0:00 / 0:00

That AI Page

@thataipage

INSTAGRAM · 2026-01-28Source

1.4Klikes

81comments

Remix This

Recreate with Kling 3 Motion Control

Make your own AI viral video

Prompt

GLOBAL LOCK:
Subject is a young woman with a stylized "Betty Boop" aesthetic: pale skin, very large expressive dark eyes with thick long lashes, small red lips, and short curly black hair styled in 1920s finger waves. She wears a vibrant red strapless floor-length gown with a ruffled top edge and large gold hoop earrings. The environment is a dark, cavernous stone dungeon or cave. A circular stone pedestal is in the center, surrounded by a ring of bright, flickering orange flames. Lighting is high-contrast, primarily warm firelight from below with deep shadows. The color grade is saturated with rich reds and deep blacks, featuring a slight vintage film grain. Pacing is rhythmic, with cuts and camera movements synchronized to a jazzy, vintage-style soundtrack.

[00:00–00:03]
Subject: The woman stands on the pedestal, looking directly up into the lens with a wide, charming smile.
Environment: The ring of fire is visible around her feet.
Action: She tilts her head slightly, her eyes wide and sparkling.
Framing: Extreme high-angle fisheye shot, looking straight down.
Movement: Slow zoom-in toward her face.
Lighting: Strong orange glow from the fire below, casting dramatic shadows upwards.
Speech: No speech, just a musical intro.

[00:03–00:08]
Subject: Close-up of her face as she begins to sing.
Action: Her mouth moves in perfect sync with the lyrics "Hell's bells, ringing in my ear." Her expressions are theatrical and animated.
Framing: Medium Close-Up (MCU), eye-level.
Movement: Slight handheld camera shake for realism.
Lighting: Warm, flickering light on her face; background is pitch black.
Speech: Speaker A (Woman), "Hell's bells, ringing in my ear," energetic and jazzy tone, high lip-sync strictness.

[00:08–00:15]
Subject: Full body shot of her dancing on the pedestal.
Action: She moves her arms gracefully and sways her hips. She looks down at the fire and then back at the camera.
Framing: Full shot, slightly high angle.
Movement: Smooth orbital pan around the subject.
Lighting: The firelight creates moving shadows on the cave walls behind her.
Speech: Speaker A (Woman), "It's certainly hot, the rhythm that's swell," playful delivery.

[00:15–00:23]
Subject: Close-up of her face, looking slightly away and then snapping back to the camera.
Action: Intense singing, eyebrows raised, eyes wide.
Framing: Close-Up (CU).
Movement: Fast zoom-in on the word "bells."
Lighting: Rim lighting on her black curls from the fire.
Speech: Speaker A (Woman), "Now somebody's got a hold of those bells," powerful vocal delivery.

[00:23–00:30]
Subject: She steps off the stone pedestal and walks directly through the ring of flames toward the camera.
Action: Her dress flows behind her; she maintains a confident, slightly mischievous expression.
Framing: Full shot, low angle to make her look imposing.
Movement: Tracking shot, moving backward as she walks forward.
Lighting: Intense orange light as she passes through the fire.
Speech: Speaker A (Woman), "They start to ring, they start to sing," melodic and rhythmic.

[00:30–00:43]
Subject: A montage of the woman in various poses and angles on the pedestal.
Action: Singing, pointing, and smiling. Text overlays appear: "Images With Bad Angles Are In The Past Now," "This Is Angles V2 Inside Higgsfield."
Framing: Rapid cuts between high-angle fisheye, MCU, and side profiles.
Movement: Dynamic pans and tilts.
Speech: Speaker A (Woman), "And somebody yells, it must be Hell's Bells!" climax of the song.

[00:43–01:02]
Subject: Demonstration of the AI tool. A split screen shows a man with long hair and then Leonardo DiCaprio in a tuxedo.
Action: A UI overlay shows a camera controller rotating the perspective of the static images. Text: "One Image, 12 Perspectives," "Watch The Camera Move."
Framing: Medium shots of the celebrities.
Movement: Smooth 180-degree horizontal rotation around the subjects' heads.
Lighting: Matches the original reference photos (natural daylight for the man, party lighting for DiCaprio).
Speech: Instrumental jazzy music continues.

NEGATIVE PROMPT:
Visual: Blurry features, inconsistent eye size, distorted fingers, flickering clothing textures, fire looking like flat 2D overlays, robotic or stiff body movements, background morphing, loss of character identity between shots.
Speech: Mismatched lip-sync, robotic vocal cadence, muffled audio, background noise, unnatural pauses, sibilance, clipping.

SPEECH PACK:
[00:03-00:08] "Hell's bells, ringing in my ear."
TAKE_A: (Jazzy, upbeat) Hell's BELLS... ringing in my EAR.
TAKE_B: (Theatrical, loud) HELL'S bells! Ringing in my ear.
TAKE_C: (Sultry, smooth) Hell's bells... ringing in my ear.

