{"event_embed_url":"https://heygen.wistia.com/embed/live/events/dlquslz1q6/audience","hashed_id":"dlquslz1q6","media":{"accountId":2397428,"accountHashedId":"69y4jlt51m","accountKey":"wistia-production_2397428","analyticsHost":"https://distillery.wistia.net","formsHost":"https://app.wistia.com","formEventsApi":"/form-stream/1.0.0","aspectRatio":1.7777777777777777,"assets":[{"is_enhanced":false,"is_composite_layer":false,"type":"still_image","slug":"still_image_1200x720","display_name":"Thumbnail Image","details":{},"width":1200,"height":720,"size":425066,"bitrate":0,"public":true,"status":2,"progress":1.0,"metadata":{"av_stream_metadata":"{\"FastStart\":false,\"Audio\":{},\"Video\":{\"codec\":\"png\",\"colorPrimaries\":\"bt709\",\"colorSpace\":\"gbr\",\"colorTransfer\":\"iec61966-2-1\",\"height\":720,\"pixelFormat\":\"rgba\",\"rFrameRate\":\"25/1\",\"width\":1200}}"},"url":"https://embed-ssl.wistia.com/deliveries/4684051b357fdd983ddd9f64e4965744.bin","created_at":1790787448},{"is_enhanced":false,"is_composite_layer":false,"type":"original","slug":"original","display_name":"Original File","details":{},"width":1920,"height":1080,"ext":"","size":2228988284,"bitrate":3590,"public":true,"status":2,"progress":1.0,"metadata":{"av_stream_metadata":"{\"Audio\":{\"channels\":2,\"channelLayout\":\"stereo\",\"Codec ID\":\"mp4a-40-2\",\"codec\":\"aac\",\"codecType\":\"audio\",\"duration\":4850.578866,\"sampleFormat\":\"fltp\",\"sampleRate\":44100},\"Video\":{\"avgFrameRate\":\"30/1\",\"bitRate\":\"3534967\",\"Codec ID\":\"avc1\",\"codec\":\"h264\",\"colorPrimaries\":\"bt709\",\"colorSpace\":\"bt709\",\"colorTransfer\":\"bt709\",\"duration\":4850.533333,\"height\":1080,\"pixelFormat\":\"yuv420p\",\"rFrameRate\":\"30/1\",\"width\":1920}}","max_bitrate":4996384,"average_bitrate":3676242,"early_max_bitrate":3432352,"time_to_ready":109.425853223},"url":"https://embed-ssl.wistia.com/deliveries/d1c17d04071daee8b3ff7648a4f3507b.bin","created_at":1792516910},{"is_enhanced":false,"is_composite_layer":false,"type":"iphone_video","slug":"mp4_h264_249k","display_name":"360p","details":{},"container":"mp4","codec":"h264","width":640,"height":360,"ext":"mp4","size":155109900,"bitrate":249,"public":true,"status":2,"progress":1.0,"metadata":{"built_by":"render-pipeline","max_bitrate":803848,"average_bitrate":255820,"early_max_bitrate":194400,"av_stream_metadata":"{\"FastStart\":true,\"Audio\":{\"channels\":2,\"channelLayout\":\"stereo\",\"Codec ID\":\"mp4a-40-2\",\"codec\":\"aac\",\"codecType\":\"audio\",\"duration\":4850.588231,\"sampleFormat\":\"fltp\",\"sampleRate\":44100},\"Video\":{\"avgFrameRate\":\"30/1\",\"bitRate\":\"121287\",\"Codec ID\":\"avc1\",\"codec\":\"h264\",\"colorPrimaries\":\"bt709\",\"colorSpace\":\"bt709\",\"colorTransfer\":\"bt709\",\"duration\":4850.533333,\"height\":360,\"pixelFormat\":\"yuv420p\",\"rFrameRate\":\"30/1\",\"width\":640}}","time_to_ready":73.422143312},"url":"https://embed-ssl.wistia.com/deliveries/c5b83b10db1a5cb80cf9f103d6b393f17d38111d.bin","created_at":1792517019,"segment_duration":3,"opt_vbitrate":1200},{"is_enhanced":false,"is_composite_layer":false,"type":"mp4_video","slug":"mp4_h264_190k","display_name":"224p","details":{},"container":"mp4","codec":"h264","width":400,"height":224,"ext":"mp4","size":118520470,"bitrate":190,"public":true,"status":2,"progress":1.0,"metadata":{"max_bitrate":401816,"average_bitrate":195473,"early_max_bitrate":166240,"av_stream_metadata":"{\"Audio\":{\"channels\":2,\"channelLayout\":\"stereo\",\"Codec ID\":\"mp4a-40-2\",\"codec\":\"aac\",\"codecType\":\"audio\",\"duration\":4850.588231,\"sampleFormat\":\"fltp\",\"sampleRate\":44100},\"Video\":{\"avgFrameRate\":\"30/1\",\"bitRate\":\"60977\",\"Codec ID\":\"avc1\",\"codec\":\"h264\",\"colorPrimaries\":\"bt709\",\"colorSpace\":\"bt709\",\"colorTransfer\":\"bt709\",\"duration\":4850.533333,\"height\":224,\"pixelFormat\":\"yuv420p\",\"rFrameRate\":\"30/1\",\"width\":400}}","built_by":"gen3-split-n-stitch","time_to_ready":184.175542198},"url":"https://embed-ssl.wistia.com/deliveries/cfedc0c09e1d7b705cc870d0a8239ec1246b5332.bin","created_at":1792517019,"segment_duration":3,"opt_vbitrate":300},{"is_enhanced":false,"is_composite_layer":false,"type":"md_mp4_video","slug":"mp4_h264_344k","display_name":"540p","details":{},"container":"mp4","codec":"h264","width":960,"height":540,"ext":"mp4","size":213999514,"bitrate":344,"public":true,"status":2,"progress":1.0,"metadata":{"max_bitrate":1232088,"average_bitrate":352946,"early_max_bitrate":244440,"av_stream_metadata":"{\"Audio\":{\"channels\":2,\"channelLayout\":\"stereo\",\"Codec ID\":\"mp4a-40-2\",\"codec\":\"aac\",\"codecType\":\"audio\",\"duration\":4850.588231,\"sampleFormat\":\"fltp\",\"sampleRate\":44100},\"Video\":{\"avgFrameRate\":\"30/1\",\"bitRate\":\"218406\",\"Codec ID\":\"avc1\",\"codec\":\"h264\",\"colorPrimaries\":\"bt709\",\"colorSpace\":\"bt709\",\"colorTransfer\":\"bt709\",\"duration\":4850.533333,\"height\":540,\"pixelFormat\":\"yuv420p\",\"rFrameRate\":\"30/1\",\"width\":960}}","built_by":"gen3-split-n-stitch","time_to_ready":212.432141729},"url":"https://embed-ssl.wistia.com/deliveries/666a382ca16209760886600c70e92109f7043e77.bin","created_at":1792517019,"segment_duration":3,"opt_vbitrate":1800},{"is_enhanced":false,"is_composite_layer":false,"type":"hd_mp4_video","slug":"mp4_h264_459k","display_name":"720p","details":{},"container":"mp4","codec":"h264","width":1280,"height":720,"ext":"mp4","size":285105298,"bitrate":459,"public":true,"status":2,"progress":1.0,"metadata":{"max_bitrate":1804392,"average_bitrate":470219,"early_max_bitrate":293080,"av_stream_metadata":"{\"Audio\":{\"channels\":2,\"channelLayout\":\"stereo\",\"Codec ID\":\"mp4a-40-2\",\"codec\":\"aac\",\"codecType\":\"audio\",\"duration\":4850.588231,\"sampleFormat\":\"fltp\",\"sampleRate\":44100},\"Video\":{\"avgFrameRate\":\"30/1\",\"bitRate\":\"335679\",\"Codec ID\":\"avc1\",\"codec\":\"h264\",\"colorPrimaries\":\"bt709\",\"colorSpace\":\"bt709\",\"colorTransfer\":\"bt709\",\"duration\":4850.533333,\"height\":720,\"pixelFormat\":\"yuv420p\",\"rFrameRate\":\"30/1\",\"width\":1280}}","built_by":"gen3-split-n-stitch","time_to_ready":272.126062071},"url":"https://embed-ssl.wistia.com/deliveries/25ec231f1b5558e1633f0ccb43b27948a18f32bd.bin","created_at":1792517019,"segment_duration":3,"opt_vbitrate":3750},{"is_enhanced":false,"is_composite_layer":false,"type":"hd_mp4_video","slug":"mp4_h264_743k","display_name":"1080p","details":{},"container":"mp4","codec":"h264","width":1920,"height":1080,"ext":"mp4","size":461702181,"bitrate":743,"public":true,"status":2,"progress":1.0,"metadata":{"max_bitrate":2796064,"average_bitrate":761478,"early_max_bitrate":428632,"av_stream_metadata":"{\"FastStart\":true,\"Audio\":{\"channels\":2,\"channelLayout\":\"stereo\",\"Codec ID\":\"mp4a-40-2\",\"codec\":\"aac\",\"codecType\":\"audio\",\"duration\":4850.588231,\"sampleFormat\":\"fltp\",\"sampleRate\":44100},\"Video\":{\"avgFrameRate\":\"30/1\",\"bitRate\":\"626941\",\"Codec ID\":\"avc1\",\"codec\":\"h264\",\"colorPrimaries\":\"bt709\",\"colorSpace\":\"bt709\",\"colorTransfer\":\"bt709\",\"duration\":4850.533333,\"height\":1080,\"pixelFormat\":\"yuv420p\",\"rFrameRate\":\"30/1\",\"width\":1920}}","built_by":"render-pipeline","time_to_ready":115.598652324},"url":"https://embed-ssl.wistia.com/deliveries/973c3259d5f92d51265db219eae6679317930238.bin","created_at":1792517019,"segment_duration":3,"opt_vbitrate":5625},{"is_enhanced":false,"is_composite_layer":false,"type":"storyboard","slug":"storyboard_2000x2240","display_name":"Storyboard","details":{},"width":2000,"height":2240,"ext":"jpg","size":798890,"bitrate":0,"public":true,"status":2,"progress":1.0,"metadata":{"frame_count":200,"aspect_ratio":1.7777778,"frame_height":112,"frame_width":200,"built_by":"render-pipeline","time_to_ready":53.611124995},"url":"https://embed-ssl.wistia.com/deliveries/fb61a709dcb1593702c3bee047f28a4013a5085c.bin","created_at":1792517019}],"branding":false,"createdAt":1790711335,"distilleryUrl":"https://distillery.wistia.com/x","duration":4850.59,"enableCustomerLogo":true,"firstEmbedForAccount":false,"firstShareForAccount":false,"availableTranscripts":[{"bcp47LanguageTag":"en","familyName":"English","familyNativeName":"English","hasCaptions":true,"hasVideoStream":false,"name":"English","nativeName":"English","wistiaLanguageCode":"eng","language":"eng","ietf_language_tag":"eng","alpha3_terminologic":"eng","alpha3_bibliographic":"eng","iso639_2_language_code":"en","english_name":"English","native_name":"English","right_to_left":false,"alpha3Bibliographic":"eng","alpha3Terminologic":"eng","ietfLanguageTag":"eng","iso6392LanguageCode":"en","genericName":"English","genericNativeName":"English"}],"hashedId":"lnwftig68j","mediaId":153374048,"mediaKey":"wistia-production_153374048","mediaType":"Video","name":"Advanced workshop: Branded videos with Avatar V, the Video Agent and AI Studio","preloadPreference":null,"progress":1.0,"protected":false,"projectId":10877572,"seoDescription":"a Sample live event video","showAbout":true,"status":2,"type":"Video","originalIsEligibleForDirectPlayback":true,"originalFileStillAsset":{"url":"https://embed-ssl.wistia.com/deliveries/d1c17d04071daee8b3ff7648a4f3507b.jpg?video_still_time=2425","width":1920,"height":1080,"type":"still_image","status":2,"public":true,"progress":1.0,"display_name":"Thumbnail image","details":{},"metadata":{}},"playableWithoutInstantHls":true,"stats":{"loadCount":128,"playCount":43,"uniqueLoadCount":102,"uniquePlayCount":39,"averageEngagement":0.135443},"trackingTransmitInterval":10,"liveStreamEventDetails":{"id":416233},"integrations":{},"tagManager":true,"captions":[{"language":"eng","text":"Hello. Hello, everyone.