Create Cosmetic Product Ads: CGI & Realistic Video with Magnific AI & Premiere Pro

1. Introduction: What if I told you that you don't need a photo studio, a mod alone set, or a single day of filming to create a cosmetic ed that looks like it came straight out of luxury campaign. Let me show you this example. Boom. Wow. What No. Stop. Pi, it just takes two images to create this whole advertisement, a model image and a product image and the right workflow that works with the video generation. Hi, I'm Karan. I've been working with AI Generation tool that is Magnific formerly known as Freepik for a while now to create high end editorial CCI loop that brands actually pay for. In this class, I will walk you through the exact workflow that I used to create this advertisement. I will show you all steps one by one, not just theory. This is for you if you are a designer, a freelancer, and a small brand owner, someone who wants to learn a skill that's generally in demand right now. So grab any photo of a model and a product that you have at your home, and let's jump into Magnific AI and start building your cosmetic ad step by step. I will see in the class, let's get started. 2. Magnific interface generate first image: Hi. Welcome to this first lecture. We're going to understand how Magnific actually work. After that, we are going to jump into the advance section. We are going to create CCI and creative AD. Once you go to Magnific, which is formerly known as Freepik, also, once you go to Freepik, it will redirect you into the same page. Creating any account on Magnific or Freepik, it just like creating it just like creating a Gmail account. Once you can see here, here is the spaces. If you click here, here is the species. Once you click here, you will see a new creative space here. What we are doing here, we are going to create a new space. Like click here, once you click on New space, you are going to create a new space here. So your new space is ready. You can name or rename it. This is the new name that I arranged. Now you can see a small tick here. Also, if you want to add a node here, once you click here, UC, all the nodes has been listed here. Once you right click, you can see same structure has been shown to us just like this. It's totally simple to navigate this. So right now, I just want to add a new image. This is the image node. And if you want to add image, a reference image, you can do so. I'm just taking a screenshot here and just pasting it here, control paste. So you can see the skin shot that we took has been uploaded just like any image. You can dragon drop image also, so it doesn't matter at all. I just had to connect it as a reference image. And if you want to remove the connection, you can do, just like this. If you want to generate a video, you can generate a video just like regular node that we have created. You can type anything here, you can type anything here. And if you want to change the model here, you can do so Chat GPT, Google, Nano Banana. It's your choice. Whichever model you want to use, you can use. It's really simple to use. If you want to change the duration of the video, you can do so. If you want to change the output to one or two, you can do so. So this is just a simple image generation or video generation tool. And if you want to generate the image, you can just run it and you can see the credits here, right? So this is just a simple interface introduction. We are going to jump into higher version so you can understand it better. So let's generate a image here, like cat jumping from done. Now I'm using a simple Nano Banana seed dense, let's use seed dense and run it. You will see the timing here under 1 minute, your image will be generated. Sometimes it shows you 1 minute, but it happened within seconds also. So if you want to delete the frame, you can delete. And if you click, if you can delete this also like this. And if you want to delete that, you can simply press the delete button like this. So our image is generating, and this is the first image that we are generating, right now, all these are simple connection, node connection. Wow. So our image has been generated, and you can see it looks somewhat funny and somewhat okay. So this is our first image that we have generated. Now we are going to generate so many things and understand the prompt structure also. That's it for this lecture. See you in the next lecture. 3. How to use the template in Magnific: Start by going to space and start by creating a new space. Once you go to new space, you can clearly see that space is ready. Really cool to see, right? It's just like gravity working on my mouse arrow. It's really cool. So let's move to our main objective, that is the node. Let's create a node of text here. And you will see the text here. If you want to change the text also, you can see, you can just simply type or give the title to your text. That's all you can do. This is for just a fast video generation and image generation. If you want it, you can use it directly. And if you do not want you can simply enter into space and start working right away for the greater and bigger projects. So we have created a space here, untitled space. If you want to change the name of this space, you can do so just like this. Simply rename it here and you can rename it here. And if you do not want it, you can just simply get back to your main space. Now, if you have any shared space, you can see here. And once you shared the space, it will give you only a three month access for this space. Also, for the template, let's start by using the templates here. Once you go to templates, you can see, create images, discover design notes, discover its notes. You will see so many ideas by going to space, I just do not want to use that. Let's get back it. I accidentally created the space. Let's get back to our space and you will see there are so many templates that had been created, anything in our mind. Let's try to create the images. Let's create image. This is the simple image generation of our prompt. What it has done, what the Jif and follow the step. They have used the assistant here by assistant. They have connected this node with the image generator and the image has been generated. After that, this is the use reference to guide your generation. We can use image to image generator. So the image here that has been generated is now used here, and the image has been used here. And just like this, in the video generation we have talked about, we have used both the reference images here, like extra information if we want to add. Let's try to use this template so that we can understand it better. So this is the image, and let's try to use. So this is the image. This is the reference image. Now let's try to add. So this is the prompt I just do not want to it this prompt at all. I just simply connect this image with the reference image extra as an extra because it's just a image generator. Now, let's add character set current reference image guy man is riding the horse. Now let's see what we got here. Let's run it. The image has been generated and it looks really cool, right? The image is really cool to see. It's working really fine with my face also. Now for this one, let's use the same image that been generated. Let's delete this one and delete this one. Also, use this as a copy and paste. Let's disconnect it and disconnect it. For this one, let's use my reference here. The image has been generated here, and this is the reference we have. Now, let's try to change this reference also. Basically regenerate the image. Replace it. Let's replace this image. Let's try to go to image generator, type here, and let's add dense forest with big trees, and let's try to generate it with a cinematic look, I guess, let's try to get a cinematic look, and we are going to use it as a reference image here. Let's try to delete this specific image and use it here as a reference image. It's just asking me to understand the camera and use the camera of use the camera setting it asked me to use, but right now, we do not want to use it. For the character, we do not want to use any of it right now. For this one. So the image has been generated, let's use it as a reference image here. Now with this one and with this one connected, you can see there is no image that has been connected and they have actually cross it out. So we have to add our images herer Like once you click it, add our reference image, you can see the reference generator two and the reference image three. So I have replaced the image with my new imageaer and once black hose galloping image style from image generator three. This is the generator three. You can see the title name here and high angle shot. Now, let's try to generate the image. Let's run it again. 4. Understand complete node system: I did, I simply add the image generator to, this is the image generator comes in reference in this generator three motion below the involvement to match both the result. This is what I did, and you can see the result has been now updated. This is the result that we got, and it is actually better than before. With this one, let's use for the multiple version of one image. You can actually generate the multiple version of your image. That's sad to delete this image. And let's use copy and paste it here, cut the connection, and simply use it for the variation, just like right click. Now, you have to add a node here, that is variation. This is the new node we are talking about. Once you drag it here, let's remove this one and place it here. Simply drag it here. Now, if you want to run it, you can run it. You can select the grid if you want to see. It is just like a storyboard. It will work in the storyboard also. And if you want to select it, you can select any of the view if you want to do that specifically. And if you want to split the images, you can split the images. You can see the images has been splits. Now let's try to generate the model that I'm using Forex two for the eighth generation. This is the grid size for eighth generation. You can actually determine which generation do you want. Which grid do you want? You can actually test it out. Now let's try to run it. Once you earn it, you will see a magic will appear. 