0:00 / 0:00

▶

Maria Kallevik †

@maria.kallevik

INSTAGRAM · 2025-04-29Source

23.0Klikes

232comments

Remix This

Recreate with Sora 2

Make your own AI viral video

Prompt

GLOBAL LOCK:
The subject is a young Caucasian woman with long blonde hair. In the tutorial segments, she wears a thick, textured green and grey knit sweater and blue jeans, sitting in a mossy forest. In the cinematic segments, she wears a flowing, sleeveless white sundress and is barefoot. The environment is a dense, temperate forest with moss-covered ground and tall evergreen trees. The lighting transitions from natural, overcast daylight in the BTS/tutorial shots to a warm, ethereal golden glow in the final VFX shots. The camera language alternates between direct-to-camera MCU for the host and dynamic, handheld tracking shots for the cinematic sequences. The speech is a clear, enthusiastic female voiceover with a slight European accent, delivered with a helpful, instructional cadence.

[00:00–00:03]
The subject (host) is in a Medium Close Up, sitting on a mossy rock in the forest, wearing the green sweater. She speaks directly to the camera. A quick, 0.5-second flash of a golden portal appears as a teaser. The lighting is soft and natural.

[00:03–00:09]
Behind-the-scenes footage. The subject, now in a white dress, runs barefoot through the forest. A cameraman with a gimbal rig follows her. The footage looks raw and slightly desaturated. The subject looks cold but determined.

[00:09–00:17]
Screen recording of the Artlist website. A cursor navigates the "AI Image & Video" tab. An AI prompt is typed: "hundreds of small glowing orbs and natural, with a warm, ethereal, peaceful and slightly nostalgic." The screen shows the generation of a fantasy forest image.

[00:17–00:24]
Screen recording of Blender software. A 3D view shows a single glowing sphere (orb) being modeled. The interface shows nodes and a particle system being applied to create a trail of floating dots in a 3D forest environment.

[00:24–00:29]
Host returns to MCU, gesturing with her hands as she explains the process. The background is the blurred forest.

[00:29–00:35]
Screen recording of After Effects. The "Roto Brush" tool is used to create a pink outline around the subject in the white dress as she walks. This separates her from the forest background.

[00:35–00:47]
Screen recording of the compositing process. A circular, fiery portal element is added behind the subject. The colors are adjusted from purple/blue to a warm orange/gold. Light rays are added, emanating from the portal center. A "mesh warp" is used to animate the portal closing.

[00:47–00:52]
Screen recording of DaVinci Resolve. The colorist moves the color wheels to shift the shadows to green and the highlights to a warm golden-yellow, creating a "cinematic" look.

[00:52–01:00]
Final Cinematic Reveal. The subject in the white dress runs through the forest, surrounded by hundreds of floating, glowing golden orbs. She approaches a massive, swirling golden portal. The camera follows her from behind. The video ends with a side-by-side comparison of the AI reference image and the final rendered video.

NEGATIVE PROMPT:
Visual: robotic movement, distorted face, inconsistent hair color, blurry textures, floating limbs, flickering light, watermark, low resolution, messy masking, flat lighting.
Speech: monotone delivery, background noise, muffled audio, robotic voice, long silences, stuttering, poor lip-sync.

SPEECH PACK:
[00:00-00:03] "How I made this video inspired by Narnia."
[00:03-00:08] "I started by running barefoot in the forest for two hours straight, which was freezing."
[00:08-00:17] "Next, I needed some reference for the effect. So I went to Artlist and used their new AI image generator, which gave me a great visual direction."
[00:17-00:24] "And it can even animate images, which is so cool. I then hopped into Blender, modeled a glowing orb, and turned it into a particle system."
[00:24-00:29] "After doing this process on all the other 16 shots, I finally hit render."
[00:29-00:35] "In compositing, I added the orbs along with some glow, and then rotoscoped myself out to place me in front of them."
[00:35-00:47] "For the portal, I used this stock element and changed the colors, added glow and light rays, and then used this video I made a while ago for the background. To close the portal, I tweaked it using a mesh warp and scaled it down until it was gone."
[00:47-00:52] "As a final touch, I did a color grade and found some epic sound effects from Artlist."
[00:52-01:00] "Here are Artlist's AI reference images, and here's how my version turned out. You can find the full video on my page."