\n\nDo you hear me well? Do you see me?\n\nGood morning. Good afternoon. Good evening.\n\nI I think we have we have people from all over the world joining today.\n\nNice.\n\nYou can drop in the chat where are you joining from.\n\nI will just wait, like, one one, two minutes so just more people can join.\n\nAnd we are going to start, like, super happy, super happy to see people from all over the world. Like, you for joining and spending your morning, afternoon, evening with me today.\n\nToday, we're going to talk about how to create branded videos with Hagen.\n\nSo you can see pinned in the chat, my colleagues would pin the presentation for today.\n\nI also want to mention that this session is going to be recorded and you're going to receive the presentation and the recording of the session after the workshop. So don't worry, if you cannot follow-up right now, or if you need to jump off earlier. We are going to send everything as I would show a lot of things today, and I think maybe not everyone is prepared to follow-up. So just don't worry.\n\nDon't worry if you cannot follow-up live. You don't need to. We're going to send everything to everyone so you can do this at peace, at home, at at your own timing.\n\nJust wanna mention something what we're going to need for today. So we are going to need Kagen and ChatGPT or CLOT or Gemini, whatever you are using as your preferred model. I'm going to use today the ChatGPT free version, as I think this is the most popular one. And we're going to talk about a little bit how to use Charge GPT cloth or Gemini or any other model to help you to create better looks, better videos.\n\nI think if everyone is ready, you can give me a thumbs up, and I can start by walk you through very quickly of what we're going to cover, show you some examples, and we can start. I'm going to stop between the slides to check for any questions. My colleagues are here also, so they would answer questions in the chat.\n\nSo, yeah. Okay.\n\nLet's get started. So as the presentation said, we're going to learn today, I'm going to show you how to use lookpacks and how to create branded videos with Heijen. I'm just going to play two video examples, and by the end of this workshop, you would be able to create such videos yourself.\n\nSo here is one of the example we're going to we're going to show...\n\nI'm going to show you two different workflows. One will be of creating branded avatar videos with a product, and other will be creating branded avatar videos for your service as I think these are like two separate workflows. So here is a video you would be able to get by end of this workshop. So this is my Europa Coco video example.\n\nHere it is. I used to run on coffee cup after cup, but the jitters and the afternoon crash finally got to me. So I stopped drinking coffee and switched to Europa Coco. It's made with rich real cocoa, and it's naturally caffeine free.\n\nThat means no jitters and no crash. Now every morning starts with a warm cup of cocoa instead, and I feel energized all day. If you're rethinking your coffee habit, give Europa Cocoa a try.\n\nSo this is one example. This is an avatar with product example. And the second example, it's about service video. So let me show you.\n\nAnd this example would actually explain the Hagen look packs that we are going to cover today, so this is my second example.\n\nThis is look packs. It lets you give a single avatar a whole range of different looks. To start, you open your avatar and select a look pack. Then you pick the looks you want, and each one is generated on your existing avatar. And because it builds on your real avatar, your identity stays consistent across every look with no facial drift.\n\nYou can create studio looks, casual looks, and seasonal looks all from one avatar with up to 500 looks each. So instead of rerecording for every scenario, you have a full wardrobe ready whenever you need it.\n\nThat's look packs. Start building your custom wardrobe today.\n\nNow I want to I want to say something, like, both of these videos both of these videos were created with just a prompt.\n\nThis is...\n\nLet me just show you an example. So you can see here all the info I said, this is my prompt, and this was the result with my end end video. So I'm going to explain you today how to create videos like this.\n\nWe're going to cover... Let me just go really quickly. So we're going to cover like, first, how to create avatar looks, generate different looks with your avatar, how to better... How to get better looks results as we do have, like, sometimes, like, people do not get their best results. So I'm going to show you how to get much better results. You can see the difference here, before and after after this workshop.\n\nThen we're going to... I'm going to show you like few examples of how you can how you can prompt for looks with your logo or with your product.\n\nAnd then we're going... I'm going to explain a little bit more of how to edit these looks, as we do have like many options of how to create better images of yourself with the Insight agent.\n\nAnd then we're going... I'm going to show you very quickly of how you can use Chargebee with your quote to create prompts for you that you can use in agent video agent and create and create a finished video. We're gonna cover also a little bit basic editing in the AI Studio and how you can do small adjustments to get to get better results better results with your video.\n\nSo I think let's get started. As I say, you don't need to you don't need to worry about following through all of... Through through the entire presentation. We are going to share all the all the research all the resources.\n\nSo, how to get started? Now, I'm just gonna go to my agent account, and I'm going to show you I'm going to show you how to get started with loops.\n\nNow, you have your avatar or avatar clone already done, but maybe you are not happy with the results you are getting as images.\n\nI want to show you before training my look back. So this was... I was trying to prompt for a look, but I didn't train my look back model. So this is what I was getting as a result. You can see the difference here. Like it doesn't it doesn't catch up my eyes correctly. It's different color. It doesn't look a lot like me.\n\nThis is happening because to get better looks, it all... The only thing that depends to get better looks from it depends from your footage and the footage you upload to Heijen.\n\nNow, I'm going to show you how to create this from scratch.\n\nNow, you have your avatar ready, let's say you have your video recording, and you uploaded a few photos of yourself.\n\nNow how to generate looks?\n\nWe have an option here called ad looks, and you can click on this option. And then of course, you can upload more looks of yourself, you can upload video footage, photos of yourself, whatever you you like, whatever you think you would like as an avatar result.\n\nNow, I'm going to cover today a little bit more of how to design new looks with AI. So, this is in case you want to change your background, or you want to change your clothing, you want to change the setting of your look. So, in this case, we're going to choose the option Design with AI.\n\nNow, here you're going to see this option where you can prompt for a new look. So, I can write some prompt here, like I need a professional look or professional settings. So, very important when we generate looks of ourselves is the footage we give as the so called reference footage. So, as better these images, as better results you will get.\n\nNow if you don't like your video, in this case, you can see like, I don't like... I don't think this is a good reference. Like my face, it's it's mid motion. The image is a little bit blurred, so I wouldn't use this as a reference to create a look. You can click here on these dots, and you can either upload the look. So I prepare some photos of myself, where I can I have like a few images of myself, and I can upload some different image to create a better look? So I'm going to choose this selfie of mine, and I'm going to upload it as a look reference.\n\nNow, very important, especially for people that are maybe changing like their hairstyle, or because I have different hairstyles, sometimes I'm with the hat, sometimes I have my hair down, the look will be as close to your reference image. So, when you want to generate a new look with Hagen, I would strongly recommend choose a look like hairstyle or type of image that it looks at it's close to the look you like to create. So now I can choose this look, and I can be like, create an image of me in office, sitting in an office.