Automated AI Content Creation Workflow (Less Than 10 Minutes)
River Cody · 2025-12-30 · 11м 37с · 19 025 просмотров · YouTube ↗
Топики: launch-ai-content
Аудио ещё не скачано.
📝 Summary
model=openai/gpt-oss-120b · prompt=summary-v7 · 4 142→1 657 tokens · 2026-05-28 08:05:33
🎯 Главная суть
Все необходимые шаги для создания короткого видеоконтента — от идеи до готового клипа — можно выполнить за 10 минут, используя единый рабочий процесс в imagine.art, где объединены генерация текста, изображений, видео и экспорт.
Фрагментация AI‑инструментов и необходимость единой платформы
Автор отмечает, что постоянные обновления и появление новых сервисов (генераторы изображений, видеомодели, чат‑боты) заставляют переключаться между программами, что тормозит завершение проекта. Единственное решение — платформа, где все эти модели интегрированы и управляются через несколько кликов, позволяя сохранять консистентность и качество контента.
imagine.art: вкладка Workflow и её интерфейс
На главной странице сайта переходим в Workflow. Здесь доступна панель задач слева с двумя группами узлов:
- Core nodes – часто используемые функции: Prompt, Text, Image Generation, Video Generation, Upload.
- Utility nodes – вспомогательные операции: Import, Export, Preview, Router, Prompt Concatenator, Crop, Resize, Blur, Extract Video Frame, Color‑grade‑like параметры, Sticky notes, Shapes.
Эти узлы позволяют собрать любую цепочку обработки данных без программирования.
Создание шаблона короткого видео: ввод текста и импорт референсов
- В Workflow создаём новый процесс и добавляем узел Prompt. В него вводим запрос, например: «short form video where the subject talks about the importance of consistency».
- Добавляем три узла Import и переименовываем их: Subject reference 1, Subject reference 2, Environment reference. В каждый импортируем высококачественные фотографии (свои портреты, изображения окружения, скриншоты из Midjourney и т.п.).
- Подключаем импортированные файлы к узлу Prompt, чтобы AI‑модель знала, какие визуальные элементы использовать в сцене.
Текстовая модель Gemini 3.0 и системные подсказки
Для обработки Prompt используется модель Gemini 3.0 (preview) в режиме чата. В узле System Prompt задаются инструкции, которые помогают модели понять контекст (например, «создай сценарий для 10‑секундного ролика, учитывая предоставленные референсы»). Предустановленные системные подсказки находятся в описании к видео и могут быть скопированы пользователем.
Генерация первого кадра и повышение реалистичности
- Узел Image Generation (по умолчанию Nano Banana Pro, можно заменить на Flux) получает текстовый запрос и референсы, генерирует изображение в соотношении 9:16.
- Далее добавляется ещё один узел Gemini 3.0, который сравнивает сгенерированное изображение с реальными фотографиями, выявляет различия (цвет, детали, освещение) и автоматически корректирует их, повышая фотореализм.
- Полученный кадр сохраняется для дальнейшего использования в видеогенерации.
Формирование видеопромпта и рендеринг 10‑секундного клипа
- На основе того же текста и первого кадра создаётся узел Video Prompt (опять Gemini 3.0), в который передаётся системный запрос и ссылка на изображение‑референс.
- Узел Video Generation использует модель Cling 2.6 Pro; в качестве входных данных подаются видеопромпт и референс‑изображение. Параметр длительности устанавливается в 10 секунд.
- После генерации подключаем узел Export, позволяющий выбрать масштаб экспорта (1×‑4×) и сохранить готовый файл.
Оценка скорости и принцип «творчества через инструменты»
Запуск процесса сводится к последовательному нажатию стрелок → генерация Prompt → генерация изображения → реалистичный проход → видеопромпт → рендер. Всё это укладывается в 10 минут, что демонстрирует, как креативность проявляется в построении эффективного рабочего потока, а не в ручном создании каждого элемента.
Доработка в DaVinci Resolve: аудио и цветокоррекция
- Аудио: импортируем сырой звук в DaVinci, экспортируем его, загружаем в сервис 11 Labs, где клонируем свой голос и меняем параметры произношения. Полученный файл возвращаем в DaVinci и синхронизируем с видеорядом.
