line skirt, and he's in a light blue short sleeves, T shirt, and a bage cargo shorts. I didn't change this during all the different poses. And let's see what the get. Look, a couple looks into the camera pose with normal body size. Everything is the same. I changed the pose, and I just changed the dress and shorts. Just take a look at this one. If I give the specific color code, the result would be better. I tried this to add this to my formula. I added the color code, just kept everything the same, and I got these kind of examples. Flowy white sun dress. This is the white color white dress, the T shirt code, the beige cargo is this color, the T shirt is this blue color, and I just kept it the same during the whole prompt, and I just changed the posing. The first one, white, blue, and beige. The second one is exactly the same. The color code, it's 90% is effective, and this is another one. The color code is almost the same. We have shades in the day. We have a sunny day, based on the light, the color shifting a bit, so it's normal. I decided to keep everything the same, just change the background. This is the 20 landmarks, like Eiffel Tower in Paris and the rest, the best landmarks around the world, and I got this. I just changed a couple stand in front of the Eiffel tower at Sunset. I got this. Based on this formula, like ft couple and everything is the same. They are wearing summer outfit. Just make sure the outfit is the same in all the locations and scenarios. Give me four scenarios with four different landmarks around the world. This is for text to image prompt, so each location is one prompt. So I got this. This is the whole prompt, and I got this beautiful image. In the last formula, I just find out that the color code is not really effective. It's working sometimes, but it's not really understand the color code, so I removed it. Basically, you can put the color code or you can remove it. Both ways works perfectly. Take a look at this one. The prompt is the same, just the varying and the outfit is different. This one is yellow and purple, but on the right side, we have different color of hair and outfit as this one. As exactly like this one. These two colors are exactly the same, the style and everything. Let's try one. I change the location and the color of the outfit, take a look at this two. These two steel are the same. I ask CGPT, based on the formula, give me four vetting photography ideas, like this. Put this idea, capture the bride and groom during the golden hour. Shortly after sunrise or before sunset, the couple stands together in an outdoor setting such as a garden or open field, the groom's arm are wrapped around the bride's waist and based on this formula. Make sure the bride and broom outfit is 100% the same in all the scenarios. Each scenario is in one prompt with no spaces or gaps or title. Git result is this one. I got this beautiful image. Trust me, these are one of the best photos that I have so far. And I showed this photo to my wife, and she didn't know if it's AI or if we really took this. Just take a look at this one. On the right side, it's like a couple taking a selfie, the bride is in a white wedding gown with intricate lace, detailing and a long train. And the groom in a black tuxedo, with a white dress shirt, a black bow tie, and as you see, it's amazing. So anytime you wanted something very realistic, try a couple taking a selfie. Try this one. A couple looks into the camera, both with the normal body size and average fit. The bride is wearing what you want to wear, and just put it in a back of taxi, in a back of car, in a balcony, and in different spot or different location, and you will get a very cool and realistic result. 8. From Pinterest Wedding Photos To Reality with AI: I don't know about you, but I've always wanted wedding photos like Pinterest photography. You know, they are amazing, like boat photography, wooden boat photography. They are absolutely stunning, but they are also expensive. You have to travel to those locations, hire the best photographer, and cost you a lot of money. But what I'm going to show you is just going to change your life. Just download any photo that you like. Go to cha GPT, and click on this letter icon here. To upload your photo, upload from computer. In reading, I'm just going to add one of the photo that I downloaded from Pinterest. This is the photo that I love to have something like this, and I just ask ChagpT to describe this photo. I'm just going to copy this, and I'm just going to ask GPT. Use this formula and give me five best boot photography prevetting. The formula was a couple looking into the camera, the same formula that you know already. Let's do this. Use this formula, which is this one, paste it for this description, and I'm just going to the photo description that I got from C GPT, and just paste it here and in. And magically, you will get the description of the photo inside our formula. Just copy that, paste it and replicate, and, you will get something like this. The Chagpt is just describing the image. It's in the boat, the sunrise, the feeling the mood is the same. Let's try again. Oh, my God. I love it. So basically, we can just change the posing and run the program again and we'll get beautiful images. Just make sure the dress and the outfit is the same and rerun it with different poses, as you already know, and we can get amazing photos. So just find any image you like, put in the GPT, get the description, put in the formula, and. You will get what you really deserve. And just take a look at this one. Based on this photo, I got this prompt, and these are the two images that I got. I believe the two images that I got is much better than this photograph here. Because it's very far, I can't see the faces. Just take a look at this one. The yellow highlight, it shows the formula, the green high lod it shows the photo description. Paste the photo description here, put it in the CG Pt. You get the full prompt, put it in a replicate and get what you want. Now, I ask GPT based on the same prompt, give me something else. I'm just going to paste it here, and I'm getting this beautiful image and absolutely gorgeous, and it's totally me and my wife. I love it. And you can tank me later by sending me your couple photography. Now, in this prompt, I just change my outfit. And it's amazing. Here, I change the pose with the same outfit, and I got this. Just take a look at these beautiful images. You can just change it to look at the camera, you can just to look at the sunset, look at the sunrise, and you will get these beautiful images. I change the hair color of my wife to blond color, and this is beautiful. Just take a look at this one. How romantic is that? Beautiful images with the same bland color. My outfit is the same. Take a look at this one side by side. A couple is seated in a small inboard on a calm lake and you will get something like this. Now let's talk about your assignment. 9. Final Assignments: You have two assignments to do. The first one, if putting your idea, you as a couple, describe the location where you want, put it into the formula, and change the poses, the location, and the mood, and send me your CGP result, and the photo. For the second assignment, pick up the best photo on Pinterest or anywhere you like as Chagp describe it, use this formula, and send me the prompt and the result you get based on the original image that you wanted. Make sure to upload your photos for my feedback. Happy AI photography,