5. How variation node works: These are all the image generation has been created, one, two, and three, four, five, six, this is not me. This is not me. This, this. All these are really good generation. We can check this out. So you understand how the image generation work, how the variation work, how a text prom work, how we can use both the generation or we can use the reference image and using that reference image, we can create our own. And also, this is all the output we have understood. So we have understood how the image generation model work in the spaces. Now let's go to our new space and understand more of the template. So this is the creating image space we have explored. Now let's try to use designer node, discover designer node. Now let's try to use it. 6. Create model image of yourself: At first, it looks web and it looks scary, but we will going to learn it part by part so that you can understand it better. Just like before, I simply enter in the use template and your space has been duplicated. Now let's try to use the template here. Let's make it your start from use this template. Now, let's try to use the template. Being duplicated, you will notify. Okay, so it is getting duplicating. Don't worry, let's wait for some time, and let's get started again. Let's jump to the future. First of all, we have to understand how this prompt or workflow or we can say how this node system work. In this node, there is a product. There is two variation of the model first with the model and with the product. After that, as you can see, it is connected with this also and with this one, you can see this is the model with the product, and this is the model with the product. After that, go value, this is the product. This is the product logo, and all these reference has been updated in this sequence. This is the designer node. This is the special node that has been introduced recently, and this one is really perfect. How we can use it. Let's see. This is the image. Once you double clic it, you will see the image will pop up, this is the image, and all these are variation of same product and two models with the designer variation. Also understand how designer works. So we are going to create by understanding it, our own generation, how, first of all, let's create a product here. How we can create it, let's create right click text prompt. Or we can simply directly delete this node image generator and generate a product. Which type of product? Let's go for the mascara or Cajal. With the assistant, we are using assistant here give me prompt for the product with clean background. Let's try to use it as a prompt studio shoot, and let's try to connect it with the image generator with the image generator, select for the context max. 150 at, let's use for the image Jen not using image, let's use Nano Banana two. And let's go for the two variation right now. With this one, let's run it. So we have this product, and this one is looking better, and this one is also looking better. Not to worry about. So we have two images right now. Now let's go to our second image. That is what they have done. They have created a image that editor pose of a woman in tense direct gaze into the camera. Lip met they have used this Mt lipstick flawless. This is the specification they have done with this. And let's try to use copy. This is the prompt and they have used CDM 4.5 for the realism. Let's go for the image generator. Again, with the image generator, and I am using a reference of the character so that we can work with this specific character because it is a AI generative character, right? And we want the character consistency. This is the main goal of this generation. Let's try to use this generation as a reference. Let's try to generate it in the reference, and with the text prompt, let's see if I had generated a tax prompt. No, they have not used any tax prompt here. Now let's try to generate the image, paste the text here, and I do not want the Matt ipisty here. I I want help of this causal pencil with the mascara and top. Let's try it at a p of a young woman paste, yes. So this is the option I have paste and add the reference, use this pace for generation with better As. And makeup, matching the cardial editorial gauge into the camera, red, met linguistic flawless skin with natural texture, visible, soft diffuse studio lighting with slight warm, pale, natural background, ultra sharp focus on eyes and lips, high fashion beauty photography, and also, let's add wearing modern modern professional black coat, film gain texture, photorealistic no text, no product, no props. Now let's try it. Let's try it with Nano Banana. Let's try to generate it. So you have understood how I actually change the prompt, a minor change in the prompt, and this is the image that is simply click from the mobile and this is the output that we got. And this is the professional photo shoot. I love the output. It actually really awesome. Now let's download this one and I will share it with my wife and she will gonna love it. This is the output. We have got the character consistency. Just like it, we have got it. Now we are going to use designer note out of it. We have model product. They have generated up model and they are going to use it with the model and the product. Now, how they are going to use 7. Generate image with the product: Have the product, they have the product. They have the model, they have the product, they have the model, they have the product with the model. We have generated the product, we have the model, and now we are going to use product and model. We are going to create product and model. Now let's copy the same prompt here, just simply click copy. And with the new image generator, we are going to right click Image generator. Let's describe it, page here, and what they will use, they are going to use it as reference. This is the reference, and this is the product as reference. So we are going to use same technique for our workflow. This is the reference, let's use it for the reference. And also use this as for reference. You can see we have two generations. First of all, we have a pencil, cardial pencil with the tip slate is merged on cling, mat white background, soft like key studio. We are going to use studio short sleek mascara tube. We are going to use mascara tube. This is the main image, this is the main prompt we are going to use so that our generator will understand how it's going to apply it. We have to see the woman is applying Dali pistic not Dalit applying paste, the sleek mascara tube with no cap, editorial, commercial pose, unconventional camera angle, the lipstick has a size the mascara or we can see sleek mascara has a size proportional to the woman's hand. So you can see, we have two generation we have reference. First, this is the reference, you can see the connection, and this is the reference, you can see the connection. Now we can generate it. Google Nano Banana two, I'm using Google Nano Banana two. Let's see if they are using a yes, Nano Banana two, they are using, and it's perfectly working fine with my generation also. Let's generate with Nano Banana two now generate it again. The generation is okay, but I don't like that much. Let's try to use it again. And if it's not working, then we are going to use another model. Yes, this one looks better. Yes, Velvet less. Can we see the velvet less? Yes. This is the same image that we got, but we have a little Musca, it's working really fine. It's good. If you want to see the previous version, you can also click here. This is how you can look. This is how you can use. And the generations are good. Let's try it with another model with context, max Pro, Flux context, Max, best for Ps, working with reference images, uses reference images with reference images, and now let's try to generate it again. 8. Why choose Nano Banana over Flux: The Basic structure has been changed and that is why I don't really trust all the models. I actually trust the Nano Banana. That is why I trust the Nano Banana because the character consistency is really amazing with this. Women applies set mascara tube with no cap, at commercial pose, unconventional camera angle with sleek mascara, unconventional, let's change it to not unconventional Angle, let's change it to focused on I camera angle. The sleek mascara has si proportional to not focused. Let's try to better camera angle. Let's try it with this. The sleek mascara has C, proportional to the woman's hair, and with this one, if you want to change the setting here, you can do so. This is how you can open the settings, and let's generate it again. Yes. This is what I wanted. Better camera angle, focused on the eye. Can you see it's actually the same as we want. But I want it to be red with the red nail paint. Let's try to change it further with that. Let's try to edit this image. Dit image, add. Change the color of nails at nail polish on nails, same color as lips tick. Now, let's generate with the Google Nano Banana tool, and let's generate. Yes, now it's really cool. Now let's save it, add to Canvas. So image has been cannot connect with image, try refreshing the page. Let's not don't worry. Let's try to add the image that we have already. Upload it. Let's try to repress it. Let's repress this angle as it's not able to use the image, yes. Now we have a really good result. Now let's try to add the designer node that we are going to use in our project. Now let's a designer node, add a designer node here, right click, scroll down and you will see or you can simply search for the designer, start the template or an ampithe document. How you can create it no design has been created. If you want to create it, simply want to create. If you want to start from scratch, 9. Understand designer node just like Canva: With designer note, we have selected. Now let's try to add this as a reference as a reference, not this one. Let's try to. Yes, this is the one that we want to add, not this one we want to add, add it with this one. Let's add this one, use this for the reference. Yes, now this is the input and output we have, and this is the output we wanted for the reference. Now, you can see all these are updated and what we want, let's create it and it will ask some questions to us, where do we want to use it? We want to use it for our story or f our portrait. Let's use it four stories or. Let's use it four stories. Pixels, millimeter inches. It depends on our requirement. Let's use it for pictures now. This is anything, do we have? No. Now let's try to create. What we want it, we can add. This is the image, here is the image, and here is the product image. Let's scrop it out and use it here. Got it. And what we are going to do, we are going to add text here. Sleek mascara, we are going to add sleek mascara. Now let's try to. If you want to add the details, do you have the product details, you can add so. So this is what designer node has done. It has created a design for me. And if you look into it directly, they have so many elements, right? So you simply have to double click. It will open up. But if you want to edit them, you simply have to dt and you will see the same window will pop up with the generation products, and you can see all these are images that has been uploaded here and all the generation, you can update it here, if you want to remove it, you can remove it and adjust the image accordingly, remove background. You can edit the image directly. You can crop the content. So it is work just like Canva. You simply have to upload all these structure and images if you want it. These all things are uploaded. You can type here, you can add a box here, how you can add a box. You simply have to add a shape here like this. All the shape has been updated like this. And if you want to create a page, you can create a page. If you want to add specific abstract, you can add and add a text. If you want to add a text, you can do so. This is how you can add a text here. This is works just like Canva. You simply have to dragon drop the element, dragon drop the text, and design your own elements. That's all you have to do. I hope you have understand the basics of this designer note, and I would love to see what you have created for your own workflow. See you in the next lecture. Thank you for watching. 