Narek Ghazaryan’s Post

AI can generate remarkable images, sound and motion, but it cannot stay consistent, stable, or predictable across a sequence. Raksha Wave technology solves this problem by introducing a dedicated layer of creative control that works with any model or engine, allowing studios, directors, animators, VFX supervisors, artists and musicians to steer AI output with clarity and intention. Raksha Wave plugs into existing production pipelines without forcing changes to tools or workflows. It is fully model agnostic and compatible with current generative image and video systems used in AI assisted filmmaking. The underlying control logic acts as a universal AI cinematic control system that can direct the behavior of many different models. TSP sets the standard for professional hybrid content workflows. Raksha Wave gives AI systems what they normally lack: reliability, precision, and directorial control. Raksha Wave is a proprietary creative framework and AI pipeline developed by filmmaker Vladimir Raksha and Raksha World Ltd. It is used for concept development, visual direction, story design, experimental animation, and advanced AI driven cinematic workflows. Raksha Wave is not an AI that guesses. It is DI, Deterministic Intelligence. You decide. It executes. Traditional AI relies on probability and interpretation. Raksha Wave translates intent into a strict structured control language. The same intent always produces the same result. No drift. No improvisation. This is why Raksha Wave can command AI. It removes agency from the control layer and turns generative AI into a reliable execution engine. Human intent remains the source of will. Deterministic intelligence is not artificial thinking, It is intelligent control. This model agnostic approach allows the same Raksha Wave setup to guide different engines without rewriting prompts or templates. At the core of the system is TSP, the Temporal Syntax Prompting language, which describes motion, timing, scene structure, and emotional rhythm in a clear, time based format. TSP was designed as a cinematic language for directing AI and can be adapted to multiple backends and engines. Even on models that do not support native audio, the TSP core can drive convincing lip-sync behaviour by controlling facial articulation, mouth shapes and timing as if sound were present. This is achieved through its own internal timing and behaviour engine, rather than relying on built-in audio features of a given model. https://raksha.world/wave/

👋 AI Cinematic Lab - Weekly Selections (Week 4) Every week, I discover 5 creators who are making a difference in the world of AI storytelling. 🎬 The highlight is Jan-Willem Blom, I was immediately hooked by this short. I'm a sucker for old illustrative visuals, and I loved the fable quality here. But the real magic? It was co-created with his 9-year-old twins. That collaboration IS the special creative spark. 💡 The others to watch: Dale Williams: Let's flip the mood completely. What happens when you turn Christmas into a dark comedy thriller? The first shot is already hilarious and dark at the same time, and Dale walks that fine line perfectly until the end. Dave Clark: A mystery vertical short that proves GenAI allows us to explore new genres and interaction models. Dave masterfully provokes the viewer to solve the mystery through non-standard, immersive camera angles. Brilliant serialization. Kelly Boesch: Continuing the exploration of AI beyond linear narrative. This is a beautiful, almost pictorial example of AI used as a digital paintbrush. Enhanced by a coherent color palette, the motion and music turn this into a visual meditation. Maciej Bakowski: Walking the path of pure abstraction. This theme borrows from neomorphism and fluid simulation. If you're a UI creative or FX artist, you'll feel an immediate familiarity with these shapes. A beautiful technical exploration. ⚡ Lab Shoutouts: Big respect to the awesome work shared in the group this week: Dario Novesi, Sharizal Abd Halim, Brian Levin, 🤙 Thomas Wegerer, Alena Llopis, Robert Gaudette, Yousef K.. ------ I’m Rudy. 15+ years building worlds for Film & Games. Now transforming AI chaos into art and craft. Follow for insights and experiments in AI cinematography. ------ 👇 LINKS TO SELECTED FILMS IN THE FIRST COMMENT

