AI :: Avatar :: Training :: The AI Model Method :: Course :: AI Videos
⪼ Made with 💜 by Polyglot.
Hey guys, welcome back! Now we're diving into one of the most exciting features - creating AI videos of your model. This opens up a whole new dimension of content that can dramatically increase engagement on social media platforms. Video content typically performs much better than static images, so this is a crucial skill for building your model's presence.
The video tool on Affogato is what's called an "image-to-video" tool, which means you can't just write a text prompt from scratch to create a video. You need to start with an existing image that becomes the first frame of your video. This is actually a good thing because it ensures consistency with your model's appearance and gives you complete control over the starting point.
Take one of the full-body images you created in the previous lesson and click "Convert to Video." This image becomes your foundation, and then you write a simple prompt describing what you want to happen after this initial frame. The AI will animate your static image based on your instructions.
Key Takeaways: • Image-to-video requirement: Must start with an existing image, can't generate from text alone • Consistency advantage: Starting image ensures your model looks exactly right • Foundation control: You choose the exact pose, outfit, and setting for your video • Animation focus: Prompt describes movement and actions, not appearance details
Writing prompts for videos is completely different from writing prompts for static images. The platform specifically recommends not over-specifying details and keeping prompts simple and straightforward. Complex terminology and lengthy descriptions actually hurt your results rather than help them.
Think of video prompts as describing simple actions and movements rather than detailed scenes. "She gently moves her body and smiles" works much better than a complex description with multiple specific movements and details. The AI needs clear, simple instructions to create believable, natural-looking movement.
Video Prompt Guidelines: • Keep it simple: Short, basic descriptions work best • Focus on movement: Describe actions, not appearance details • Avoid complexity: Don't use technical terminology or over-specify • Natural language: Write how you'd describe movement to a friend
For your first video, try something simple like "she gently moves her body and smiles." This creates subtle, natural movement that looks authentic when posted on social media. The key insight here is that basic, understated movements work much better than dramatic or over-the-top animations.
Dramatic movements and dancing often come across as inauthentic and obviously AI-generated. Subtle movements like gentle body sways, natural breathing motions, and authentic smiles create much more believable and engaging content. Remember, you want your audience to focus on your model, not be distracted by unrealistic movements.
Movement Best Practices: • Subtle is better: Gentle movements look more authentic than dramatic ones • Natural actions: Breathing, slight body movements, genuine smiles work well • Avoid over-animation: Dancing and dramatic gestures often look fake • Social media ready: Basic movements are perfect for platform posting
Once you've mastered basic model movement, you can add camera movement to create more dynamic and professional-looking videos. Add "camera moves around the subject" to your existing prompt to create the effect of someone filming your model in real-time.
This combination of model movement and camera movement creates much more engaging content that feels like authentic, professional videography. The camera movement makes the video feel more immersive and realistic, as if someone is actually there filming your model rather than just watching a simple animation.
Camera Movement Benefits: • Professional appearance: Looks like real videography rather than simple animation • Increased engagement: Moving cameras create more dynamic, interesting content • Realistic feel: Simulates someone actually filming your model • Combination power: Model movement + camera movement = compelling videos
Affogato also offers a talking video tool that can make your model appear to speak. After creating a basic video, hover over it and select "Create Talking Video." This opens up additional content possibilities but comes with important limitations and legal considerations.
You have two options: write a script for your model to "say" using template voices, or upload your own audio file. However, you cannot use someone else's voice without permission - everything must be original or legally obtained. The face swap feature shouldn't be used since you're already using your own model.
Legal and Ethical Requirements
This is absolutely critical: you cannot take someone else's video and swap your model's face onto it. This violates social media terms of service, breaks platform rules, and is potentially illegal. You must create all content yourself using your own model and original images. This is non-negotiable for legal and ethical reasons.
Talking Video Process: • Script option: Write what you want your model to say, choose template voice • Audio upload: Use your own recorded audio (with proper permissions) • Legal compliance: Never use others' content, voices, or videos without permission • Original content only: Create everything yourself to stay legal and ethical
The talking video feature produces results where the lip-syncing isn't perfect and the overall effect is less authentic than the original silent video. This is a limitation of Affogato's simple platform - the technology isn't advanced enough for Hollywood-quality results.
While the talking videos give you insight into how AI lip-syncing works, they're generally less convincing than silent videos with natural movement. For most social media purposes, the silent videos with good movement will perform better and look more authentic to your audience.
Realistic Expectations: • Lip-sync limitations: Not perfect, may look obviously artificial • Authenticity trade-off: Talking videos often look less natural than silent ones • Platform constraints: Simple tools mean limited quality capabilities • Strategic use: Consider when talking videos add value vs. distract from authenticity
Use video content strategically to enhance your model's social media presence. Silent videos with subtle movements often perform better than talking videos because they maintain authenticity while adding the engagement boost that video content provides.
Consider creating a mix of static images and simple movement videos rather than focusing heavily on talking videos. The goal is to build authentic connection with your audience, and sometimes less complex content achieves this better than more advanced features.
Video Strategy Tips: • Mix content types: Combine static images with movement videos • Prioritize authenticity: Simple, natural content often outperforms complex animations • Platform optimization: Different social platforms favor different video styles • Engagement focus: Choose video types that genuinely enhance audience connection
This basic video creation on Affogato gives you fundamental understanding of AI video generation, but remember that this is just the beginning. The simple platform provides decent results for learning and testing, but more advanced platforms will offer significantly better quality and more sophisticated features.
Use this experience to understand what types of video content work best for your niche and audience before investing time in more complex platforms. The principles you learn here - simple prompts, natural movements, authentic expressions - will apply to more advanced tools as well.
Key Takeaways: • Foundation learning: Basic platform teaches core video generation principles • Quality progression: More advanced tools will provide better results • Strategy development: Test content types and audience response before upgrading • Skill transfer: Techniques learned here apply to advanced platforms
Remember, while Affogato's video results aren't industry-leading, they're perfect for understanding the fundamentals and testing what types of video content resonate with your target audience!