10. Getting started with commercial production: Today, we are going to learn about the makeup product at. Last time we have created the photo. Now, let's start with the making of product at. Let's use it. Once you click in on it, this is the template you can use directly, or if you want to create your own, you can do so. But right now, we want to understand how this work and we will get to know how we can create our own. Okay. So let's try to use this template, click on it. It will take some time about one or 2 minutes maybe, and the template will ready for you. Okay, so this time, it's getting really fast, and let's try to name it product, video. So this is what we got and what we have to create. So this is the branding. This is the video production, and this is the video hero. Let's welcome to making of Makeup product ad. Now let's try to use this now let's watch this video first. How it will work? Let's try to check this. Boss, baby. Boslos baby. Step in the room and I make a statement. No introduction needed for this kind of payment. That Glassy finish that high shine glow. I'm running the game yet. It's my own show. Call it what you want. It's Bos gloss, baby. Boslos, baby. So video is really realistic and you can't say that it is generated within the AI. It looks like we have created everything in three d of plus. All these are real, real, real person. But no, all these are created using AI. So lots of money has been saved by using this. Let's see Ait video. If you want to edit the video, you can do so. Now, let's start by using the video production. So once you click here, you will see nothing is open, right? So what we have to do, you can actually create pages inside the space. Where you can find or create the new page, this is where you can create the page. So this is the welcome page where you can see this is the branding brief branding visual on this is the index of this page. That is on first page, you will see Clientin branding on second page, video production, and on third page, video hero and final video. Closed showing the product meet the models. Let's start by using. Once you click here, you can see video production, client and branding, and video hero. Let's start. Let's drag it above it, and you will see client and branding should be on the first on the second and on the third. On this, you can see how many nodes and work production has been created. Let's open this one and you will see the product. This is the product one, two, three, four. This is the PNG, and this is the image that has been created, and this is the color palette we can use. Natural, fun, youth, empowered. All these are branding or all these are brand values. How you can decide all these n values, you can find it on the Internet. I will showing you advanced corporate branding so that you can understand where you can find it. So this is just an example. So this is a corporate company, this is an authorized company. So this is a software company where they create a software, and it is really well known company. You can see visual system and logo system. You can see how you can use the video, how you can use the logo, our brand, our principle, our story. So here you can see optimistic, retentless, brave engages, trust it. So all these are brand value. You can just copy and paste here. Or if you are working on different brand value or if you are creating a branding for another company or any startup, you can actually add anything here and create your own. How you can do that, you can simply right click insert shape add reference. If you want to add reference, you can do so. No, how you can do that. Right click. You can add a sticky note or stickers here. Let's try to add a sticky note. You can add the sticky note here. And if you want to change the color, you can change the color, type anything here. Extra large, you can change bold. You can do anything with it. Got it. So let's remove this one. So here is the client brief. The client has given you the details of everything the product, how the branding works, their moto, their background, their objective, and their target audience on which audience they are targeting and creative direction. You can add anything here. How you can add this one. You can simply right click and you can try to add the list. If you want to add the list, you can see the list. Got it. And now with this one, their product, they have shared the product images also. We are going to work with our own product, so you don't have to worry about the product. But right now you can see a product photography of a capsule, shaped lip, close, bottle, isolated on a seamless, solid white background, a bottle features a smooth cream color round it cap, translate, lower sexin field with the text on it is printed vertically in bold, white, probably. Type on the center with the number 03 positioned neatly at the base. All these are means or texture and background. You can actually check it out. This is really good to see. You can add the palette. You can actually create the image. Basically, what you are doing, you are creating the image using prompt and all these are information that has been shared by the client. This is created by the client. This is the client color combination and branding. You can actually see here, branding action, brand color, brand visualization, e. You can actually use this as a reference and use it in your branding. You can see that this is the color palette they are using. This is just an example and you can see everything in action. This is how everything works. Now let's go to another page. We have understood the branding a little bit of branding information we got here. But now let's start by using create production video. Once you got here, you will see 68 of nodes are working together, and this is how you can work in the production. And we are going to debug everything one by one, so you will not get confused. In your production level videos or in your advertisement videos, we are going to use our own references so it will help you a lot in understanding the video composition, images composition. Eth. Let's open this one. You can see there are four nodes here, what it's work. This is how it works. This is the final output. Start by Double click to video to watch and hear the finals. Just like Double click, you can open this video. We know already about it about this one production video production, this is the main step and this is where we are going to create our own video. Don't worry. We will going to debug everything one by one. I will going to explain you every step to generate, regenerate, and learn everything at once. If you feel overwhelmed, I recommend you to take a deep breath, take a coffee break, then come again because it's going to take a lot of time. Thank you for watching, see you in the next lecture, and we are going to debug everything in our upcoming lectures. S in the upcoming lecture. 11. Professional Look using prompt assistant: What they have done, they have used a reference here. What I'm doing, I'm not using everything at once. I'm replacing one model in this entire video, and you will understand how each model and each product does. So we are working with replacing this model in our current product video by the model that we have created the product before. So now let's start by using the character sheet, and there is one more step that we are using that we call the character set. Let's use the character sheet here. So start by uploading the character sheet. We have uploaded the character sheet here, let's increase it a little bit, and we have a reference here. You can add reference either by generating the image like they did. They have generated the image by using Google non Banana, and in 1k4k, it depends on your choice. Or you can add the reference by selecting right click, search for Stock and click here. A click on Stock, you will see photos, any expressions if you any confident or happy, sad, anything you have in mind, you can use it or you can use design also. If you want to use branding, social media, product, brand identity, you can select anything from here. If you have digital art background, you can use anything in your project. The advantage of using the stock images and videos from the pre peak friends, what they have done, they have used assistant here. So what assistants done they basically use these to format. So how you can see if they have used this specific image, there is a key to understand that once you click any of the image, you will not be able to see the highlight on the node. If you click this, the node is not highlighted. But once you click this, you can see the highlight node. This is how you can spot it out that this specific image and this specific image has been connected with this assistant node. So now let's try to use this image and recreate our images with the product and entirely new model. With this one, let's try to generate the image of it. But first of all, let's try to delete this one so we can understand the better. All these are references that has been generated. Once you click on it, you will see this is the prompt. Let me copy this prompt, copy and paste. Prompt for assistant. And what the prom says, extract oval style, lighting setup, and location mood from the full photo pojet Sess on focusing on maintaining background, consistency, cohesive, color grading and control light direction. And soft throughout or short skin