Traditional luxury lifestyle commercials, especially shots involving street choreography, colliding talent, slow-motion prop physics, and liquid splashes, require studio rigs, high-speed phantom cameras, and days of set production.
For this spec commercial for % Arabica, the objective was to prove that agency-grade visual storytelling, strict brand consistency, and complex physics can be achieved directly from static reference assets using a hybrid AI generation + editorial post-production pipeline.
Uploaded alongside the final spot are the exact three visual inputs used to anchor the commercial:
Talent & Wardrobe Anchor: Locked the model's facial identity, skin tone, red-and-white tweed mini dress, gold watch, pearl bracelet, and pendant necklace from a single indoor portrait into an outdoor sunlit street walk.
Eyewear Hero Prop: Integrated the rimless pink-tinted Miu Miu MUA51S sunglasses with gold hardware onto the talent, and subsequently animated them as a standalone mid-air prop.
Packaged Product Anchor: Preserved the exact cup geometry, layered espresso-milk gradient, eco-straw, and crisp % logo typography throughout intense motion.
Leveraged Higgsfield AI and its Seedance 2.5 video engine to fuse the three separate image references into one cohesive, physically believable scene. Engineered prompt choreography and camera trajectories to execute a continuous low-angle tracking tilt-up, starting at the red stilettos on the pavement and rising smoothly to eye level without texture warping or identity drift.
Directed a secondary passerby interaction in a beige linen suit to create a natural kinetic trigger. Using Seedance 2.5's motion control, the collision launches both the sunglasses and the % Arabica cup toward the lens in hyper-slow motion. The camera flies straight through the translucent pink Miu Miu lens—using it as a natural optical wipe, while iced coffee droplets splash upward in crisp detail around the suspended cup.
3. Editorial Rhythm, Depth of Field & Brand Finish
Took the generated sequences into post-production to tighten the pacing, enhance shallow depth-of-field separation between the floating product and the blurred city backdrop, sync every stride and impact frame to an energetic rhythmic vocal beat, and resolve into a clean minimalist % ARABICA® brand lockup.
Tools & Deliverables
AI Video & Generation: Higgsfield AI (Seedance 2.5), Multi-Reference Image Conditioning
Post-Production & Finishing: Adobe Premiere Pro, After Effects, Photoshop
A 60-second fan short made entirely with AI video, with my own face as the lead. Nine shots, seven characters, built from one selfie and a set of reference images. No cameras were used.
I started from a public prompt guide by kaye_creatives on Higgsfield and treated it like a shooting script: built my own character sheet, cast the missing character, generated each shot separately, then cut, captioned, titled and mixed it myself.
What I'd pass on to anyone trying this:
Give every reference image one job. Face, room, prop or camera angle, never two at once.
Put exact counts in the prompt. Models add people and props unless you tell them how many.
Generate long, cut short. My first assembly was 67 seconds. The final is 60, and the pace is where the quality came from.
The big stacked WORK / RESULTS / RATES menu next to the loose purple stickers keeps the portfolio playful while the navigation stays easy to scan. Do the stickers react on hover, or are they tied to scroll?