[00:08-00:15] "It's certainly hot, the rhythm that's swell."
TAKE_A: It's CERTAINLY hot... the rhythm that's SWELL.
TAKE_B: (Fast) It's certainly hot, the rhythm that's swell!
TAKE_C: (Playful) It's certainly hot... [giggle] the rhythm that's swell.

[00:15-00:23] "Now somebody's got a hold of those bells."
TAKE_A: Now SOMEBODY'S got a hold... of those BELLS.
TAKE_B: (Intense) Now somebody's got a HOLD of those bells!
TAKE_C: (Rhythmic) Now somebody's got a hold... of those... bells.

[00:23-00:30] "They start to ring, they start to sing."
TAKE_A: They start to RING... they start to SING.
TAKE_B: (Melodic) They start to ring... they start to sing!
TAKE_C: (Whispered) They start to ring... they start to sing.

[00:30-00:43] "And somebody yells, it must be Hell's Bells!"
TAKE_A: And somebody YELLS... it MUST be Hell's Bells!
TAKE_B: (Shouted) And somebody yells! It must be HELL'S BELLS!
TAKE_C: (Grand finale) And somebody yells... it must be... Hell's... Bells!

How thataipage Made Betty Boop Higgsfield Angles V2 Tutorial — and How to Recreate It

This case study analyzes a high-performance AI video showcase featuring a stylized "Betty Boop" aesthetic character in a dramatic, fiery dungeon setting. The video, posted by @thataipage, serves as a powerful demonstration of Higgsfield's "Angles V2" feature. It blends nostalgic character design with cutting-edge AI video generation, specifically highlighting the ability to manipulate camera perspectives from a single static image. The visual language is defined by high-contrast lighting, a saturated red-and-black color palette, and dynamic, wide-angle "fisheye" cinematography that creates an immersive, slightly surreal experience. By combining a recognizable pop-culture archetype with a "how-it's-made" technical reveal, the creator successfully captures both aesthetic appreciation and professional curiosity.

What You’re Seeing

The video is a montage of AI-generated clips featuring a woman who embodies a modern, live-action interpretation of Betty Boop. She is characterized by short, curly black hair with distinct finger waves, oversized expressive eyes with dramatic lashes, and a petite frame. She wears a vibrant red, strapless, floor-length gown with a ruffled neckline. The primary setting is a dark, cavernous stone room where she stands on a circular pedestal surrounded by a ring of actual flames.

Shot-by-Shot Breakdown

Time Range	Visual Content	Shot Language	Lighting & Tone	Viewer Intent
00:00 – 00:03	Subject looks up at the camera from the pedestal.	Extreme High Angle (Fisheye)	Warm fire glow from below, deep shadows.	Hook: Immediate visual impact and "uncanny" charm.
00:03 – 00:08	Subject singing "Hell's Bells" directly to camera.	Medium Close-Up (MCU)	High contrast, saturated reds.	Reinforce persona and demonstrate lip-sync.
00:08 – 00:15	Subject dancing; camera pans around the pedestal.	Full Shot / Orbiting	Dynamic shadows moving with the camera.	Showcase environment depth and motion consistency.
00:15 – 00:23	Subject singing with intense expressions.	Close-Up (CU)	Dramatic rim lighting on hair.	Emotional connection and detail showcase.
00:23 – 00:30	Subject walks through the ring of fire toward the camera.	Full Shot / Tracking	Flickering firelight, high drama.	"Wow" factor; demonstrating complex interaction.
00:30 – 00:43	Rapid cuts of different angles with text overlays.	Mixed Angles	Consistent with previous shots.	Educational: Explaining the tool (Higgsfield).
00:43 – 01:02	Demo of camera rotation on DiCaprio and Brad Pitt images.	UI Overlay / Split Screen	Varies by reference image.	Proof of concept: Showing versatility on famous faces.

Why It Went Viral: The Mechanics of Attention

The Power of Stylized Nostalgia

The choice of a "Betty Boop" aesthetic is a masterstroke in audience psychology. It taps into collective nostalgia while modernizing it through a "live-action" AI lens. This creates a visual friction—the character is familiar yet looks entirely new. The exaggerated features (large eyes, tiny waist) trigger a biological response similar to "baby schema," making the character inherently captivating and memorable. The "Hell's Bells" soundtrack adds a layer of 1930s jazz-age rebellion that perfectly matches the character's origins.

The "How-To" Reveal as a Retention Engine

The video doesn't just show a pretty animation; it solves a problem. By including the text "Images With Bad Angles Are In The Past Now" and showing the UI of the tool, the creator transforms the video from mere entertainment into a valuable tutorial. This encourages saves and shares among other creators who are looking for ways to improve their own AI video workflows. The demonstration using recognizable celebrities like Leonardo DiCaprio and Brad Pitt provides instant social proof—if it works on them, it will work on anyone.

