MCM House
Case Study12/06

MCM House

MCM House already had a finished campaign shoot. We generated the missing product shots to sit inside it — same locations, same light, same grade.

ClientMockberg
Year2026
ServicesAI Ecommerce Visuals

Introduction

PMCM House is an Australian furniture brand built on relaxed, lived-in interiors — natural timber, soft linen, and pieces made to feel personal rather than styled.

Their campaign was already shot and graded, but a handful of products from the collection never made it in front of the camera. Those pieces needed images that belonged to the same set, not a second shoot with a different look.

Year

2026

Duration

1 Week

Client

MCM House

Challenges

The new shots had to sit next to real photography without giving themselves away — same location, same light, same grade.

Furniture makes that difficult. Fabric weave, timber grain, and reflections are exactly where a generated image usually falls apart.

The process behind this case.

01
GROUND RULES

Locking the moodboard and specs

We set the original shoot location, the colour grade, and the specs — resolution, aspect ratio, crop. Every frame after this gets measured against them.

Locking the moodboard and specs
01
LOCATION DEVELOPMENT

Building a setting for every shot

Several location options per frame, all built against the moodboard. MCM House picked one per shot.

01
PRODUCT PLACEMENT

Putting the real product in the room

Each shot got a brief — product, location, grade reference, styling note. The client then chose between two finished placements.

Putting the real product in the room
01
REFINEMENT & UPSCALE

Pushing the frame past "good AI"

Textures, reflections, foliage, hardware, and fabric weave rebuilt until the shot holds at full size, then upscaled to 4K for print and web.

Pushing the frame past "good AI"

Solutions

We rebuilt the world of the original shoot first, then generated the missing pieces inside it.

Location, colour grade, and technical specs were locked before a single frame was produced, so every shot could be measured against the campaign that already existed. The client chose the location and the placement on each shot before it went to refinement.

media1
media2
AI-Generated Product Catalog
media3
media4
The missing pieces of the collection, produced to slot straight into the existing campaign.
media5
media6
media7
media8
media9

media1
media2
AI-Generated Product Catalog
media3
media4
The missing pieces of the collection, produced to slot straight into the existing campaign.
media5
media6
media7
media8
media9

media1
media2
AI-Generated Product Catalog
media3
media4
The missing pieces of the collection, produced to slot straight into the existing campaign.
media5
media6
media7
media8
media9

Results

The collection is now covered end to end — the pieces that never went in front of a camera sit beside the ones that did, with nothing marking the difference.

A missing product no longer means a reshoot. The campaign can be extended from the material the brand already has, in the same look, without returning to the location.