If you have ever generated a great shot, felt excited, moved on to the next panel, and watched your character turn into a completely different person... this module was built for you. Module 3 is all about one word: Consistency. - Character consistency - Background consistency - Prop consistency - Wardrobe (clothes) consistency - Style consistency - Sound, Music, Voice Consistency All of it, shot by shot, the way a real director controls a production. If you are part of the AI Animation Masterclass... Access Module 3 here Here is what you are walking into: - You will build your Director's Reference System. Character reference sheets, prop reference sheets, wardrobe sheets. You will learn the exact prompts to generate them, and then use them as your source of truth across every panel you create. No more guessing. No more hoping the AI remembers what your character looked like three shots ago. - You will learn to work with 3D environments for background consistency. Using Marble, you will scout virtual locations the way a real filmmaker scouts a set. You will find your angles, plan your shots, and create a background that stays consistent across the entire scene. You will even learn how to add elements directly into a 3D space using Nano Banana, and how to connect two separate 3D sets so your story moves seamlessly between locations. - You will get a lesson that most tutorials completely skip. Your audience is smarter than you think. You do not need to show every single detail of the action. You do not need to animate a sword handoff, a transformation, or a complex object pass. You just need to show enough for the story to land. As a director, your job is to choose what people see, not to force the AI to do things it cannot do. This is one of the most freeing ideas in the whole course. - You will learn which platforms let you reference specific images. Freepik, Galaxy, Higgsfield, Dzine, OpenArt. You will know exactly which tools support the kind of image-locked consistency workflow this module teaches.

I got a Discord notice this morning that Marble had just dropped a new version, so to satisfy my curiosity I tried it from a great pano that previously produced a horrid world. I was very impressed. I get the feeling that they are now using a much finer mesh, though that's just my guess. I say that because complex objects with a lot of shape detail have gone from completely unusable to respectable (if not perfect). Since we now know that OpenArt's world creation uses Marble's engine, I have to wonder if they are using the new version. I still have some Marble credits to use before I switch to OpenArt for world creation.

Ok guys, this one I am really excited about. Live replays just got a serious upgrade available for VIP users in the AI Animation Masterclass. From now on, every live class gets uploaded into its own dedicated section inside the Classroom, with a proper title based on what we actually covered in that session. No more hunting around trying to figure out which replay is which! But here is the part I am most excited about. Each replay now comes with notes. I recently signed up for Rev.com, which uses AI to transcribe sessions and generate notes automatically. I have to be upfront with you: I will not be personally reviewing every set of notes before they go up. I will copy and paste what Rev produces. BUT, I reviewed the first one and honestly? I was impressed. The quality was good enough that I felt confident this is going to be genuinely useful for you. Here is what you get with every session: - Key Points so you can see the big takeaways at a glance - Notable Quotes with timestamps so you can jump straight to the good stuff - Detailed Insights with timestamps so you can dig deeper into the parts that matter most to you And here is a bonus I love → Because the notes live right below the video inside Skool, you can use the Skool search bar to find specific content across all the lessons. Looking for that thing I said about character consistency three sessions ago? Just search for it. Done. All of it is waiting for you right now in the Classroom. Check the screenshot below to see exactly where to find it! See you inside! 🎬

This is my final animation for which I used Luma, being the best at keeping the 2D characters and backgrounds, and not turning them into a 3D hybrid like other AI models, which eventually don't even look like my original characters. It also does parallax and fills the missing parts very well, and does not add weird and random stuff in the video. Unfortunately, Luma doesn't have lip-sync. So I used something else, not completely satisfied, but for now, it works. This is my first animation with this type of complexity, so there is room for learning and improving.

Inconsistency between generations.