Making an AI‑driven movie is a blend of traditional filmmaking and tech‑savvy tools. Here’s a streamlined workflow one can follow: 1. Concept & Script– Start with a solid story. AI can help brainstorm ideas, outline plots, or even draft dialogue, but the core concept still comes from you. 2. Script Refinement– Use language models to polish dialogue, check pacing, or generate alternate lines. Keep the human touch for emotional nuance. 3. Visual Planning– Sketch storyboards or use AI‑powered image generators (Midjourney, Stable Diffusion, DALL‑E) to visualize key scenes and set the visual tone. 4. Production - Actors & Location– If you need virtual actors, employ deep‑fake or generative‑actor tools (e.g., Synthesia, Runway). For real shoots, use AI‑assisted cameras for focus tracking and scene composition. - Audio– AI can synthesize voice‑overs, create background music, or even generate Foley sounds. 5. Post‑Production - Editing– AI video editors (Adobe Sensei, Runway, Pika) can auto‑cut footage, suggest cuts, and generate rough cuts quickly. - VFX & Compositing– Use AI for upscaling, background removal, and generative fill to speed up visual effects. - Color & Sound– AI tools can auto‑color‑grade and master audio, ensuring a polished final mix. 6. Distribution & Marketing– Deploy the film on streaming platforms or a dedicated OTT portal. AI can help tailor trailers, create personalized thumbnails, and target audiences with predictive analytics. 7. Feedback Loop– After release, gather viewer data (watch time, sentiment) and feed it back into the AI pipeline to refine future projects.

AI video creation is moving from “prompt-and-pray” to real control. A new AI workflow now allows creators to animate motion by manually moving objects or images within a scene, even using something as simple as a stick. Instead of guessing what motion should look like, creators first show the AI how something moves using a rough video or basic animation. The system then: • Tracks the demonstrated motion • Removes unwanted visual elements • Rebuilds the scene with accurate movement • Keeps the background clean and consistent • Uses a short text prompt to finalize the video This approach brings AI video closer to traditional animation workflows, where intent and precision matter. It’s a meaningful step toward AI tools that enhance creative control rather than replace it. #AI #GenerativeVideo #CreativeTechnology #VideoProduction #AIVideo #FutureOfMedia #DigitalCreation #TechInnovation #CreatorTools

Unlock Creativity with AI: From Text to Animation! Did you know that crafting the right prompt for AI can do more than just generate text? With advanced prompt engineering, you can now bring stories to life — literally! By carefully designing your instructions to Large Language Models (LLMs) and AI animation tools, you can create custom animated scenes, characters, and narratives without any prior animation experience. It’s all about knowing how to “talk” to the AI! Want to see it in action? Check out this short animation created by Emmanuel Abiola O. entirely through LLM-guided tools Whether you're a content creator, educator, or just curious about AI’s creative potential, learning prompt engineering opens up a whole new world of possibilities. Have you tried using AI for creative projects?

🔥 Hot take: Everyone says AI animation is “soulless,” but are we sure humanity is always present in the animation we celebrate? Behind the scenes, the human side of animation has often been anything but humane. Before attacking AI for “devaluing creativity,” let’s acknowledge what has existed for decades in traditional and digital animation: • Poverty wages behind billion dollar franchises • Outsourced labor with weak protections • Extreme crunch and burnout • Artists pushed out early in their careers This industry has produced beautiful art while relying on low pay, high pressure, and unhealthy working conditions. Here’s the uncomfortable truth: There is not always humanity in human made art. AI did not create this problem. It is simply further exposing it. If we truly value the human touch, we should value the humans who create it. AI is not pure, but neither are the alternative tools. Thoughts? 👇 PS... Happy Holidays to all PPS - Back with longer content in the new year (or next week?)