\n\nAnd I can just click generate.\n\nNow, it will take a little bit of time and Heijen will generate my images.\n\nNow, what I want to show very important to get better results from your look, especially if you just upload a selfie or if you just upload like an image that is not with a good angle or something, you probably won't get the best look results. Now we do have option for that called improve likeness. This this allows you to add more images and to learn and to teach HeiGen a little bit more about how you look. Now to get better results, what I would recommend is that you can click here on improve likeness with personal model. And here you have this whole option where you can upload up to 30, 40, 50 images of yourself.\n\nTo get better results when you're prompting for looks, I would strongly recommend this option. Now what I'm gonna do here is I'm going to click on the folder, And here I've selected over 40 images of me in different settings. Now, why this is important? It's important when you train a look model that you have some images where your full body is visible, where your face is visible from different angles.\n\nThis would help for when you prompt for new looks, this would help the model to understand better how you look in different settings in different environments.\n\nAnd you get much, much better results than just prompting for a photo. So, what I'm going to do here, I'm just going to select all these images, and I'm going to upload them to Heigen.\n\nLet me just while waiting for the uploading, let me just check my previous images.\n\nSo, let me go back to my avatar, to Dora.\n\nAnd here you can see the results which I just generated a while ago with me sitting in an office. And this is exactly the thing I was talking about. It's very similar to the uploaded look I already have. So it's it's very similar.\n\nNow, if I want to have a look where I stand up, or I'm in a different setting, I, of course, I can prompt for that.\n\nBut I do want to get like much better results from this.\n\nNow let me go back to here. So you would see this like validating photo. This this will take a while. So this basically is checking if the images all look like you and if it matches your face. So what I'm going to do here, I'm going to select all the images I have uploaded of myself, and I'm going to click here on train.\n\nAnd this will train my model, and it will help me create better... Much better looks in future.\n\nDidn't pass content review, it can show you can you can get this message when you're uploading photos.\n\nMaybe the light is too bad or your face is not visible, or it looks much different than your normal face. So you may have some photos that are not approved.\n\nBut you should be able to get results. Now, this will take like ten to fifteen minutes.\n\nI already pre generated looks with my train model under my another avatar, so I can show you the results here.\n\nNow, these are looks I have created, you can see here I uploaded all my photos.\n\nAnd I have here I have generated some different looks with after I train my model. Now, let me go back to the presentation because I want to show you a little bit...\n\nA little... Some examples of how I did generated these looks and what prompts you can use to create better looks.\n\nNow, here in the presentation, we can go back to prompt your new look.\n\nAnd here you would see there is this clickable expand fields where I added some prompt examples.\n\nYou can use these prompt examples, copy paste them with your own product or with your own logo to generate looks. These are just different examples. You can see the prompts here and how the different prompts will give you different lightening, different style. As you can see, one of the looks is like this look is like more polished, more blurred.\n\nThis one is also much more polished, much more blurred, maybe for professional content, while you can create looks like that, that look more natural and have this more UGC style vibe of feeling.\n\nThese prompts you can, again, as I say, you can copy paste, you can reuse them with your own, you can reuse them with your own products, with your own looks.\n\nAnother example here, this is for how to incorporate your logo.\n\nSo here are another examples. As you can see here, this look, for example, much more polished. I add here the Heijen look back logo incorporated in my image.\n\nHere I add it on the background. This is again like a generated look. And now I'm going to very quickly show you how you can generate looks like that by yourself. So we'll get back to the avatar.\n\nSo you can see here, this is still training. So as I said, I'm going to use my previously trained model for that.\n\nAnd here I'm going to go to AdWorks Design with AI.\n\nYou can see here this little toggle. So, once your personal model have been trained, you can switch it on or off. I would totally recommend to keep this on all the time. You also have option here to choose aspect ratio. So depends if you want a portrait look or landscape look.\n\nAnd here you have something called reference.\n\nNow first, again, I would make sure this is not a good image of me. It's with glasses and stuff. I just have this as a reference image. So if I want to create a look with with glasses or something, I can use this reference image.\n\nBut in this case, I don't like this image. I don't like my hair here. So I'm going to switch to another photo of myself, and I'm going to use this photo of me because I like my hair, I like my clothing here. And here I can click on reference.\n\nSo here on reference, you have option to upload images. So here you can upload your product image or so if you have like a product image, if you have a logo, you can upload to the ad reference your product or logo and incorporate this product or logo inside your look. You can add multiple reference images. So, let's say you have product and a logo, or maybe especially for real estate, for example, you have an image of a house or of a room and you want to create an image of your avatar inside this room, you can also do that.\n\nNow I'm going to show you very quickly the two examples with product and with the logo. So I'm going to choose my product, so this is the Europa co op package, and here what I'm going to do is I'm just going to copy paste one of these prompts here.\n\nLet me just copy paste that.\n\nI'm going to add it here. I'm going to choose portrait as I want more UGC style of image, and I'm going to click generate.\n\nNow these prompts, these prompt examples, you can use it with your own product. So what you can do, for example, it's you can just make take a screenshot of this prompt. Right? So you can be like that.\n\nAnd you can go to chat GPT, and you can upload your own product, maybe your product is, I don't know, a glass or a bottle or some food packaging, whatever your physical product is, you can just upload image here of your physical product, and you can either copy the prompts from the sheet, you can just drop screenshots from the presentation if it's easier, and you can be like, can you create similar prompt for me, but for my product that is, I don't know, perfume?\n\nAnd you can upload image of your perfume also as a reference.\n\nAnd you can adjust this prompt. So I can be like, that is a perfume, and I want to be dressed in a red suit.\n\nSo you can use these prompts from the presentation and you can adjust them depends of your product or your service. You can simply go to to ChatGPT to cloud, add a little bit of information of your product and service, drop this prompt, and ask to rewrite it for your use case. This is this is the easiest way because you kind of can see already the result you're gonna get. So if you prefer like more natural look, you can just drop like this prompt as an example. Or if you want more professional look, you can just drop this prompt. You can adjust even like here the prompt says both hands are raised.\n\nYou can just again, like tell Charjipiti, don't want the hands to be raised. I want them come like down on the table or whatever you prefer. You can just copy this prompt, explain how you want them to be changed. And it will create and generate a new prompt for you that you can after this reuse in HiGen design look back section to to generate that. Now, same with the logos, I do added few examples with logos.\n\nOf course, you can incorporate logos in different ways, you can incorporate logos in the background, you can incorporate logos on your clothing, you can incorporate logos on product, You can have maybe a cap or something.\n\nSo let me just copy this prompt, for example.\n\nAnd what I'm going to do here, I would remove the product, and instead of that, I'm going to add my logo here and add the prompt.\n\nHere we go. Okay. Let's make it... Okay. Let's make this one landscape, and you would be able to see the results. Let me just show you what we have so far. So here is what I have generated from my prompt example here.\n\nYou can see it's pretty, pretty... It's pretty similar.