should appear real and unretuched showing natural texture while keeping shine balanced and professionally managed. This is a polished studio shoot, so highlight should feel intentional, never excessive. And they have used Google Nano Banana three flash for this generation once you click here, and it will automatically enhance your prompt large with this larger prompt. So we already know how assistant work. With this one, they have generated using these two images. Now, let's try to replace this specific image. With this one, you can see what this prompt says, medium portage short of a young black woman with natural texture here, wearing soft gold, eyeshadow, and a colorful crossed top. Her skin shows visible texture, blemish, bright studio lightning with clean shadow photograph against a plain white background. So this is the specific chrome that has been used for this specific image. Got it. But I want this to be work for this one. How we can do that. Let's try to at the assistant here, right click, search for Assistant and with the assistant, use it as a reference image with this one, paste it here. Now, this is not a young Black woman. This is a young woven with natural textured hair, wearing soft gold eyeshadow with this one, let's see if it will work or not. This is not an eye shadow will not work. For this texture, as you can see, there is a texture difference between them. We are not using a golden eyeshadow. We are using eye shadow that will actually work with this color. Or we can say soft silver eye shadow and a colorful crosstp. You can see this is colorful crosstop, but we want to change this outfit, also. Let's use Gemini for it. Now we are taking the help of Gemini, as this is the product, where should we start? So help me with the character sheet of this person. Of this young woman in which portrait and mid size shot we can create using AI within Freepik and change the clothing and the hairstyle as per modern style not very flashy look, not very flashy look, but give it more like a professional give it more like a professional look. Give me the detailed prompt for it. Now let's try to use the thinking. Let generate the image. The prompt has been generated. Now let's try to use this prompt. For it. Let's copy and paste here. Now let's try to. Either we can generate the image directly, let's try to generate it and generate this one. Not this, let's try to remove this and try to connect with this one. Now let's try to use Google Nano Banana two. And now let's use the ratio that is best for portrait. And for this one, let's try to use three ratio four for the portrait images, or let's try to use it for 69 ratio. No it's 916 ratio for the story or for the Instagram. Now let's try to generate the image, copy, remove the assistant here and add the text prompt here and paste it here. Let's delete this one and simply copy and paste here. Now, let's see what we got here. This is one extra step we are working with. Why? Because no one wants to see the AI generated images. Everyone wants to see the consistency and the character they are working with. In the industry, and let's see if it will work or not. No, it's totally unnatural. Let's try to make it a modern and funky look. Let's try to add modern and funky look with light silver eye shadow on it. Now let's try to use the image prompt assistant here and as we can see, it is not working at all. Let's try to delete this one. Gemini has not actually help us at all. We have to use another prompt. Now let's try to add this prompt now rearrange it. I hope this time we got a better result. This looks professional and it's good to see it's look good. Now let's try to use this as for our app. 12. Characters Using Nano Banana: So let's duplicate this one, duplicate, duplicate, and move this one, and let's try to remove this. Let's try to replace this from this image, delete this one and replace it here. Once you drag it here, and drag this image here. Now, let's see, medium shot of the same woman up close to chest, matching face, skin tone, hairstyle and blue eye makeup, emphasize natural skin, texture, and detail, pastel blue background, soft studio lightning with subtrle shadow and realistic highlights. This is the new prompt we are working with and it is generating the image in four K. Let's try to use this image here and let's recreate it. Okay. So there are two options to generate also. You can either generate in each frame of it where it's connected and it will automatically create the image automatically create in each frame of the sequence and you can automatically create the images or the video in each frame automatically. There are two options to generate. First of all, either you can generate for this node or and if you select this one, it will only replace the image with this node only. If you select the second one, it will automatically replace all the images and all the videos within the node that it is connected further in our upcoming generation. So the choice is yours. If you have extra cart, you can do so if you have less time. If you have less time, you can change the entire model at once. Or you can change the model one by one and see the result if it's working for you or not. I recommend you and I highly recommend you that you should use this for this node only, not from run from here, because if you might not get the result or the variation in the image that you are uploading or the model variation, this will truly the loss of your credit. I recommend you to use this node only. Let's click this. Once you click it and wait for some time, the image look really unnatural and I don't like the result. This is why I ask you not to use this for entire sequence. So this is good result, but not actually up to the mark that I wanted. So let's try to remove this and I have to work with the prompt also. Let's try to remove this one also. The prompt, I don't want it. Let's use the prompt for my character. Let's try to use it from Gemini. Try sharing the different outfit and try sharing the different outfit and makeup portrait for the image, not creating it like grid. Keep it single image. This will be used for blow advertisement video. Let's try it. Give me prompt at least with three options. So we can try it out with Google and Banana. Let's try to use this. The subtitle text loveel I don't like this subtitle. I want it to be clean. I want it to be clean. Let's not paste it here. Let's try to paste here, use it here and see what we got here and remove this one. Generate. So I think we have not connected the reference. Let's try to add reference here. This looks natural. So this is the character that we got, and it looks natural. I like this one. But there might be a change in the image. I think this is a portrait of South Asian woman. So it has changed South Asian, so this is not from the South Asian woman. The high resolution portrait of the image from reference, maintaining a specific nose and bite studio. So it has generated the eyes a little bit. Let's try to reconnect it. After that, let's try to reconnect it with this one also. Let's see we got a better result here. Now let's see if we got a better result. By using this prompt also. Let's copy and paste it, paste it here and see a single high reason South Asian woman of the woman, not from reference. Let's try to remove it. I facial feature, nice, her dark style, her makeup minimal dw lipstick, hero coated vivid, juicy pink, wet looking she's looking slightly off camera to her left with a playful engaging smile. She wearing a clean of shoulder gradient plumage, and let's try to regenerate this image, run it. It looks really nice and clean image. Now let's try to use it with the third prompt that we got, copy and paste it here, connect. This one, denote this one of the woman, not from the image, remove this one. Now let's under it. Again. So the image looks really attractive. Yes, it shows me more confident. On the face, and it looks more professional. But let's not try this one. Let's try to use this image as this image. I like this one. This is the one I like. We can use it. So this is the image I like the most, so we are going to use this one. Now let's try to copy. And paste. The reason I'm using this character sheet is that because in the agency, you will not get the exact character sheet of everything. You have to create your own character sheet from the model they will provide you. And it is really important that you will learn how to use a specific model if the agency or if the client will give you. And they will not ask you to create a random AI generated images. They will ask you to use the model pace and this is why character consistency is really important. So we are going to use this one, and with this one, let's try to add this one and see if we got the result or not. But before that, let's try to change the prompt. Let's try to change it a little bit. Oh, let's try to use this directly within the prompt, I guess. With this prompt, let's try to remove. So this is the prompt that has been connected here, right? Let's try to add a more specific color to it with this one. A young, vibrant close up portrait of the woman with short texture, alpha curt, silver hairring tilted her payful pose. She's wearing white sleep flash, seen illuminated background energetic emirs. Now let's try to use it in this one. With this one, let's try to add image that has been generated, and let's try to use it with this node, the result that we want it this time. The hair style has been changed and it looks more natural to me. This is good. The change is really good, but let's try to change the prompt a little bit, like we have nodes Afrocal adorned with silver earrings. So this is not we wanted, and let's try to use it again with the same prompt. And I'm just testing it out and experiment with it. If we want to change the image here, we can use directly this image because I like the image and the natural look on the face, it's really good. We can use it directly here also. Yes, image is good. No doubt about it. Let's try to use this specific image that we have generated as we have tried to generate this image with this one. But I don't like the exact result, but let's try to use our another model here. So this is the model that we have generated, and let's try to remove or replace it. Let's try it with this one. Use this image directly here. And remove this node with this one. They have connected this specific node and this specific node here, and this is connected with two nodes at same time. This one is connected, and this one is connected. And you can see this is the image which is connected with each of the frame