TAKE_A (Instructional): Clear, slow, emphasizing the tool names.
TAKE_B (Energetic): Faster pace, higher pitch on "so cool" and "epic."
TAKE_C (Storytelling): Warm, conversational, focusing on the "freezing" experience.

How maria.kallevik Made This Narnia Portal VFX — and How to Recreate It

This case study analyzes a high-production-value VFX breakdown tutorial by creator Maria Kallevik. The video masterfully blends "behind-the-scenes" (BTS) grit with "cinematic fantasy" polish, specifically focusing on a Narnia-inspired portal effect. By showcasing the bridge between raw iPhone/mirrorless footage and professional post-production tools (Blender, After Effects, DaVinci Resolve), the content appeals to both aspiring filmmakers and casual viewers who enjoy "magic revealed" narratives. The aesthetic transitions from a muted, mossy forest palette to a vibrant, golden-hour fantasy world, utilizing glowing particle systems and dynamic portal compositing to create a high-retention loop.

What You’re Seeing

The video follows a hybrid structure: a direct-to-camera "talking head" host in a cozy green sweater provides the narrative thread, while the B-roll alternates between technical screen recordings and cinematic final shots. The setting is a dense, evergreen forest with heavy moss coverage, providing a naturalistic "fairytale" backdrop. The wardrobe shifts from the creator's casual knitwear to a flowing white sundress for the cinematic sequences, creating a clear distinction between the "teacher" and the "character."

Shot-by-Shot Breakdown

Time Range	Visual Content	Shot Language	Lighting & Color	Viewer Intent
00:00–00:03	Host intro + quick flash of the final portal effect.	MCU (Medium Close Up) + Fast Cut.	Natural overcast forest light.	The Hook: Promise of high-quality result.
00:03–00:09	BTS: Running barefoot in the forest, camera rigs shown.	Wide Shot / Handheld.	Raw, desaturated green/brown.	Humanize: Show the "struggle" (freezing weather).
00:09–00:17	AI Image Generation (Artlist) UI walkthrough.	Screen Recording.	Digital UI (Dark Mode).	Tutorial Value: Show the starting point.
00:17–00:29	3D Modeling in Blender (Orbs and particles).	Screen Recording / 3D Viewport.	Technical grey/wireframe.	Authority: Prove technical skill.
00:29–00:47	Compositing in After Effects (Rotoscoping & Portal).	Screen Recording / Masking.	High contrast masks (pink/purple).	The "Magic": Show how layers come together.
00:47–00:52	Color Grading in DaVinci Resolve.	Screen Recording / Color Wheels.	Warm golden shift.	Polish: The final "look" transformation.
00:52–01:00	Final cinematic reveal vs. AI reference.	Full Screen Cinematic.	Golden, glowing, ethereal.	Satisfaction: The "Big Reveal" and CTA.

Why It Went Viral

The "Aesthetic Authority" Strategy

This video succeeds because it doesn't just show a result; it documents a technical journey. The "Narnia" theme taps into deep-seated childhood nostalgia and a universal love for fantasy. By starting with the "freezing" BTS footage, Maria creates an immediate emotional connection—viewers respect the effort before they even see the art. The transition from a "messy" reality to a "perfect" VFX shot triggers a dopamine release associated with problem-solving and transformation.

The Tool-Stack Transparency

Modern audiences are obsessed with "how it's made." By naming specific tools (Artlist, Blender, AE), the video moves from "entertainment" to "educational resource." This encourages Saves (to reference later) and Shares (to show other creator friends). The mention of AI specifically targets the current tech zeitgeist, showing a practical, non-threatening way to use AI as a "reference" rather than a replacement for craft.