\n\nSo these prompts, as I say, they're they're reusable. You can you can change them in a way you want and in the way you like.\n\nI'm I'm going to see the results a little bit with the logos also. I'm just going to check some questions before that.\n\nLet me leave this open here for you so you can can check maybe some of the prompts.\n\nNo, you don't need to use GPT or Clot to create mock up images for you. All images in these examples, they were created inside Hadron.\n\nOf course, of course, you can use ChatGPT to create some looks of yourself. I would recommend strongly when you train a look, or when you add looks to Heigen, do not use AI generated looks.\n\nYou can collect what I did, for example, I don't like to have to do photos of myself. So what I did, I checked all my photos on my phone from the past two years, and I choose different photos of me in different settings like standing up, sitting down, selfie style, more professional style. So I would definitely recommend when you upload looks to Hedgehog, do not use AI generated looks because after this, you would get more AI looking results if you generate a new look inside Hagen. So I would really strongly recommend, even if you don't like yourself on the photos, I don't like myself on most of the photos I've added to train my avatar.\n\nYou can see here... Let me just show you some of the examples I've used to train that.\n\nLet me just show you my avatar.\n\nSo, you can see I just added like random photo selfies of me.\n\nWhat to make sure also, add some photos from different angles. So, have like left angle, right angle, because our profiles usually look very different from left and right side. And then if we want to create looks from different from different sides, we'll get much better results if we add reference photos from different angles, again, like standing up, sitting down. You can see nothing special in the looks I have added here, just just bunch of random photos. The only thing I would recommend, it's really make sure to have variety of photos. Also, when you create more UGC looks, choose an image with a lower quality as a reference image and when you want more professional look, you can choose more photo with clear quality, as this would help like a more natural low quality image will help the AI model to create more natural look of you also.\n\nNow, let me see my results here. So, my avatar look is ready. My model is trained, and here is the look packs, as you can see from the prompt I just added.\n\nThis is probably not the best one. I don't like this example, but you would you would get prob... You would get... When you generate looks inside hedge, and you get, like, two two options at least to choose from.\n\nSo I would say I would say this option is much better. And I'm going to use this. I'm going to use this in a video.\n\nNow, edit and refine your look. I'm going to go really quickly to this option. Let me just add my charger for a second.\n\nSorry about that.\n\nOkay.\n\nNow, again, with looks, doesn't matter, let's say, if you you added a photo of yourself or if you generate a look inside HeyGen.\n\nMaybe you don't like this look that much. Maybe you want to adjust something.\n\nEven if you add photo of yourself, maybe you don't want to generate looks. Maybe you just want to use your own photo that you already created.\n\nNow, we do have these options, and I want to show you very quickly about these options that we have here.\n\nSo, let's say I have this look, right? And I love it, I like it, but the light is not good, or I don't think it looks well enough. So, we have a few different options here. So when you go into your avatar, doesn't matter again if it's uploaded look or generated look, you have this little pencil icon here that you can click and you get this option called edit looks.\n\nAnd here you have few different options. Now you can click on appearance, for example, and here you have like preset filters, very similar. We all use like Instagram, Snapchat, all that. So you do have like here these preset filters you can choose to change your look.\n\nHere you can choose intensity. As you can see here, you see the difference. It will give me like, it will enhance a little bit the photo, it will give me like more polished look.\n\nIf you don't like any of these pre setups, we also have advanced settings. So, you have this option here, you can smooth your skin or you can make it less smooth. You cannot make up different levels of makeup. You can remove blemishes, you can slim your face, you can fix the exposure, you cannot remove some shadows, you can fix the contrast of your image. So, as I say again, if you generate a look and maybe you don't like something, you have all these options here that will allow you to adjust these looks and to get it in a way that you like the most.\n\nDepends on lightning. Let's say this is this is way too much makeup for my face. I don't like that.\n\nBut I would like a little bit more of a brightness on this photo. So I would would add like brightness and maybe a little bit more of saturation. So, as I say, when you generate an image, you can play with these options here and save as a new look, and you can make your look in a way that you like.\n\nAnother option I wanted to show, which is called camera angles. So you can click here, and you can generate different camera angles of your avatar. Let me just show you an example right here. So let's say I have this image, and I would go here, and I want to generate three different angles.\n\nSo I have my avatar speaking the same...\n\nAt the same look, the same background, but I want, like, different angles to make my video look a little bit better. We do have...\n\nThis is already preset. If you don't want to prompt, you can just click here, and it will generate three different angles of you. Here is here is an example I want to show you, till we are waiting for the generation. So, this is how generate angles would look like.\n\nYou would get like few images that include you in like left profile, right profile, so you can get like this variety of angles.\n\nAgain, you have here also custom edit, so you can can edit via prompt, You can edit via prompt, you can add maybe some reference images. This is a fast way to edit looks. And we have also the remixing to scene option, which I just show you. And basically, can adhere the different types of reference images.\n\nYou can add clothing, for example, let's say, if I want to change my clothes, I can just... Let me choose a look.\n\nOkay. Let me choose this look. I want to change my clothing, and I want this clothing to be consistent across scenes. So I would just choose a reference image of the clothing I want, and I would be like, regenerate this image with me standing standing in the office wearing the pink pants and blue sweater.\n\nAnd I can...\n\nEspecially for the clothing, I want to say if you if you want consistency across scenes, across looks, instead of prompting for regenerate me with a white shirt or regenerate me with a with a red costume or something, I would strongly, strongly recommend get a reference image of the clothing you want and prompt with this image as a reference. So you would get this way you would get like consistent results, because if you just prompt for a white shirt, this white shirt would probably look different on each image you generate.\n\nNow let me check the questions again and we're going to move to the video agent where I'm going to show you how to generate...\n\nHow to create prompts for VideoAgent and how to generate finished video with your product service.\n\nFor my for my AI avatar, I did I just did... I just upload, like, fifty seconds of footage for the ones I already I already showed you.\n\nWe did... We do have a workshop for beginners when I'm talking about how to create an avatar and best practices to create avatar.\n\nSo we will send this also together after after this workshop. And you can see there I'm talking a lot about, like, the footage you add.\n\nYou don't need longer content, like longer video footage. What I would say it's better, it's you need variety of footage. So if you can upload maybe one video which is portrait mode, one video which is landscape mode, This will give you better results.\n\nA UGC look, we have a question, what is an UGC look? UGC look is the so called user generated content look, which is basically a look that is not that professional.\n\nYou can see, for example... Let me go back to the examples here. Now this would be a typical UGC style look where you have you have a little bit, you can see the lightning here. It's not that perfect. The the image itself look with a bit lower quality. It looks like a natural image that you're gonna snap in your kitchen without profession... Without having, like, professional lightning can set up. While this look, you can see the difference looks much more professional that the background is blurred.\n\nEverything look much more tidy. So the difference between UGC and the other looks would be basically if you want more natural looking result, not that professional one, you can use some of the UGC prompt. And if you want more professional look, you can use some of the other prompts for more professional avatars.