here, one, two, three, four. Let's try to regenerate our image with this one. And we are going to change the product a let's try to use the product 13. From Phone Snap to Pro Product Image: This is how you can create your own image and replace it everything at once, so you can remove everything with this image. Let's try to create our own specific dedicated product video so that we can actually use it for our UGC or production level ad. Let's try it. So this is the image we have created. This is the image we have created by using the characteristic. Let's try to copy and paste it, and we are going to use this let's to remove this one. We are going to use two or three images. Let's try to give me more options, give me more prompt options with different output and different makeup style. Also give me and also remove the Asian woman text as I will use my reference. So meanwhile, Gemini is creating the prompt. We are going to retouch images that we got from the client. So we have two images that has been clicked from the mobile itself. Now let's try to save it one by one. So both are images that has been clicked. Now we are going to upload it in Freepik. With this one, right click Image generator, use Google Nano Banana. And with this one, select selects, Selectis type product A, B from both the reference image, create a clean background image of this product consistency, same and product image should look professional studio image for online demonstration. Let's try what we got here. Let's try to run it. And I hope that we got the perfect result of this product that I have clicked from the mobile you can see from just a single image, I have extract a really good image out of it. Now let's try to use this one for our product image, we can see here. Let's try to use it, move it here, and we have the model image. We have the product image. Now our next step is to add the model as well as the product. So this is the product and this is the model. Let's copy it and paste it, remove the connection here, and we have specifically used this image with this one, we have to use the image that is created like this. So it is not a met and liquid type lip low. It is just like a lipstick. So we are going to use a stock image here. Let's see, right click Stock and see Lipstick. Yes, like this. Just like this, just like this. Yes, it is just like a lipstick. Yes, this looks more natural. And let's try to use it in our sequence, and this one is more natural also. Let's try to add this one also. Add this one, add this one. Now, let's add this in the media. 14. Product Images from Scratch: This is so I like and add this into my media. Let's wait for some time. Somehow, it's not working. Let's see if we can get it or not. Let's try to repress it. I'm trying to add a single image here. Let's try to add again. I will work or not, lipstick Search, let's try to add this one. Somehow, this node is not specifically work here. So there might be some error with the stock. Let's try to add from the Freepik itself. Let's try to directly upload it here, let's try to use it lipstick. And you will see so many examples here. I like this one. Let's download this. I like this one also. Download this. In the medium size, you can download, save it, download. Let's use the JPAC JPEC download it. I like this one also. Save to use this one also with the medium size. There are so many reference, we can find it. I like this reference also. Let's try to use this one, medium size. Save this reference. I like this one also. Save the closelips, save it. So there are so many references that we have added. Now let's go to here, upload all these reference. So all these images has been uploaded with this one and with this one, we have product, we have model. This is the image that we want to generate right now. Let's try to use this and by going to image generator, connect this image here, connect this image here, and I want to add the text here. Simply a text head text. Let's not use this one. Let's try to add the image. This is the image, use this as This is the main product and use the characteristics of the reference image from this also image is just like lipstick, but in light pink color, not light pink color. It will be in red color color. It is simple white cover. Let's try it and run the prompt. Meanwhile, let's try to add assistant here. In the assistant, this is the image prompt and describe this image. Run the prompt. Meanwhile, I have asked assistant to describe this image, and if it will not work, then I will go to ask my assistant to help me with the generation. Everything is awesome. It's working really well. I like this very much. I like the output of it as it got the characteristic specifically from this one, and it keep it white. As you can see, the studio light product shot on the bold red lipstick are rich mad scarlet, bullet set in the polished gold tube. Is I not it refer as tube, but I refer it as cover position upright against a texture, mire of matching pigment, the creamy, velvety, smur with rough irregular edges, provides dynamic contrast to the silic metallic caching, and the transparent check pop background isolate the product for easy composition. So we can actually delete this specific information because it's actually explained this check background also. So if it's not worked at all, we are going to use this one. So it is not that red in rich color, but for now, we are not going with the perfection. Let's go with this. Output. If we are working with the agency, then they will ask you to change the color of this mat. As once you apply it, it will just look like this, a light pink, glossy look. So we have the image. We have the product image. What we want right now, we want the model with the product. This is what we want here right now. Here, what I did, I simply use the prompt structure from the previous image that has been generated and simply copy and paste in my generation. Let's copy. But first of all, let's try to generate so many images of my model also. Let's try to generate it. Have to generate first the images. Let's copy and paste it here, remove the text done, and let's try to create so many images of my model right now with the product. So this is the product copy and paste. We have now product and the model. Let's try to generate it. So we want the different poses of our model. Let's keep it here for this one, and make it with this detailed portrait of a woman, copy and paste it here, paste the text, image generator, add this for the reference, and add this for the text. And let's read the text on high resolution professional Studio women. So this is the PNG, I don't want it, maintaining high special clean structure, shoulders, leak plows, flattering. Let's try to add the image generator for this one. Let's try 15. AI Product Photography Generate Consistent Model: Is the new output I got, and I think it's a little bit different from the image that I got. Let's try to not use this as a reference. We are going to use our character state for the reference because it would give more information to my output. And let's try to make it nine to 16 ratio for this one and remove this, try to use this for the reference here. So we have connected our image reference directly here. And for this one, let's try to delete this one. Now let's try to use this as a character. Use this for reference. Let's not use this one and let's try to regenerate it with different output. Also, let's try to copy this also image generator describe the image here again, try to add this specific image with the prompt, you can directly add it, and there is no image one here, so it will randomly add the image from itself if I add it. Now let's try to generate with this. Also with the characteristic, I copy. Let's try to add this structure of images in this next generation also, right click Image generator, add it here with this one, a texture, generate it texture here, remove the PNG information from here also. And for the parents, use this for the reference image here. Yes, it's connected, run it. This is the image I got. We get the result in sometime in this image also and in this node also. This is the image that we got. It's look good, but not that professional look. I want a professional look and a background like this one, and like this one, and like this background. Does it really work or not? I think I forward to generate the image within this node also. All these images are generated. Let's try to copy and in the image generator, paste here with this one. Try to add image reference and see, let's read the prompt one so that we can understand what it's trying to convey. Studio beauty portrait of a young woman with fair skin, visible fake and few natural blemish. No, we are not using this studio beauty portrait of a young woman. At reference this, yes, it looks better, and we are going to use a natural bremicrate platinum Blonder here. No, we do not want thehir description and wing eyeliner. Yes, we can use it with closyndelip sea vias rib, lime green top against a soft pastel pink gradient background lit with clean high end cosmetic stud lightning her hands are brought forward toward the camera holding the lip gloss in the foreground as a present to it to the viewer. The product is centered, labeled facing the camera with a subtle shadow tap of field that keeps her face sharp and closy slightly closer crap surely detail on her left, chunky twisty, golden. Let's try it with this reference and with this reference. So everything has been done. Now let's try to pray it around with this image generation. You can see how I'm connecting each node every time and you can see the cluster is creating automatically. 