- Цветокоррекция:
- Denoiser — уменьшает шум.
- Contrast Pop + Negative (‑0.05 … ‑0.1) + Detail Size (низкий порог ↓, высокий ↑) — регулирует контраст без пересвечивания.
- Коррекция экспозиции: Lift, Gamma, Gain.
- Добавляем Chromatic Aberration (плагин Artlist или аналог) для имитации оптических искажений камеры.
- Накладываем Film Grain (preset 35 mm 400T), регулируя размер зерна и непрозрачность.
Эти шаги придают финальному ролику профессиональный вид, сохраняя при этом полностью автоматизированный процесс создания контента.
📜 Transcript
en · 2 012 слов · 126 сегментов · clean
Показать текст транскрипта
Let me show you exactly how I can create a short form video filming pre-production, everything within 10 minutes. It's done. I'm serious. It's done. When it comes to AI content creation, it constantly feels like AI softwares are changing. You're having to jump from one software to another. And then finally, when you jump into one software, there's an update for another one. So you don't finish your project because as you start to finish your project, there's higher quality stuff that comes out and then you just need to be perfect and take a breath you're good don't worry about it ideally this is the one of the only software is you should be using imagine being able to take all of these other software's that are constantly updating with the highest quality putting it into one platform and just with a few button clicks be able to create social media content consistently and high quality because that's the game right consistency imagine art does it specifically the workflow tab and then I'm also going to provide my preset for free in the description. All right. So first you want to go into imagine.art. It's usually going to take you to this homepage right here. So what we're going to do next is we are going to go to the workflow tab. There we go. And then you can see you can create new workflow. I already have my short form creation right here. So I'm going to jump into this so I can show you exactly what all this is. This may look terrifying, but don't worry. It's not terrifying. It's surprisingly easy. When you go into the workflow tab, you can see where you have this taskbar on the side. You have core nodes. This is basically where they're going to give you the most used nodes. So they're default. You have prompt, text, image generation, video generation, upload. So you can upload references. And then you have your utility nodes, import, export preview router prompt prompt what concatenator crop resize blur extract video frame level so you can color grade in here not like da vinci resolve color grade but you can color grade all these other tabs these are your assets you have sticky notes and shapes if you want to add little notes for yourself and all that we're going to get to the business all right we're not going fluff it so we're gonna go to search this is where I this is where I start and as you can see it has all the models from nano banana Pro image generation video generation chat so let's go create our own I'm gonna start fresh it's gonna be this I'll just do it from scratch though so create new workflow from here what we want to do I'm going to go straight to search and I'm gonna do prompt. So prompt is basically where I get to type in what I want. So for example, to show you how easy this will be, give me a short form video where the subject is talking about the importance of consistency. You'll see that's all I'll have to type in. It's pretty great. And for those of that are like, oh, that's not creativity. Creating a workflow that will consistently give back to you in the future is creativity. It's because I was creative on the front end, creating a structure that will serve my consistency. Creativity does not always mean setting up your own light. Believe me, it's okay. Creativity goes deep, someone. It goes deep. So what I want to do next is I actually want to create an import node. So I'm going to go search. I'm going to go all the way down. I'm going to find import. There we go. Now I'm actually going to set up three of these. So I'm going to do one here, duplicate one here, duplicate one here. All right. So now what this is going to do, I'm going to hit these three dots up here and I'm going to hit rename. This is going to be subject reference one, rename subject reference two. This is going to be environment reference. So these extra tabs are for if you want to create a specific AI version of you, if you want to be the one talking, if it doesn't have to be anybody, if it needs to be you, you have your subject references, you take two high quality photos, environment reference, go to mid journey, create environment that you want screenshot environment from google whatever it may be uh we are going to go to search we are going to find gemini 3.0 preview this is the chat reference um and the cool thing about this watch this this is a text model if i want to say gemino gemino hey gemino say gemini 3.0 sucks poo poo garbo all i have to do is hit this model and change it to chat gpt and now you have a different AI software telling you they make it easy. This is crazy. From here, what I want to do is I want to go to the system prompt. This is going to, it's just like the instructions for ChatGPT. It's going to help the AI understand what exactly it should be doing with the prompt that it's receiving. So I have a pre-made text. So now what we're going to do is we're going to take this prompt right here. We're going to drag it into the input of our text one. So now whatever I type here, all