The inevitable rise of AI in Filmmaking: Simply put: AI is a revolution and the portal to a world of new possibilities for Cinema: One of the most realistic, immediate and technically feasible ways AI will reshape filmmaking is in reshaping stylistic characteristics, the palette and final rendu of a film...For example, AI can already get surprisingly close to Golden Age noir aesthetics, and within a few years it will be production-quality. Let’s break it down practically: 1 )The Look of Film Noir Is a Recipe — and AI Can Learn Recipes: Classic noir visual identity consists of very knowable ingredients: Cinematography: • Deep shadows, low-key lighting • Hard sources, sharp falloff • Venetian blind patterns / gobo lighting • High contrast ratios • Narrow depth of field  • 1940s/50s film stocks • Grain structure, highlights, etc. AI models can learn these patterns beautifully. 2 ) Modern digital cameras are not a barrier: RED, Alexa, Venice — all of them actually capture more information than needed to enable AI to emulate old film looks. Digital footage gives you: • Huge dynamic range • Clean shadows • Flexible color spaces, etc. This makes it ideal source material for AI-driven “film aging"... AI can add constraints and artifacts—what it cannot do is recover what isn’t there, so these amazing new digital cameras are not a negative factor to achieving "classic" esthetics... 3 ) What AI can do right now (2025, soon to be 2026...) AI tools today can: • Convert modern color grades into black-and-white with period-accurate contrast • Add authentic grain patterns • Redefine film response curves • Introduce lens simulations • Create lighting illusions • Generate fog/dust atmospherics, etc. 4 ) Where AI still struggles: Artificially reproducing: • The behavior of light through vintage lenses • The micro-contrast of older film stocks • The subtle movement imperfections of analog cameras • Grain that responds dynamically to exposure levels • Changing halation These aren’t trivial because they’re physical phenomena. But deep-learning tools are getting extremely good at simulating photochemical behavior. 5 ) What AI will be able to do in the next 2–5 years This is the realistic part: AI will soon offer: • Scene-to-scene stylization with continuity (a big limitation today is that grain/contrast often vary from shot-to-shot) • Camera-aware styling AI will know how to stylize differently for wide shots vs close-ups... • Lighting reconstruction • Lens fingerprint reconstruction • Old film-stock emulation down to the chemical level will be possible... This will allow a filmmaker to: • Shoot clean modern footage • Feed it into an AI film-noir transformation model • Get an output that looks indistinguishable from a 1949 RKO classic... And crucially this will be at one point controllable and consistent enough for full-length features... 6 ) Conclusion: AI will soon become a standard post-production tool, like a LUT on steroids...

"Prompt to VFX" is here 😮 IC-Effect has just been announced, a research paper from Communication University of China and NUS that demonstrates prompt-based VFX editing at a level I didn't think was possible yet. Add flames, lightning, particle effects, or complex animations to video using text instructions. The system preserves background integrity while injecting professional-grade effects with temporal consistency. Code is not available so I could not check for quality, resolution and prompt adherence. Some example shoots look pretty good, others not so much. For VFX professionals, the writing is on the wall: up-skill fast or become obsolete. The tools are evolving faster than most teams can adopt them. Those who learn to direct AI agents and master prompt-based workflows will dominate the next era of content production. Those who wait will watch their competitive advantage evaporate. We're at an inflection point in visual effects. The barrier to entry is collapsing. The speed of production is accelerating. Brand-safety and quality control in AI-generated content is becoming table stakes. The question isn't whether GenAI will transform your workflow. It's whether you'll lead that transformation or be disrupted by it. What's your up-skilling strategy for 2025? Want to see how we're building GenAI pipelines that transform content production? DM me to learn how your team can stay ahead of the curve.

AI TVC. Human alignment. Preproduction mattered more than ever. This ATOM Lucky Fortune AI TVC marks the final commercial of the 2025 ATOM Lucky Fortune campaign — and one of our most challenging AI-driven projects. The real challenge wasn’t using AI. It was making AI visuals work within an existing brand system. The more powerful the tool, the clearer the alignment needs to be. Strong preproduction, shared expectations, and respect for brand systems are what turn AI visuals into a real commercial — not just generated images. Every scene had to stay aligned with: -Brand guidelines & SUPER graphics -Existing mascots and mashhead UI -Approved characters, moods, and cultural settings That’s why preproduction played a bigger role than usual. Before generating anything, we locked scenarios, moods, character behavior, winning-moment structure, product journeys, and even frame-by-frame flow. From there, AI became a tool, supported by human-made CGI, VFX, editing, and creative direction — through multiple revision rounds — to deliver the best possible outcome within the available branding assets. Big kudos to our Creative, Post, AI, CGI & VFX teams for the discipline and patience. And sincere thanks to ATOM Myanmar and Passion Point Agency for trusting Magical Creation Myanmar Media Production throughout the Lucky Fortune Campaign 2025. AI may speed things up — but alignment and preproduction still define the result. Check the Video at Youtube https://lnkd.in/gH2CWZJu Linkedin https://lnkd.in/gj9hUYmt #staymagical #AITVC #PreproductionMatters #Brandfirst #CreativeProcess