16. Pose Replication Mastery: Imagination is awesome, but I do not want these patches at all on the face. Let's try to reuse it and re edit our specific prompt. Visible fake and few natural blemish. I don't like this. I think we have to remove this one and bold minting with glossy nude lips. Yes, this is the lips, likenatural lips and keep her expression prayful and slightly potty. Pouty eyes glazing sideways. Yes, but let's try to regenerate it. So I simply copy this prompt and paste it here with the reference image directly connected to my new generation. That's all I did. Let's see what we got here. Let's not wait for this creation. Copy this prompt here. And with the reference image selected, generate image again, the prompt, I just copied, not here. Let's delete this one, paste it here. First kiss bow playful with tail chin, one kis, relax, background, portrait, report or realistic. And let's generate this one. A. What we did, we did the same pose, but with the focused one. Now let's try it with this one. Yes, generating it's generating the image. With this character, we have to use this. These three pose I want it. It will going to apply the lipstick on her face, right? Let's try to use this one first, copy. I'm going to use this one. Image generator, paste the image here and see use blue background girl in the subject. Yes, I have added the age match the exact pose frame of the yellow background wrapping light. This is for this portrait, right? But I want it with this portrait. Let's change the prompt a little bit. Use the blue background girl as a subject, use the girl as the reference. Girl as the subject match the exact pose and framing of the yellow background reference. No, close up tight beauty close up. Yes, it's already done it. Match the tight beauty close up. It's not match the tight beauty beauty, close up, head slightly tilted, eyes looking at camera, lips relaxed with glossy fuchsia lipstick. Yes. Her left hand is raised next to her cheek holding the same lipstick product. This is the same lipstick product. Yes, this is the same lipstick product, and I have connected this with this one with this one. Got it. With the same lipstick product. It is not compsy to add. Don't worry. Label facing camera cap on top, bottle close to the lower cheek, Zine finger. Positions naturally around the bottle, like the reference, around the bottle, like the reference. I don't know what is naturally around the bottle, like the reference. So like the reference keeper style. So I want it and paste it here. If we get it. No, it's not working. Those space to space copy and paste will not work, and I think we have to download this image. Download this, save it, and upload this image here. And for the reference, Freepik use the reference and which reference is talking about, we have to add it here, the Freepik. And if we have too many images, then we have to also add for the lipstick also vertical at this one, keep her hairstyle, earring, necklace, makeup, blue eyes shade and skin texture consistent with the blue image replacing only the pose and hands, product placement, match, the yellow reference. I think it will not work. Let's try it if it will work or not. If it will work, it's good and if it will not work, then we have to change the prompter a little bit also. So this is actually picked the same image, but I don't like this. Let's try to remove the image here and run it. Once I remove this image that it took actually the same pose and everything from this image, let's see. So this was the last image that we used, and I don't like the result. The model has been changed or the model is really same as like we have used for the reference. But right now, this is the image that has been selected for this one. This is the image. And now you can see the output is really good with the image, and this is good to 17. Product Shots from One Image: Has been generated for this one also. And right now let's try to create another pose with this image. Let's try to copy it, copy the prompt, and you can actually generate all these prompt within the Gemini and Chat GPT. Once you ask for it, they will give you the result. But I'm not going with all these prompting work. We are actually creating the images with our own character, so you don't have to worry about the consistency or if the project comes with a new character or with the new model, actually, you know how to work with that specific model. Or you can use a man, a woman, anything, you can use it. It is just a mobile clicked image. From just a mobile clicked image, we have achieved so much higher level of output. With this one, let's try to generate image, paste it here. Yes, past here and you can see this is just image. We are going to use this image and try to generate the image again. Photorealistic, first case, and let's try to generate the image again. But with this, use it with two K and run it again. Result is not what I expected because I expected the result of this. Let's try to change the prompt a little bit. First case, chmove leg, shoulder, show only pace no hands and product. Let's try it again. This is the pose that we got here. But still, the hair has been changed and this is the color set and the clothing has been changed. But don't worry. Let's try to use another pose for our model. This is the model. This is great image. All these four images are great, one, two, three, and also let's try to create another output of the image. Let's try to use this prompt so that we can play around with the hair also with the image generator, delete this specific node, at this hair. With this image generator, let's try to add this image connecting with this one again, at medium poahot of a woman with slides side parted here, greening glitly eyeshadow, and no freckles I want it. She looks to camera. She's laughing with her mouth open, showing her teeth, sevi neon green hair clip, long silver earrings and zebra printing high neck top with line stretching, bright white studio background with clean lightning. Mouth opening, mouth open. Teeth Yes. Now let's try to generate with this one. 18. Product CGI Images : Like this image, so the result is good, actually same as we thought, but it's not like the real touch that I wanted from the model because you don't really portray the model or don't really show the kindness of model on the face. So let's not use this one, and we have this, this, this. And let's try to use another one, if we got it. This one. No, this one, this one, no. Let's try to use Freepik for the reference. This one, I like this one. Let's use this image for the reference. Now for this one, we have this and this let's use with the image generator with this reference image, add here. Now let's try to add the description here. Like use this is the main character, use the reference image and use the reference image, change the pace of the character. In keep the consistency same. And also, I want this to be used as a apparent for this image. Let's use this one, done it, replace the product with this. Now let's try to generate it again. The results are really good. As you can see, the characteristics has been used like this, but let's try to use another image of the model with the same image generator. Let's try to generate the variation here. Let's try to generate the variation of the image variations. And if we want to generate 16, nine is to 16 with height three is three, and if we want any of the variations, let's try it with one is to three. Let's try it with two is two variation. Now we are using the variation here. It is really important. Let's try it with this one also. Variation, two is two, two, nine is two, 16 expressions. Let's use it for expressions and generate again. We have used two types of variation, one with the expression, one with the reprime and also let's try it with this also. Let's try it with the variation. No, let's try it. With the variation with different output, and this is the input, and this is the input. So at one time, you can select only one variations for the frame. I'm not asking for reframe here. I want to ask it for custom. Let's not use this one first before moving to this one, let's try to see what we get here. Yes, one, two, three, four, five, six, seven, eight, nine, ten. Yes, the variations are good. Let's move forward with this variation also, and now it's going to reframe after that, we are going to edit this whole sequence within the Premiere Pro because as always, I'm a video editor and I have so many things in mind, how we can edit all these types of videos. So that is why I'm going to give you more variation of touch. If you want to select the specific variation, you can simply select and you can either open in preview or then you can use it as an ard wod and directly use in your video. Time out after 5 minutes. So let's try to nine, 16 ratio, two, three, expression or storyboard, let's try to use it as a storyboard, split images, yes, and now with this one, let's try to create a storyboard out of it only this node. So right now, what I'm trying to do here as it's not able to generate this time, so I'm trying to use it with the storyboard. Let's see if it can give me more variety of sequence for the video. After that, we are going to after that, we will jump after that after that, we will jump into our video generation part. Also, we have to edit this product video also in the three D. Like here, they have created the video and demonstrate the video how they can show the product in the video format and with the background, we are going to create our own product visualization workflow. This is the product variation. No, I don't like the result. Yes, I like this one. Yes, I like this one. This one, this one, this one, and don't like this one. I like this one. But I think let's change the color to blue one like we have. We are going to use this one, not that one. So the variations are good. No doubt about it, we are going to use this variation for our video. They have showed me the angle of the images that I think they have give me the result of not the storyboard, but the reframes that we have selected before they gave me the result just now. It's okay. It doesn't matter right now. Let's try to move forward with our product shoot or product demonstration or product visualizations. We have this product. This is the product we have. This is the product we have, and we are going to use it for our demonstration. As you can see, this is the video, create an image featuring just two products floating amongst jelly, sweet, pink product, and collar product. For this one, let's try to use the assistant to get the result. For this one, let's try to add assistant. For this one, we are using Chat GPD five, and we are going to use not the image, but the texture. Me ideas to generate product demonstration of lip bomb demonstration of lip bom. Let's try to air it. We are going to use the reference for the image generator. Again, the same step we are using connecting this node here. And let's try to generate it. I have not read it out. Honestly, I simply add it in my image generator, and I want to see what it will give me. And if I don't get the result, a better result, then I have to move forward with Gemini or Cha GPT. I don't like the result. Now let's try to move forward with the prompt here for product video shoot. I want prompt for image generation with PAPI and modern look for lip bum. Let's say we are going to use this, save this one, save this one, and use it in the here as reference image, this is the render product. Basically, I'm trying to see what we got basically what I'm trying to see, I'm trying to