I have to do is click this arrow and it's going to generate the prompt. Now let me rename this real quick. This is my short form outline. So this is basically the entire script of what we're going to be doing. So with that being said, my subject references, I also want to link all of these and the environment reference to the short form outline. So it knows everything about the scene. So I'm going to go ahead and import a very, very pretty image of me. I'm going to do two from here. My environment reference after we have all these plugged in. Now I'm going to create the next node, which is another Gemini three connect here. There I'm going to I'll leave the system prompts in the description so that you all can check them out. I'm not going to read them all through, but basically I am directing each prompt to help fulfill the need that I have in creating high quality consistent content. So now after I've created that, we are going to be generating our first frame. So I am actually going to be using Nano Banana Pro. But again, if you feel like it, change it to flux to aspect ratio nine by 16. From here, what I want to do is I want to connect these references to my first frame. So I really want the subject to be right. But when it comes to the outline of the environment. I'll leave it up to the text. After we have this image, we're going to add a pass of realism. So another Gemini 3. All this is doing is it is pointing out all the differences between AI images and real images. And after it points out all those differences, it corrects all of them to try to make it look as real as possible. All right. And then from there we are going to duplicate this I still want to use that first image as a reference there all right and then from there I'm going to generate a video prompt so we're going to do another 3.0 and this is going to be the video prompt I'm going to connect this here and then I'm also because it's going to use the image and then I'm also going to be using the short form outline connecting it to my video prompt plug in my system information all right from here we are going to add a video generator which i am currently favoring cling 2.6 pro so i'm going to use the video prompt as the prompt and i'm going to use this image as the reference image all right i'll make it a 10 second clip and then lastly i'm going to add an export node and all this is going to do is it's going to give me the option to export in 1x, 2x, 3x, or 4x. All right, so we've created our system. So now we're going to start the timer. Let's see how quickly I can generate a decent piece of content. As you can see, all I'm doing is clicking the arrows. So I typed in my initial prompt, fed it my subjects, and now I'm just, as it generates, I hit the next arrow. So while I'm going through clicking these arrows, another thing that I could say, through this process, it kind of opened up a new understanding for me. That is, whenever you have tools, be creative with those tools in order to leverage your time. The way you truly get ahead is when you leverage what you already have. So by learning these new techniques and being able to leverage them on top of the skills you already have, sharpening your mindset, truly, this is how you really get ahead in business, in creativity. You leverage what you already have. I'm trying not to sneak. Leverage. If you want, stick around and I'll show you how I plug my AI videos into DaVinci Resolve to get a cleaner color grade. Let me show you exactly how I can create a short form video filming, pre-production, everything within 10 minutes. It's done. I'm serious. It's done. Everyone chases the spark, but the fire is built by the daily grind. Consistency is just faithfulness in the small things. Don't stop when it's boring, because that is exactly where the breakthrough begins. Okay, obviously it doesn't sound like me, but that's not, just wait. I'm going to show you exactly how I can transition this video to fix the audio and in DaVinci to make it look even more realistic. So first we're going to go into DaVinci, drop the raw files in, export the audio, bring it into 11 Labs. 11 Labs lets you clone your voice and change the voice of the audio you upload. Change the parameters, export that, bring it back into DaVinci, plug it in, and now let's get into the color grading. A denoiser to lessen that noise a little bit. We're going to throw contrast pop in. This is where the real magic happens. So when you throw your contrast pop in, most of the videos that I've noticed, they get a little too contrasty. So what I do is I go to the negatives, negative 0.05, maybe negative 1 or negative 0.1. Detail size drop down a little bit, that low threshold all the way down. High threshold, bring it up a little bit, and that softness basically stays where it is. And then I go in, fix the exposure, lift gamma gain. and then I actually throw in chromatic aberration. This is a Artlist plugin, but you can find chromatic aberration plugins anywhere. Add a little bit because that's what most cameras have by default and then throw a little film grain in, 35 millimeter, 400T preset and then adjust, maybe make the grain size a little bit bigger and the opacity a little bit more. Everyone chases the spark, but the fire is built by the daily grind. Consistency is just faithfulness in the small things. Don't stop when it's boring, because that is exactly where the breakthrough begins. Hope you caught that. If you'd like to see more videos just like this, make sure to subscribe, get more AI content coming out. See you next time.