Platform Perspective: The Retention Loop

From an algorithmic standpoint, the video uses a rapid-fire editing rhythm. Every 2-3 seconds, the visual changes—from a face to a forest to a software UI. This prevents "scroll-away" boredom. The "Comparison" at the end (AI vs. My Version) is a classic high-retention tactic that forces the viewer to watch until the very last second to see if the creator "beat" the AI or matched the vision.

5 Testable Viral Hypotheses

The "Struggle" Hook: If you show the physical discomfort of filming (e.g., "running barefoot in the cold"), viewers are 40% more likely to stay for the result out of respect for the "grind."
The AI-Reference Bridge: Using AI to generate a *concept* rather than the *final video* positions you as a "Director" rather than just a "Prompter," increasing perceived value.
Software UI Porn: Fast-motion screen recordings of complex software (like Blender nodes) create a "hypnotic" effect that boosts watch time, even if the viewer doesn't understand the tool.
Nostalgia Anchoring: Explicitly naming a famous IP (Narnia) in the first 3 seconds provides an immediate mental framework for the viewer, reducing "contextual friction."
The "Layer Cake" Reveal: Showing the rotoscoping (the pink outline) makes the final composite feel more "earned" and impressive than if it just appeared.

How to Recreate (Step-by-Step)

Concept Selection: Choose a high-fantasy theme (e.g., Narnia, Harry Potter, Lord of the Rings). Identify one "Magic Element" (a portal, a floating object, a transformation).
AI Moodboarding: Use a tool like Artlist AI or Midjourney to generate 4-5 "Visual Direction" images. Focus on lighting and color palettes (e.g., "Golden orbs in a dark forest").
The "Raw" Shoot: Film your subject in a matching real-world environment. Pro Tip: Use a tripod for shots involving complex VFX to make masking easier. Wear high-contrast clothing (like a white dress against green moss).
BTS Documentation: Have a second person film you filming. Capture the "un-glamorous" parts—the cold, the equipment, the setup.
3D Asset Creation: In Blender, create your magical elements. For orbs, use a simple sphere with an "Emission" shader and a "Particle System" to make them float and follow a path.
The "Roto" Grind: Import your footage into After Effects. Use the Roto Brush tool to separate your subject from the background so the 3D orbs can fly *behind* them.
Portal Compositing: Use stock footage of fire or smoke. Change the blending mode to "Screen" or "Add," and use a "Mesh Warp" to give it a circular, portal-like shape.
Color Grading: Use DaVinci Resolve to unify the 3D elements and the raw footage. Add a "Glow" effect to the highlights to make the portal feel like it's actually emitting light onto the trees.
Sound Design: Layer "Whoosh" and "Sparkle" sound effects. The sound should peak exactly when the portal opens or the subject passes through.

Growth Playbook

Opening Hook Lines

"How I turned a freezing forest into a Narnia portal."
"Stop using AI for everything—use it as a reference instead."
"The secret to cinematic VFX isn't the camera, it's the layers."

Caption Templates

The "Process" Template:
I spent 2 hours running barefoot in the forest for this 5-second shot. 🥶 Worth it?
Here’s the breakdown of how I used AI, Blender, and AE to create this Narnia-inspired portal.
Which part of the process surprised you most? 👇
#VFX #Filmmaking #BehindTheScenes

Hashtag Strategy

Broad: #filmmaking #vfx #creative #digitalart (High reach, high competition)

Mid-Tier:

Niche:

FAQ

What tools make it look the most similar?

Blender for 3D particles and After Effects for the final compositing and rotoscoping.

How do I make the portal look "real"?

Add "Light Wraps" in compositing so the portal's glow bleeds onto the edges of the subject.

Is AI necessary for this workflow?

No, but it speeds up the "concept art" phase significantly by giving you a lighting reference.

How can I avoid making it look like "bad" AI?

Only use AI for the background or reference; keep the human subject and the primary movement real.

Which platform is best for VFX breakdowns?

Instagram Reels and TikTok both love "Magic vs. Reality" transitions, but YouTube Shorts has better long-term search intent.