\n\nNow for... I'm going to go... There is some questions for the audio. I'm going to I'm going to go through that a little bit later now when I'm going to show you when I'm going to show you about how to create videos with VideoAgent.\n\nNow let's go back to the next step.\n\nSo for everyone, I want to ask in the chat, how many how many of you have used VideoAgent? If you can give me a thumbs up if you if you have used VideoAgent inside Hagen.\n\nNice. Okay. Okay. Great. Great. Great. A lot of a lot of people a lot of people have been using the video agent.\n\nNice. Okay. I'm just going to do a small introduction to what video agent is for the people that haven't used it yet, and we we... We're just going to move... We're gonna move directly to the steps.\n\nNow what agent video agent is...\n\nAnd here is my photos, actually. Let me just check. Okay. You can see here the example photos with the clothing.\n\nAnd let's go back. Let's go to the video agent.\n\nNow what video agent is? Video agent, it's agent... It's like similar to Charge Jupiter to cloth, and it allows you to generate full videos with motion graphics, with b rolls, with just a prompt.\n\nNow here are the two examples I've created with Heigen as a video agent. So you can see here, these are the examples I show you in the beginning.\n\nThese videos, they are not they are not edited. They were just created with with the video agent directly.\n\nSo I'm going to show you now a few examples of how to create this video for your product or for your service using the video agent.\n\nNow how to access video agent? You can access two ways. I've added link to the video agent inside the presentation. You can click in there, or what you can do is you can go to apps and click on video agent and you're going to you're going to access the video agent app.\n\nNow, few few things about creating video with VideoAgent. The first thing, very important, what we what we need as an information first will be your avatar.\n\nNow, you can see here before creating a video, can click and choose the exact avatar you want.\n\nNow, this is an avatar I don't want to use in this case, so you can click here and you can choose the exact look you want to use. Now, in this case, I would like to use the Arupa Cocoa look, so I'm going to choose this look. So before creating video inside the video agent, make sure you choose the exact avatar look you want. Here you have also option of the voice, so you can hear the voice you choose.\n\nHi, I'm excited you're here. This is your voice clone pre... Okay, so I don't like this voice.\n\nIt selects a voice that I don't like. So what I would make sure is that I choose the Okay. Wait. Language.\n\nThis act...\n\nI want to use, so I'm going to click on my voice, and I can pick another voice. So you would probably have like few different voices inside. Hey, Jen. So I don't like to use this voice as is with some accent, and then I will get very, very bad video result.\n\nSo what I'm going to do here is I'm going to choose the voice I really like.\n\nAlso, important update, Heijen has launched Heijen Voice, which is at the moment the number one voice model.\n\nWe are very proud with this launch. So I would strongly recommend everyone, you can click here and choose like under voice language, And you can test and try Okay. Wait. Agent voice.\n\nSo actually changed how I edit. So you can select here different voice engines. You can also hear how the different voice engines are sounding. So again, before generating any video in video agent, make sure you choose the correct voice and you you go to the voice and language and you hear the different voices and you choose the the one you like you like the most and you Okay.\n\nWait. There's a voice. Okay. Wait. This act So I would say this is very important as one of the one of the biggest complaints I I heard from users is that they don't like the voice.\n\nNow this is this is a way for you to to select really the voice you like. If you don't like any of these voices, you hear...\n\nYou have here option rerecord audio where you can record your audio again in a in a quiet environment, or you can upload audio you already have and recreate this voice in a way you like, and you can hear it again and again till you like the results. So I would definitely recommend like using these options.\n\nNow, here we have another option, brand system. So if you have if you have your website with with specific brand colors or logos, you can upload your brand system here. We have a guide for that in the previous workshop, which we're going to send after this one of how to do that exactly.\n\nSo you can choose your brand system, so you make sure your videos are your videos are consistent across the different videos you are creating. And here you have something called brand glossary.\n\nBrand glossary is something you can create.\n\nFor example, if you don't like if you don't like the... Let's say you have a brand, but HeyGen doesn't pronounce it in a in a in a in a good way.\n\nYou have here this brand option where you can add a brand glossary, which will be basically you can add Excel sheet or CSV file where you have the exact pronunciations of your product or service, so you make sure that it's pronounced correctly across your videos.\n\nNow, let's get back to the presentation here.\n\nNow, again, I have have few templates here for for different for different videos, And this is how I'm going to show you now how you can how you can use this with Cloth or ChatGPT to create these prompts for you.\n\nNow, in this case, what I would recommend everyone, if you haven't done this already, before before asking Quote or ChatGPT to create prompts for you that you can copy paste into the video agent, I would give I would give this few things as as an information because to get the right video is very important, the information the information we give.\n\nSo what I would recommend before that... So before you start the chat, let's say, before you start, like, creating prompts for your video agent, I would add here all the information about my product or service. So if I have my website, if I have if I have marketing plan, basically, as much information for your product or service that you have, I would add it before starting starting the chat. This will give you much, much better results, and it would create much better prompts for you based of your product or service.\n\nAnd I would do this... Of of course, there is there is other ways you can you can create your custom GPTs. You can create skills. This is a a bit more advanced.\n\nBut for just... If if you are a beginner, if you just start... Started creating this type of videos, I would recommend have a folder on your desktop with information about your brand, who is your target audience, what's your brand messaging, and I will drop this information here before starting to do anything in terms of video creation.\n\nNow how to create prompts for for your product or service.\n\nNow here I added few few prompts.\n\nYou can again, like copy paste these prompts add to chat GPT or cloth and ask to create these prompts for you.\n\nAgain, this presentation will be sent here. I have a lot of prompt examples that you can use, you can adjust.\n\nSo, this case, I'm just going to use the basic prompt here, and I'm just going to add it here. And I would be like, help me create a video prompt for Hedge and Video Agent, or you can just copy this here.\n\nAnd, also, you can you can just paste this instruction.\n\nAnd here what I what I would do... Now I would change these variables in a way that... For the video I want. Now in this case, the length of my video, this is a prompt for two, three minute video. I would do it, like, thirty second video.\n\nWhat's my product or service? Here, I'm going to add my product or service, which will be Hagen look packs in this case.\n\nTagline or one line promise, what I want to say to my audience, this would be like, be everywhere without being everywhere.\n\nYour source, so again here, I prepare... I do prepare some source images. So I would tell I have based on... I have folder with screenshots and screen recordings.\n\nObjective. What's my objective? Like, what I want my target audience to learn?\n\nIt would be, like, upload a photo upload a photo, create a look, create avatar videos. Who is this for?\n\nVery important. I would say this is this is very, very important before you create any prompts to tell for who is this video. This is very important. It will... It would create the script in a way that it will much, much better your target audience. I would add, hey, Jen.\n\nBeginners.\n\nPre order knowledge.\n\nAgain, this is something we already... If you have your branch or knowledge, you can already add this to that.\n\nAnd then the language, I want this in English.\n\nAnd the tone of the video, I want to be clear, education style, demonstrative. And here I'm going to show you, so I do prepare...