Platform Signals & Algorithmic Triggers

From a platform perspective, the video excels in watch time and re-watchability. The fast-paced editing (cuts every 2-3 seconds) and the constant change in camera perspective keep the viewer's brain engaged. The "loop effect" is strong here; viewers often go back to the beginning to see if they can spot the "seams" after learning it was made from just two images. The high-contrast, fiery visuals are also "thumb-stoppers" in a crowded feed, ensuring a high click-through rate from the Explore page.

5 Testable Viral Hypotheses

The Archetype Hook: Using a hyper-stylized version of a classic cartoon character (Betty Boop, Jessica Rabbit) will outperform generic "beautiful woman" prompts by 2x due to nostalgia and distinctiveness.
The Technical Reveal: Explicitly stating "This was made from [X] images" creates a curiosity gap that drives viewers to watch until the end to see the proof.
Fisheye Perspective: Using extreme wide-angle or fisheye lenses in AI video creates a sense of "real-world" cinematography that masks AI artifacts better than standard focal lengths.
Dynamic Lighting (Fire): Scenes featuring flickering, motivated light sources (like fire or neon) increase perceived quality and realism because they demonstrate the AI's ability to handle complex light interactions.
Celebrity Validation: Including a 3-5 second "proof of concept" using a famous face (DiCaprio, Pitt) at the end of a video increases shareability by 40% by making the tech feel more "legit."

How to Recreate: From 0 to 1

Step 1: Character Design & Base Image

Start by generating a high-quality base image. Use Midjourney or DALL-E 3. Focus on a specific "era" or "style."
Prompt Tip: "1930s live-action Betty Boop, short curly black finger waves, large expressive eyes, heavy lashes, wearing a red strapless silk gown, cinematic lighting, 8k resolution."

Step 2: Environment Setting

Generate a second image or use an outpainting tool to create the background. For this video, a "dark stone dungeon with a ring of fire" provides the necessary drama and contrast.

Step 3: Tool Selection (Higgsfield)

Upload your base image to Higgsfield. This tool is specifically designed for the "Angles" feature shown in the video, allowing you to define camera movement after the image is generated.

Step 4: Defining Camera Motion

Use the "Angles" UI to set your keyframes. Start with a high-angle fisheye shot to mimic the hook of the original video. Set a second keyframe at eye level for the singing segments.

Step 5: Lip-Sync Integration

If your video generator doesn't handle speech, export the video and use a tool like Hedra or LivePortrait. Upload your AI video and an audio file of the song you want the character to sing.

Step 6: Adding Secondary Motion

In your video prompt, emphasize "flickering firelight," "moving hair," and "flowing dress fabric" to ensure the scene doesn't look static.

Step 7: Editing for Rhythm

Import your clips into CapCut or Premiere. Cut on the beat of the music. Use "Speed Ramps" to emphasize certain movements, like when the character looks up or steps through the fire.

Step 8: Text Overlays & CTA

Add bold, centered text overlays that explain the "magic." Use phrases like "One Image, 12 Perspectives" to highlight the value proposition. End with a clear Call to Action (CTA) to follow for more AI tips.

Growth Playbook: Distribution & Scaling

Opening Hook Lines

"Stop scrolling! This Betty Boop AI is actually insane."
"I turned one photo into a cinematic movie. Here’s how."
"The end of bad camera angles is finally here."

Caption Templates

Option 1 (The Tech Enthusiast):
The future of AI video is moving fast! 🚀 I used @Higgsfield’s new Angles V2 to turn a single image into this cinematic sequence. The camera control is a game changer for indie creators. What character should I try next? 👇 #aivideo #higgsfield #creativeai

Option 2 (The Aesthetic/Artistic):
Hell’s Bells 🔔 Bringing a classic icon to life with a modern twist. The lighting, the motion, the vibe... AI is finally getting the 'soul' right. Save this for your next project inspo! ✨ #bettyboop #cinematicai #digitalart

Hashtag Strategy

Broad (Reach): #aivideo #artificialintelligence #digitalart #contentcreator
Mid-Tier (Niche): #higgsfield #aimovies #creativeai #videoeditingtips
Long-Tail (Specific): #bettyboopstyle #imagetovideo #cameracontrol #aiartcommunity

Frequently Asked Questions

What tool makes it look the most similar?

Higgsfield is the primary tool used here for the specific camera angle manipulation.

What are the 3 most important words in the prompt?

"Fisheye," "Cinematic Lighting," and "Character Consistency."

Why does the generated face look inconsistent?

This usually happens if the seed isn't locked or if the motion strength is set too high; keep motion between 3-5.

How can I avoid making it look like AI?

Add film grain in post-production and use motivated lighting (like fire) to ground the character in the scene.

Is it easier to go viral on Instagram or TikTok with this?

Instagram Reels currently favors high-aesthetic "eye candy" like this, while TikTok prefers the "how-to" breakdown.

How should I properly disclose AI use?

Use the platform's built-in "AI-generated" label and mention the tools used in the caption to build trust.