see any three D render. If we got here, this is the three D render that we got, and this is the three D render. I loved it so much. Let's try to open it and see if it has other renders also. Let's see with the modern gradient, and we can see Strawberry case. So it says strawberry case, right? Lipstick product. With strawberry. Strawberry, three D, render. Yes, we got the so many results here. But let's delete this one and simply create three D render output of reference image with gradient background and straw barry. Let's try to again generate this one and also try to generate a image here with the reference image, Theory render of product of same product with stro Bry in the middle of city. Let's try to render this one. And with this one let's try to add the image generator here and generate with this one also. 19. Commercial Product Videos: And I think I forgot to actually use the reference image here because it might not work as precisely as we want. But if we do not get the result, don't worry. We are going to generate it again. I do not want the strawberry in between. Let's try to. Same product with strawberry in the middle of Okay, in the middle of the city. No. I think same product with render of same product in the middle of the city, strawberry all around it. And make it strawberry all around it. Keep the product in focus. Let's generate it again. I like this one. But let's try to remove the background and keep the background. A simple one. I like this one also. All these things, I like this one. I like this one and this one. Yes. But let's try to add to add with the middle of the strawberry all around, keep it and keep the product in focus. Product size is bigger than the building. Let's try to generate it again. This is the image, and this is the image. This is with the reference, this is without the reference. That is why it is important you connect the reference image for better understanding of the prompt as well for the image generator. With plain background flying on tro berry above the clouds in pink colors regenerated. I simply add simple generation pomp here. Whatever my mind think of, I simply a it here. The generation is also good here. No doubt about it on pink strawberry. Let's try to generate an image also that this person is trying to kiss the strawberry also. Let's try to generate the image. Kissing the strawberry background will be in pink studio. Let's zoom in lightning. Let's not go this deeper. Kissing the strawberry. Let's try it. Yes, it's good. No doubt about it. It's good. So we have so many generations here, and let's try to use it this one also. We are going to generate this one also with this one, with this one selecting, generate a image with pink and white gradient back ground floating in middle with with red strawberry like jelly. Let's try to generate it again with this one and let's see if the image it's working. No, the image generation, I think might be fail here. Let's try to generate it again. With this one, we are going to generate the image here also. The image generation is good, but I want it to be let's remove. Let's generate it again. But remove mate color of lip, remove black mat color. Let's try to generate it again. Meanwhile, we have asked the image generation. Yes, keep the eyes close here, and let's try to generate the image that is kissing the straw berry, eyes closed, and for product shoot. Let's generate this one. And we are going to see no, the variation of the image is totally different. I think we have to. This is the selection I want. But with this met, remove meat generator, keep everything same except remove the background of lip pm that is color code Mt. Let's try to use it. If it's not work, then I will ask Gemini to help me with the prompt. I like this one. So I might not able to explain it. Let's ask our assistant so that it will understand the image and we can identify what we call it. So there is a way to describe this image and let's run it. A vertical studio, red light, be centered, musty rosy pink background. The white tube labeled with blue Heaven and love charm with shadow name strawberry and most rising tint iba reveals a vivid met red bullet. Met red bullet, okay, glossy downsp around the product. So glossy dony strawberry floating around the product, right? So with visible droplets while a textured swipe of matching red bums sits behind the tube, textured swipe. Adding da, paint. Tary contrast, the lightning is clean and bright, emphasizing the smooth lipstick surface, the future shine, and the overall fresh fruity cosmetic ethic. Let's use this and again, change. Let's let's let's again, try to did the prompt. Remove textured swipe of matching rubra seed behind the tube. Now let's try to regenerate it. I hope this time we got the result. And for this one, let's try to remove for this one also. So this is how you can identify the specific area of the image and change it according to your need. Now we have changed it. Now it's working fine. Now let's try to create a video out of all these images that we have. Now let's try to create videos out of all these images. This is the one I don't like this one. Let's remove this one, remove this one and we have. This. So you can see how we have created our own product video, so you can see how we have created our product visualization. And now let's try to create our own product video by using video generations. So we have to understand the prompt also, how it's working and how it will give us the result. Let's try to use our specific video generation here. For this one, a slow dolly in short with soft diffusing light, the lip glos tube, the lip close tube remain the focal point while drifting slightly. Let's copy this pro and paste it here in our video generation and our video generation, let's try to right click add Video generator with this one, paste it here, Noto, we are using Kling three point for this 20. Still to Motion Using Kling AI: Are moving forward to our video generation part. Video generation is somewhat tricky. If you think you can brainstorming your ideas into the video section. But right now I'm using Gemini to teach you how you can use or animate the video for the prom purpose. As if you have in mind in any particular video generation, you can type that and you can get the result that you wanted. But for this generation, I'm using right click. Let's move on it by clicking right click. It's not working. Yes, now it's working. Done. So with the video generation, I have used a video generation part. With this one, let's connect it with this cling three I'm using. And for the reference, it's not the reference start image and image references. Okay. Now with this press type reference image sound with 720 and generation 550 with 108 generation will be 700. So I am using 720 P four just now for the reference image. Image kiss the strawberry looks into the camera with the smile. Smile. Now, let's with ratio 916, five second cling 3.2 I'm using. Let's make it work. We have to simply test it out if it's working with this or not. With this one, I want the generation, video generator, start frame, frame. This is the end frame I used, and this is the starting frame I've used. And I've used the Kling 3.2 model and using the sound as usual. So this is not working at all. It's used the afference, but it's not the video that I wanted. So I have to use a model with the generation start image. It added the sound, but I don't like this animation at all. Let's try to create another. Let's see if it will work or not. With the start and end frame, let's describe it using Gemini. No, let's start it using assistant here. Use assistant for this one. And with this one image and image. With this, prompt for both animation, both video for Kling three point model in which it gave natural movement. If it's not work, we will move forward to our next step on advanced step. But we are testing it out with everything we have right now. So with the start image as a reference, let's see. It has changed the face of the subject a little bit. Let's see how we can use it, but vertical close up of 8 seconds. Now let's try with this image and with this text, let's try to animate it, expect ratio. Let's try to animate it and I hope the generation will work. Yes, a little bit of video transition we need it here. It's look good. Yes, it's look good. Let's try to animate. And with this, let's try to use this frame also. I like this frame. And we can see if we use this prompt as for our video video generator generated pasted reference here like this. Close up, paste studio, soft inward manner cinematic movement, fork no Waterbuk, no clips and see if it will work or not done. So what we are doing, we are basically creating so many videos and see which video will work with us done. Move it here, move it here. Can you see the hair style is different here. Let's see if we got any thing different in our prompt, close our beauty. You So let's read it out, close of over Beauty video of young adult woman in the pestle studio, soft pink background matching the reference face and makeup, sees presses and squeezes her cheeks inward with both fists, puffing her lips, slightly, creating a cute, playful expression while keeping her hands on her cheeks. She tilt her head side to side. In a small, childlike manner, gentle rhythmic eyes glazing slightly off camera laughing at the end, smooth cinematic movement with a subtle zoom in and zoom out sync to the heat push in on one tilt, pull back on the other soft diffuse light, shallow, tapped or field, crisp skin texture, natural freckless high detail four k, 24 FPS table, framing, no cards, no tax, and no watermark. And there is a change with the here, right? Let's try to regenerate it and see if we got the result right or not this time. I have copied this frame. Right click, Video generator, with this one, connected here, and with this one. Connected here and connected here. Let's try to use another model here. Now, let's try to use this one disconnect this. And again, I'm trying to use the grok here. Let's see if we got the result right here. And with this generation, I'm using a closer beauty video of young adult woman in the pastel studio. No. Let's try it too. There is this frame woman uses the lip bum on her lips natural movement, and then smile. Now let's try to generate. And I hope this will work. Yes, this looks cool. I love it. Don't worry. We are going to use this frame. Easy. Basically, I'm going to use this frame by frame animation in my video, so I'm not going to animate it. And also, I have to animate this video and video and this video. Let's try to animate this one with the video