\n\nI do have prepared my assets. So what I'm going to do here, it's here I have a bunch of screenshots of the steps of how I want to to create agent looks in this case. And I'm gonna just like drop it to the chat also to have have some examples. So, me drop them like here.\n\nAnd I'm going to go like this.\n\nNow this prompt examples I gave, again, you don't need to use them exactly. They're they're just some examples of how a prompt for a video agent can can look like.\n\nWhat... Again, what I would strongly recommend when you when you create these prompts, you can you can use the voice mode. I love I love both GPT or Cloth voice mode. So instead of writing all this stuff here, I wouldn't even write this manually. I would just drop this prompt template in there. I'm going to turn the voice mode, and I'm going to be... Look, so this is my business.\n\nI will give some information. I'm trying to create a thirty second video about Hagen lookpacks.\n\nI want this video to be for Hedgehog beginners. I'm creating this prompt to add into Hedgehog Video Agents, so if you can please use this prompt template I gave you and help me create this video.\n\nIn my case, I don't want b roll footage, I just want avatar footage.\n\nSo if you can create a prompt with just avatar footage and some motion graphics, etc. So, I would totally recommend I would talk there as going to talk to someone which I hired to create videos for me.\n\nSo I love the voice mode. I would just copy paste any of these prompts.\n\nI would just use the voice mode. I would explain what I want, how I imagine the video, what exactly I want.\n\nAnd I think this is a much easier way than writing.\n\nI would imagine I have new... I hire a new video editor, and I'm trying to give them all the information I can give them to create the best video for me.\n\nNow here is a here is a prompt example I have created already. So what I'm going to do here is I'm going to copy this prompt, and I'm going to go back to the Hagen lookpacks. I'm going to choose the lookpack videos. I'm going to paste this prompt here, and here I'm going to upload my assets. As I say, I have prepared... You can see here I have a ton of screenshots and video recordings to create this video. So I'm going to upload all the assets I have into the video agent to use it to use it as a reference to include in my video.\n\nLet me show you a little bit the assets I have prepared. So let's say these are the assets I have prepared for the Avatar look video.\n\nSo you can see here I have some screen recordings, like desktop recordings. So if you if you have a special educate... If you want to do educational videos or how to guides, this is a great way. You can just, like, screen record your screen when you are doing, like, something. Or if you have, like, product or service, you can add videos of your product or service. I've added a bunch of screenshots for the... From the Hagen interface that I want to be included in the video. For the product video in Europa Cocoa, what I used here, it's...\n\nI have a bunch of b roll footage that it's a real generated footage.\n\nSo I'm adding all this footage into the agent, video agent to use as a reference for my video.\n\nNow this would take a little bit time to load. I'm going to show you the the ready examples as we don't wanna waste, like, more time, and I'm going to bring you faster to to the results. So let me go back to my projects here.\n\nSo you can see here, this is my video agent chat.\n\nAnd let me zoom here so you can see you can see the exact prompt I added. So as you can see here, I choose the avatar with the RoboCocoa.\n\nI choose the voice I like, and I upload it to the video agent all all the footage, all the bureau footage I already have prepared. You can see here the exact prompt, create personal story style video about switching from coffee to Europa Coco using the video plan below.\n\nYou can see here the exact prompt I have used.\n\nHere is... This is a scene by scene prompt. So here I described the scenes I want to have included.\n\nVery important where I need arrow arrow meaning when I need avatar on camera and in which scene I want bureau footage Dural footage meaning all my all my photo and video assets. If you don't have assets, the agent will automatically take some stock footage that you can after replace.\n\nNow, as I say, these prompts... All the prompts, they will be sent so you can you can see them you can see them exactly, you can incorporate in your video. Now this is the result I get as a first result.\n\nSo let me show you here. I used to run on coffee, cup after cup. But the jitters and the afternoon crash finally got to me. So I used real cocoa.\n\nAnd it's naturally caffeine free. Okay. So this is my first result, which I don't like. Why I don't like it? There is not enough avatar footage, too much of the B roll footage, the music kind of doesn't match what I want to do.\n\nSo I'm going to show you very quickly how you can edit this stuff to get to get a finished video which looks like... Which looks much more natural and it does look much more finished.\n\nNow here is the other example I have for the service video. Again, I'm going to show you right here.\n\nSo you can see here again, I added all my screenshots, all the all the footage I have in my phone... On on my desktop folder. Here is my prompt, and this is the first result I got.\n\nNow This is Lookpacks. It lets you give a single.\n\nNow this is the result without editing. As you can see here, you're gonna see all these motion graphics and that you get out Avatar your identity from the video agent.\n\nNow, in this case, I don't like this motion graphic because it it doesn't tell me anything. It's it's a little bit morphing with the background. It's not well visible. Now you don't need to keep writing the agent and try to fix that because this will burn your credits as this is going to regenerate probably the entire scene. If you just want to make small adjustments of these parts, you can just click on edit, and you can just write the edit you want for this particular scene. In this case, I want to remove I want to remove the motion graphic. Remove the motion graphics.\n\nAnd this would regenerate this particular scene for me without motion graphic.\n\nNow, again, this will take a little bit of time. I'm gonna go back to the coffee video and show you how how you can adjust a little bit.\n\nNow what I'm going to do is I'm going to open the AI Studio and adjust some of this of this stuff manually.\n\nNow, in this case, I think there is like too many scenes. So I'm going to I'm going to delete some of the scenes as I don't need them.\n\nSo, what I'm going to do is like, I would manually delete the scenes.\n\nAnd as I say, I want a little bit more of avatar footage. I want to compare like few different looks.\n\nWhat I'm going to do here... So let's say I'm... I want to replace this scene with an avatar scene. So I will just click on the scene right here, and I'm going to choose one of my avatar looks.\n\nAnd I want to have different looks in the same video, so I'm going to choose this look for this scene. So you can see here, this would create a video where it's going to mix different looks of me in different settings with with the B roll footage I already have.\n\nAgain, also here, for example, I want to replace this scene with another scene I have.\n\nYou have this option in the studio where you can go to media, my media, and here you can see all the uploaded footage you already have. So let's say I want to replace this scene with another, so I will just click here and replace the scene. So I make the video in a way I want to make it.\n\nAgain, I don't like the music. I wouldn't waste credits and time to ask agent to replace the music for me. This is much quicker. So I can just click here on this audio track, just delete this audio, and here you have option music. So I'm going to choose the music I like, and I'm going to add this new music to my video.\n\nAgain, I don't like this. I don't like to have, like, the same avatar images in all the scenes. So, again, I'm going to click here.\n\nI'm going to click on the scene, and I'm going to change this look with my podcast look. So in this way, I would have I would have the entire video with few different scenes, which will make it look more much more natural.\n\nAlso, I know often we don't like the voice. So you can hear how your voice sound.\n\nLet me play that.\n\nSo here you have, like, these different voice agents.\n\nNow I can choose... Again, I would recommend now the new agent voice a... Voice an... Engine. So I would choose the new voice agent and update my default settings to hear how how this sound. You can click here on each scene and listen and listen to your voice also. So just one second. I think it needs to refresh.\n\nI used to run on coffee. Cup after cup. I used to run on coffee. Okay.\n\nSo I think this voice, it's...\n\nFor example, this scene, it sounds too slow. So what I can do here, you have all these voice settings. You can adjust also scene by scene. So let's say for this type of video, would like...\n\nI would adjust the speed and I want to do the tone more, like, friendly instead of excited.