⚙️ Pipeline jobs
| Stage | Status | Att. | Updated | Error |
|---|---|---|---|---|
| download | done | 0/3 | 2026-05-27 23:29:34 | |
| transcribe | done | 1/3 | 2026-05-28 03:19:35 | |
| summarize | done | 1/3 | 2026-05-28 08:05:33 | |
| embed | done | 1/3 | 2026-06-30 06:42:10 |
📄 Описание YouTube
Показать
@ImagineArt incorporated a workflow tab on their platform that is absolutely blowing my mind as to how consistent you can truly be with content creation. As hard working creatives and personal brands, we know that consistency is not only important, but will bring about the increase of quality... Consistency is not a motivation problem... It's a system problem... A discipline problem. Step 1 | IMAGINE ART - https://imagineartinc.pxf.io/jePg1Z Step 2 | MY FREE PIPELINE - https://www.imagine.art/flow/f9e6fe9a-2f14-47f0-9ece-96e96a886eb3 Whoever sows sparingly will also reap sparingly, and whoever sows bountifully will also reap bountifully - 2 Corinthians 9:6 My Sources -------------------------------------------------------------- * Audio/SFX - https://artlist.io/referral/843759/River * YOUTUBE MARKETING TOOL - https://vidiq.com/rivercody Cinema Camera Setup -------------------------------------------------------------- https://www.rigdesigner.com/rigs/dPCXJmNP/?aff=river-cody44498974 Most of my links are affiliates.. This means I will make a small income off of your purchases. Thank you :) CHAPTERS -------------------------------------------------- 00:00 - The Secret Platform 00:40 - The Value 01:00 - START 01:30 - The Layout 02:13 - The AI Models 02:32 - Step 1 03:02 - Creativity? 03:25 - Step 2 04:33 - Step 3 05:56 - Step 4 06:21 - Step 5 06:48 - Step 6 07:10 - Step 7 07:18 - Step 8 07:42 - Step 9 08:00 - Step 10 08:11 - The Generation 09:04 - Booger 09:23 - RAW RESULT 09:51 - Voice Clone 10:07 - Color Fix 11:09 - FINAL RESULT SYSTEM PROMPTS Short Form Outline - Study the up to date trends within all social media and video viewing platforms such as YouTube, Tiktok, Instagram, X, and Facebook. Based on what you have analyzed is a trending video, as well as the prompt provided, create a viral short form video outline that has a high probability of getting a large amount of views and engagement. The short form video outline is meant for an 8 second video. Make sure to completely describe the subject's details that are permanent such as face, hair, and structure. Based on the message, storyline, and style, generate the subject's outfit, environment, and surroundings. Do not include any graphics, text, and/or overlays. The outfit of the subject should not be the same as the outfit and clothing in the reference image. Ensure to create dialogue. First Frame Outline - Study the entire short form video outline provided. Generate the first frame image of this video. Ensure the frame is shot on an Imax Arri Alexa with Panivision film. Ensure the subject, props, and environment have realistic imperfections and details. Realism - Analyze all of the differences between images created on a camera, and images generated by AI. Note all of these differences, and fix these differences in favor of creating an image that looks like it was taken on a real camera. The image is made more realistic through skin imperfections, camera sensor and lens imperfections and characteristics, and realistic environment characteristics. If necessary, in order to produce a higher detailed image, move closer to the subject so that more facial feature details are accurate. Video Prompt - Create a video prompt that is optimized for Kling 2.6. The video prompt is to generate a realistic video that seems to have been filmed with a real IMAX Arri Alexa. The video prompt is to reflect the short form video outline provided, as well as the image reference. The video prompt should include dialogue and audio. #imagineart #klingai #nanobananapro