generation. Let's add. I love it. I just want this frame Olney, and let's move it to this frame. And with this generation, let's copy and paste it here. No connection, I need it. And I'm going to use this frame here, video, for text. Let's try to use the reference frame. And I'm using Groce here because I'm getting better results from Grock Mas the reference frame and as the reference frame, and come very light brown hand comes and pick up the blue heaven product. And let's see if it will generate or not. Otherwise, I'm going to use Kling. We are basically testing out with other other models. Whichever model give me the best result, I will use that and edit my entire sequence of advertisement. It's cool. Yes. It's amazing. It's really amazing. Now, let's try to again, let's try to use this same but I have to change it, pick up the product, and bring out the product from the frame. Let's try to regenerate. No, I think we have to change it a little further or not. Let's see. Let's download. First download this one. Don't worry, we are going to animate. I think we have not get the result, right? Let's try to use it with Kling and wait, wait, wait. Now, I think it just updated right now. It looks amazing. Now, let's save this one also. Done. So we have saved this. This is the frame. This one I also saved. Done. This one I also save. I love this one also. Done. Now, let's see. Yes, this one, save this one. Yes. It's like, save. But I want it to be like, let's change this frame and also 21. CGI style videos from images: Image generator, and we are using the frame. Add this and add this. We have again, add this and this frame, and I would like to replace the strawberry in with the reference product inference product in this strawberry done. And now let's try to run it. Basically, I got the idea just now that I'm going to replace this by the product. It's totally depending on your ideas and how you think of everything. Yes. Now, let's try. Same frame, same logic, yes. Now with this one, let's try to animate with this video generator. Again, reference image. Yes. With this, I'm using a Grock here. Again, let's try to reference image, kisses the product and smile. Again, let's try to animate. It looks like it's going to eat this, but don't worry. Yes, yes, we can use this frame, right? Because it matches the frame perfectly. And let's try to download this video first. And this video, again, download this one. Now we are going to use everything at once, wait, wait wait. I forgot to animate this also. Let's try to rotate this into 360 format, and let's try to rotate this one. Have we generated any images yet in this one? Yes, yes, yes. This one and this one. Now let's try to animate this with the 360 rotation. Right click video generator, generate this. It's so messy at start, right. But once you created all your videos and the generation has been completed, then you can arrange them one by one in each group, just like this. Right click, make a group of it, and you can name it all. Now let's try to animate this with a frame. But I want it make the blue heaven, lip bam product, rotate, rotate 360 in motion, natural movement of clouds and throw barry levitating. Now let's try to animate this frame. It's superb. It's really superb. Let's save this one. Now, let's try to animate this. Whatever source tinted up, glossy, overall, fruity, aesthetic. Let's try to remove or give animation to this frame for video model. And let's see what we got here, and let's try to generate the text. Right click Video generator again, reference frame using rock, text here, sleek tube charming. Let's try to animate it first. And what we get it I don't I don't really read much about the video generation, but if you have in mind, you can read it out and change it everything. It is just I'm teaching you everything so that you can learn that easily. I always create videos for the corporate clients. Where they need everything in perfection. It's good. It's really good. The randomness in strawberry jelly is good. How the product is levitating, it's good. Let's save this frame also. You can hear the sound. It's also did the sound here. We have this video generation also, where I think I have already saved this both the video. 22. Using Internet for Inspiration: Add another film. I would like to add another fime here with this. Let's try to connect it with the video generation. In the video generation, it's working automatically, right. Then we have to change the reference image to our 361, but don't worry. Let's try to use it in the image generation also, and reference drop into the show. By shake and now generate. I just looking for another frame so that we can work it out with so many visuals because more visuals means more eye catching, more retention, and the more the visuals of the product, it looks really expensive products. If you have more if you have so many frames or so many visuals in your product videos, it feels like that this advertisement is expensive. It looks unnatural. Let's try to delete this one. No, let's try to generate it. Let's drop the product into the strawberry Let's try to use assistant here. No, taxate. Let's try to use assistant here. Again, a creative frame needed for the image or the product demonstrate keep it very creative and eye catchy. Let's try to see what it will give us. Let's generate it and now again generate this one. And it feels like it will not work at all, why because I have forgot to add the reference frame so that our assistant can understand what product I'm talking about. So I have to regenerate it with my product also so that it will give me a better result. Let's try to at the reference image here. No. Let's try it again with radio. Let's try to animate it. Not animating it. Let's try to regenerate the image or the frame. After that, we can animate it. It's really simple. I don't like the animation at all. Let's try to use free pick I just want it to be a ending frame of my product. That's all. From the Internet, from the lipstick, I found a reference so I can use it in my video, try to screenshot it. Let's screenshot this frame, and in the Gemini, I want this frame as a prompt for my ad video. But I want a change that my product drops on a light pink powder. Let's try to use Gemini Fast and again, use it. This is the right example that I asked the prompt. Specifically for prompt, I gave me the image. That is why I don't like this. Let's try to save this and use it in our free peak space. So once you with this frame, right click image generator with these spaces, use this in the reference. Again, our product in the reference, Google and Banana now replace product in this with the product and remove the text and move the text, LialParis again, let's try to generate. It looks better. Now again, let's try to use this with with the reference image image generator, remove the product in reference image and make it like nothing happened to this powder frame. Again, I don't like the result. Let's try to generate image again. Remove all the chaos in the frame. Let's try to generate it. If it's not work, then I'm going to create a simple frame of color and animate that. Let's see what we got here. I think we got nothing. No, we got nothing. Let's try to generate the image again. Yes, I got the right result for now. Now let's animate both this frame. With a generation of this video and data, and I'm using two point Kling 3.2 with the start frame and with the frame. Before that, let's try to use our assistant because assistant give better result in my video production, so I'm using it again. 23. Good result in animation a perfect prompt: And with this one, let's try to delete this one because I don't want this one also. With this product fall onto this and powder is spread all across natural movement. Now let's try to use it and enhance our prompt using a C stand. Now let's try to add this text prompt here and see if we got anything from this animation. Now, the video has been generated, let's see. It looks not that much good. Let's try to see the video has been generated. No. I it's not that good. We have used Kling. Let's see how we can define the prompt here, defined prompt and space node sequence for the t plaque and the product reference image text to prompt photorealistic closer of the liquid surface, viscous liquid surface. We don't want as cus surface as a lib um tube plunks down, catenm not radial, dynamic pla and no pulse I wanted here. And natural powder, spread all over product center cat mid impact with suspended drops, microplass, no drops. I wanted a Product centered, I close, press, energy stake, follow the fluid glossy wet highlights, no, and top left high sado daptofel. Yes, I want it. And rich palette with supple mesena highlights, ultra detail texture, fluid flow simulation I wanted in here. Drive is Plass shape from the product. Yes, radial particle drops. I don't want any particle drops, motion blur, I don't want it. And after that, I want it with a denoise amount for clean highlight detail output export at high resolution with Alpha if needed for further layering. I don't want anything, I just want to create it again and see if it will work or not. Let's say if we got I don't want any liquid fluid here, so I have to regenerate it. I don't not want any liquid. Let's try it again. Cross natural movement, but no liquid needed. Let's try it again. I hope this time we got a better result. Now I like this animation. Let's save this. 24. Project: For project, I want you to create any of the video. If you want to create a complete video, you can do so. Within the Premiere Pro, I have created this. Boom. Da da da da da da da. Wow. Who. Stop. Bye. So this is the Premiere Pro video that we have created. And for your project, you simply have to create a image, or if you want to create one video of CGI, you can do so and share it in the project section. If you want to create a complete sequence video, you can create and share it in the project gallery. I would love to see that and give a feedback on it. I will be sharing this project file and all the files that I have used here in the project and resources tab Blow. You can download them, unzip it, and use it as your convenient. If you want to change the sound, you can do so if you want to change the graphics, you can do so. It's totally your creativity. I just want you to see what you have created. Also, in the resources section, you will find this prompt library that we have used within the lectures. You can experiment with them and try to be creative with this. And if you have any problem, if you have any questions, you can ask in the discussion. So see in the next class.