\n\nYou also have, like, adaptive avatar look, so the voice will adapt automatically to your look. So I can hear all this all this I used to run on coffee cup after cup.\n\nOkay.\n\nLet's say this is better. You have you have also the direct voice where where you can direct the way you want your voice to sound. So, again, I would recommend, like, for small changes, if you if you if you just don't like a certain scene, you can go to the AI studio. You can replace manually all the scenes with the scenes you like. You can mix your avatar look. So you can generate videos that looks a little bit more natural, that sounds more natural, and you have a little bit more of a control over what what you're getting as a result.\n\nAs you can see here, so this is like my final result from my previous edit.\n\nI used to run on coffee, cup after cup. But the jitters and the afternoon crash finally got to me. So I stopped.\n\nSo you can see how from this long video with 12 bureau footage and with slow voice by a little adjustments in the AI studio, I was I managed to create a video I actually like.\n\nAlso, can turn on, turn off captions always before you download your video. So if you don't want captions, you can you can just turn them off by clicking here. You can also change your caption style before downloading your video.\n\nAnd, yeah, I think I think this is it for today.\n\nLet me just check let me just check some questions, some answers. You can scan here, as I say, we are going to we are going to send this also are going to send this also to everyone who attended the workshop.\n\nSo you can access all these resources, you can access the prompts.\n\nWe have all the recordings live. We have live recordings of this workshop at... In our YouTube, so you can check this.\n\nMy colleagues would refer to that. I will just check now some questions I maybe didn't manage to answer.\n\nLike, yeah, tips to... Tips on how to best edit and minimize the credits used, I would say what I just show in the studio, it will save you a lot of credits because I see a lot of users when they when they get when they do video with the video agent and when they just want to do some change, they would keep... Like like, people would usually keep writing here, like, change this scene, change that scene, etcetera.\n\nYou don't need to you don't need to you don't need to change you don't need to change to write here to regenerate your whole video and boost credits. Again, as I say, you can you can just edit the scene by scene here if there is, like, small edits. As you can see here, I asked for the motion graphics to be removed, and they were removed just by prompt here.\n\nIf you need, like, to change, like, B roll footage or something, I would strongly recommend you can open the AI Studio and do this manually and make sure that all the bureau footage, the voice sounds correct, and everything sounds good before you generate. And this this will save you a lot of credits because running running every change through regenerating with video agent, this will waste much more credits than if, for example, you want a small change and you go in AI Studio and do this manually. Again, it's a little bit of extra work. Maybe it will take you maybe ten or twenty minutes extra.\n\nBut with AI Studio, you can for sure as everything is still manually, you can ensure that you get the scenes in a way you want your scenes to be made.\n\nAnd you can hear the voices before generating. You can adjust as many times how the voice sound, and this this will save a lot of credits for generation. Again, like, the video agent is getting better and better, but this would be my recommendation. If you want to save save credits, try try AI Studio, try to maybe adjust some things manually.\n\nLet me go back to the chat to see if there is some more questions.\n\nNow to...\n\nIf you if you if you just want to create... The question is, like, how how we can create an avatar that work just with the script.\n\nIf you just want a scripted video with avatar, you don't even need to go... You you don't need to go to the video agent.\n\nYou you have this option here. So the best avatars, it will be from your video recording. That's for sure. So if you want the best possible avatar look, video video reference of your avatar, I would recommend use your video recorded footage. They look they look more natural. If you just wanna run a simple script video, you have this option here, a Darro. So this is my recorded footage.\n\nHey, Theodora Kirova. Your digital twin is So I can just click Darro here and just choose the quick create option, and I can just create an avatar.\n\nI can just add my script here and just create avatar scripted video without motion graphics, without bureaus, etc.\n\nNow, the infographics sometimes the graphics get into avatar phase.\n\nThis is this is a little bit of an issue. We are working on improving video agent, and we have less results with motion graphics covering the face.\n\nAgain, if the motion graphic cover the face, let me show you here really quick.\n\nSo you can either prompt for it. So you can you can prompt here under edit and adjust the scene where the motion graphic is covering the face.\n\nOr if you don't want to prompt about it, you can open AI Studio.\n\nAnd you can very quickly move these graphics around. So let's say I don't like I I don't like this graphic to be here if the graphic is over my face. In the AI Studio, I can just take the graphic and I can place it where I want to place it. So if you want to keep the graphic, but it's covering the face or it's somewhere that you don't like it inside your video, you can just go to the studio, click on the graphic. You can even like change the colors of the graphic, for example, if you want. You have different options here, and you can move it around.\n\nThis would be the best way of how to how to control that.\n\nHow big of a library of bureaus? I think I think I think in Hagen, you can you can provide...\n\nI think it's up to fifty fifty files for bureaus. I'm not actually sure. I need to check what's the limit for adding bureau footage to the agent. But I would recommend...\n\nI mean, of course, maybe you need to... Maybe you need you need a video with 50 or 60 bureaus.\n\nYou can add that. This may get like a little confused if it's like a lot of assets.\n\nI would recommend just add the bureau footage or assets that you want to be used...\n\nThat want to be used in the... Inside the video.\n\nNow... Yeah. With the video agent, how to prompt better for the video agent? I I gave I gave this example prompt. You can try to use my prompt templates. Let me know how the results are.\n\nAgain, it's still it's still video agent. It's still an AI. Sometimes even if we give if we give, like, the perfect instructions, it won't work perfect. I'm not going to lie to you. Even if you give, like, the most perfect instruction, sometimes it it won't get it.\n\nI mean, it's getting better. I can see I can see it's improving a lot week by week. The team is working **** ** that. This is why I spent a little bit more time of how how you can edit results that you don't like. And I would I would really recommend if if some... If if it doesn't work like this, you can just edit the edit the scene that you don't like. So you can prompt for extra motion graphic or you can prompt to remove the motion graphics or you can prompt to replace the scene.\n\nBut again, if you want like to save the most out of your credits, you can just go to the AI studio and you can very, very quickly manually edit, replace the scenes that you don't like and get a very, very good good result after the video agent. I would use the video agent. I will set it as something like as a start point for my video.\n\nOf course, I sometimes get, like, perfect results with video agents. Sometimes I don't. So I would look at the video agent as a starting point of my video, and then with a little bit, maybe like five, ten minutes of extra edit, I can I can get I can get to the results I I actually I actually want? Again, in the AI Studio, we have different image models, different video models.\n\nWe are going to run another workshop in November, probably, where I'm going to spend more time of how you can generate AI cinematic footage. This meaning your avatar walking with the product or using the product or or use using your product in a scene.\n\nSo we are going to have a separate workshop on that, specifically, as I know we have a lot of users, they have different services and products, and generating good AI video footage with their avatar doing action. It takes some time.\n\nIt needs some knowledge for the right models to use.\n\nSo, we do have a lot of feedback for this type of webinar. So, we're going to cover this probably in November.\n\nYes. So, the question is like in portrait mode, if you can just replace the bureau footage, but keep the but keep the graphics, keep the motion graphics. Again, yes, this is this is this is completely possible. You can just go to the AI studio here.\n\nSo let's say I want to...\n\nI like to keep...\n\nSo these are my motion graphics. Right? And I do like my motion graphics, but I don't like the background footage in the end. I can just click on the background footage, go on media, my media, and choose another footage that I like.\n\nLet's say I can replace it with with this one. Sorry.\n\nOne second. So, yeah, here I can replace that with the footage I like.\n\nNow, yeah, I think I think this is it for today.\n\nI know I know we do have we do have a lot of questions, A lot more questions. I'm going... We're going to check after after the webinar. I'm going to check... We are going to send a feedback form. So please, please cover... Please cover... If you have five minutes to fill up the feedback form so we can cover in next webinars more of your questions or more specific things you need.\n\nI would be really I would be really happy to hear your feedback. We build these webinars now based on feedback from each webinar, so your opinion is really, really important for us.\n\nAnd yeah, just want just want to to say thank you to everyone for being here today.\n\nThank you for spending almost one and a half hour of your life with me.\n\nAnd as I say, we'll be super happy to hear your feedback. Give us suggestions of what you want to learn, what was easy for you, how we can improve for next time.\n\nThank you so much. Have a great day.\n\nHave a great evening. And, yeah, I hope to see you soon in next webinars. Thank you so much one more time for being with me today."}],"hls_enabled":true,"embed_options":{"opaqueControls":null,"playerColor":"#00C3FF","playerColorGradient":{"on":false,"colors":[["#00C3FF",0],["#66ffff",1]]},"plugin":{"captions-v1":{"on":"true"},"chapters":{"on":"true","chapterList":[{"id":"11977714-250e-4d29-9580-e24d36374ab3","time":"1014.21","ai_title":"Effective Prompting for New Looks","title":"Effective Prompting for New Looks","deleted":"false"},{"id":"178cba28-0b9d-45dc-aaa6-66087b9b49d9","time":"754.555","ai_title":"Improving Likeness With Personal Models","title":"Improving Likeness With Personal Models","deleted":"false"},{"id":"31e87e3e-f7ae-4be0-b70f-cb4e2832eb26","time":"1937.975","ai_title":"Generating Multiple Camera Angles","title":"Generating Multiple Camera Angles","deleted":"false"},{"id":"50359e5a-c7e8-4835-9adb-77e1a66ccbc2","time":"2754.56","ai_title":"AI Prompting Strategy for Video Agent","title":"AI Prompting Strategy for Video Agent","deleted":"false"},{"id":"66b36b81-9d29-4b9a-9f15-2df3c0b1434b","time":"4092.7","ai_title":"Credit Saving Editing Strategies","title":"Credit Saving Editing Strategies","deleted":"false"},{"id":"7f2cc2a8-a637-48b6-a214-96300b204169","time":"205.78","ai_title":"Workshop Goals and Video Examples","title":"Workshop Goals and Video Examples","deleted":"false"},{"id":"846f1571-f14c-4ae6-ad3f-1799cebb9690","time":"3945.44","ai_title":"Advanced AI Studio Manual Editing","title":"Advanced AI Studio Manual Editing","deleted":"false"},{"id":"86bddfc7-8af9-4111-b6b6-5b87171b17f9","time":"3423.655","ai_title":"Reviewing and Refining Video Results","title":"Reviewing and Refining Video Results","deleted":"false"},{"id":"ae13ce17-a7cb-4f16-93e4-495659e46e54","time":"491.07","ai_title":"Generating Custom Avatar Looks","title":"Generating Custom Avatar Looks","deleted":"false"},{"id":"b0b20805-40d4-4293-ab1e-a65f98c3ebb4","time":"1735.39","ai_title":"Editing and Refining Avatar Appearance","title":"Editing and Refining Avatar Appearance","deleted":"false"},{"id":"b0b5cbe3-8c04-471e-83b3-584f847df60a","time":"2384.17","ai_title":"Using Video Agent for Full Videos","title":"Using Video Agent for Full Videos","deleted":"false"},{"id":"c3a6fe8d-fb67-49ef-9f9f-7cca15d51040","time":"314.455","ai_title":"Understanding HeyGen Look Packs","title":"Understanding HeyGen Look Packs","deleted":"false"},{"id":"c45ea77c-2fe7-4cc8-bc28-f467d155b125","time":"4623.28","ai_title":"Upcoming AI Cinematic Workshop","title":"Upcoming AI Cinematic Workshop","deleted":"false"},{"id":"c6e31ff3-d5d0-4446-8ea5-ea1ccf638643","time":"74.885","ai_title":"Creating Branded Videos With HeyGen","title":"Creating Branded Videos With HeyGen","deleted":"false"},{"id":"cfd47d71-384a-49ab-b1dc-a1b9fb063ebf","time":"4755.735","ai_title":"Workshop Feedback and Closing","title":"Workshop Feedback and Closing","deleted":"false"},{"id":"d5b582f9-3be0-4d1f-b4ed-ac9cc33462b0","time":"4266.98","ai_title":"Creating Scripted Avatar Videos","title":"Creating Scripted Avatar Videos","deleted":"false"},{"id":"e59e99ca-94db-4954-9cb9-cbc1e22e0133","time":"3842.38","ai_title":"Voice Agent and Audio Settings","title":"Voice Agent and Audio Settings","deleted":"false"},{"id":"e6fe2a7b-2b87-472d-8fb7-168ec3d54081","time":"1204.465","ai_title":"Incorporating Products and Logos","title":"Incorporating Products and Logos","deleted":"false"},{"id":"fac96a48-3b28-43dd-b584-d2ccbc2e00c6","time":"4499.175","ai_title":"Effective Video Agent Prompting","title":"Effective Video Agent Prompting","deleted":"false"}]}},"volumeControl":"true","fullscreenButton":"true","controlsVisibleOnLoad":"true","bpbTime":"false","roundedPlayer":"8","vulcan":true,"stillUrl":"https://embed-ssl.wistia.com/deliveries/4684051b357fdd983ddd9f64e4965744.jpg?image_crop_resized=1280x720&image_quality=100","bodyFontFamily":"Inter","contrastIcons":true,"newRoundedIcons":true,"shouldShowCaptionsSettings":true},"embedOptions":{"opaqueControls":null,"playerColor":"#00C3FF","playerColorGradient":{"on":false,"colors":[["#00C3FF",0],["#66ffff",1]]},"plugin":{"captions-v1":{"on":"true"},"chapters":{"on":"true","chapterList":[{"id":"11977714-250e-4d29-9580-e24d36374ab3","time":"1014.21","ai_title":"Effective Prompting for New Looks","title":"Effective Prompting for New Looks","deleted":"false"},{"id":"178cba28-0b9d-45dc-aaa6-66087b9b49d9","time":"754.555","ai_title":"Improving Likeness With Personal Models","title":"Improving Likeness With Personal Models","deleted":"false"},{"id":"31e87e3e-f7ae-4be0-b70f-cb4e2832eb26","time":"1937.975","ai_title":"Generating Multiple Camera Angles","title":"Generating Multiple Camera Angles","deleted":"false"},{"id":"50359e5a-c7e8-4835-9adb-77e1a66ccbc2","time":"2754.56","ai_title":"AI Prompting Strategy for Video Agent","title":"AI Prompting Strategy for Video Agent","deleted":"false"},{"id":"66b36b81-9d29-4b9a-9f15-2df3c0b1434b","time":"4092.7","ai_title":"Credit Saving Editing Strategies","title":"Credit Saving Editing Strategies","deleted":"false"},{"id":"7f2cc2a8-a637-48b6-a214-96300b204169","time":"205.78","ai_title":"Workshop Goals and Video Examples","title":"Workshop Goals and Video Examples","deleted":"false"},{"id":"846f1571-f14c-4ae6-ad3f-1799cebb9690","time":"3945.44","ai_title":"Advanced AI Studio Manual Editing","title":"Advanced AI Studio Manual Editing","deleted":"false"},{"id":"86bddfc7-8af9-4111-b6b6-5b87171b17f9","time":"3423.655","ai_title":"Reviewing and Refining Video Results","title":"Reviewing and Refining Video Results","deleted":"false"},{"id":"ae13ce17-a7cb-4f16-93e4-495659e46e54","time":"491.07","ai_title":"Generating Custom Avatar Looks","title":"Generating Custom Avatar Looks","deleted":"false"},{"id":"b0b20805-40d4-4293-ab1e-a65f98c3ebb4","time":"1735.39","ai_title":"Editing and Refining Avatar Appearance","title":"Editing and Refining Avatar Appearance","deleted":"false"},{"id":"b0b5cbe3-8c04-471e-83b3-584f847df60a","time":"2384.17","ai_title":"Using Video Agent for Full Videos","title":"Using Video Agent for Full Videos","deleted":"false"},{"id":"c3a6fe8d-fb67-49ef-9f9f-7cca15d51040","time":"314.455","ai_title":"Understanding HeyGen Look Packs","title":"Understanding HeyGen Look Packs","deleted":"false"},{"id":"c45ea77c-2fe7-4cc8-bc28-f467d155b125","time":"4623.28","ai_title":"Upcoming AI Cinematic Workshop","title":"Upcoming AI Cinematic Workshop","deleted":"false"},{"id":"c6e31ff3-d5d0-4446-8ea5-ea1ccf638643","time":"74.885","ai_title":"Creating Branded Videos With HeyGen","title":"Creating Branded Videos With HeyGen","deleted":"false"},{"id":"cfd47d71-384a-49ab-b1dc-a1b9fb063ebf","time":"4755.735","ai_title":"Workshop Feedback and Closing","title":"Workshop Feedback and Closing","deleted":"false"},{"id":"d5b582f9-3be0-4d1f-b4ed-ac9cc33462b0","time":"4266.98","ai_title":"Creating Scripted Avatar Videos","title":"Creating Scripted Avatar Videos","deleted":"false"},{"id":"e59e99ca-94db-4954-9cb9-cbc1e22e0133","time":"3842.38","ai_title":"Voice Agent and Audio Settings","title":"Voice Agent and Audio Settings","deleted":"false"},{"id":"e6fe2a7b-2b87-472d-8fb7-168ec3d54081","time":"1204.465","ai_title":"Incorporating Products and Logos","title":"Incorporating Products and Logos","deleted":"false"},{"id":"fac96a48-3b28-43dd-b584-d2ccbc2e00c6","time":"4499.175","ai_title":"Effective Video Agent Prompting","title":"Effective Video Agent Prompting","deleted":"false"}]}},"volumeControl":"true","fullscreenButton":"true","controlsVisibleOnLoad":"true","bpbTime":"false","roundedPlayer":"8","vulcan":true,"stillUrl":"https://embed-ssl.wistia.com/deliveries/4684051b357fdd983ddd9f64e4965744.jpg?image_crop_resized=1280x720&image_quality=100","bodyFontFamily":"Inter","contrastIcons":true,"newRoundedIcons":true,"shouldShowCaptionsSettings":true}}}