Skip to content
MadooMadooDocs

I04 — Three garments, one model, one set: a lookbook with shop codes

A small fashion brand has three new pieces, each photographed flat on a table. A lookbook needs them worn, by the same person, in the same place, so the page reads as one collection; and each look must lead to its shop page — a QR code under every picture, for the printed lookbook in the shop and the PDF on the phone.

This example makes the set once, photographs the model there in each garment, styles each look as said, and lays out the lookbook with labels and codes that are exact and checked.

The workflow in four stages, with the reference run. 1, one model, three garments, and the facts: labels with name and price, product codes, the shop link. 2, one set for the collection: the courtyard, Text to Image, once, 5 credits. 3, three looks photographed in the set: overshirt with a hand on the hip, dress walking, sweater angled — Try-On with the set as the place, one shoot with the same light and true scale, styling said per garment, 5 credits each. 4, cards and the lookbook: a card per look with a white band, the label and a QR to the product, 0 credits; every QR read back, 3 of 3 match their links; the lookbook page, 20 credits.

The exact graph is the workflow (as a diagram).

The case

Section titled “The case”
Input Who provides it Reference run
Model the brand’s model photo a person in a white T-shirt, black trousers and white sneakers, made with Text to Image
Garment 1–3 the product team, photographed flat an ochre linen overshirt, a terracotta linen dress, a Breton striped sweater
Garment k Label the merchandiser Ochre linen overshirt · € 189 …
Garment k Code the shop ochre-overshirt …
Shop Link the shop https://example.com/shop/
Collection Title the brand Summer in the Courtyard

The inputs: the model in a white T-shirt and black trousers on grey; the ochre overshirt, the terracotta dress and the striped sweater laid flat; and the set the workflow made, an empty Mediterranean courtyard with bougainvillea, whitewashed walls and terracotta tiles.

What it shows

Section titled “What it shows”
  • One set, made once, for every look. Text to Image makes an empty courtyard described as a photographer would — a full-length fashion shoot, at eye level, six metres away, a 50 mm lens, a clear area of floor where a model will stand. The same picture goes into every look — the strongest way to make many pictures look like one shoot (Consistency and brand).
  • One person throughout. The same model photo in every Virtual Try-On: face, hair and body are kept; only the garment, the pose and what it replaces change.
  • Photographed in the set, not pasted on it. Given the set as background_image, Try-On photographs the model in that courtyard as in one shoot: on its tiles at true scale and in its perspective, in its soft warm light, with the colour of the terracotta reflected on her, shadows where her feet touch the ground, and the same depth of field and grain as the courtyard. The framing moves a little from look to look, as a photographer’s would; the place stays the same. Nothing about the place is said in the looks’ custom_prompt — the node does it.
  • Styling is part of the brief. Each look says what the garment does to the outfit: the overshirt worn open over the T-shirt, keep the trousers and sneakers; the dress replaces the T-shirt and the trousers. Without it the model decides — in a first run it dropped the trousers under the overshirt, in a second it kept them under the dress.
  • Words and codes that are exact. The labels are set by Text on Image in a white band added below each photo with Extend Canvas — nothing covers the look. Each QR code is built from the shop link and the product code with a Text Template, and read back by the step: 3 of 3 match (Layouts, batches and codes).

The workflow, step by step

Section titled “The workflow, step by step”
Phase Steps What they do here
A — set A1 The Collection’s Set Text to Image, 3:4, an empty courtyard for a full-length shoot
B — looks B1–B3 Look k in the Set Virtual Try-On with the model, the garment and the set; poses hand_on_hip, walking_stride, standing_angled; a styling line per garment in custom_prompt
C — cards per look Cka Photo 900 × 1200, Ckb White Band, Ckc Label in the Band, Ckd Product Link, Cke QR to the Product, Ckf QR at 180 px, Ckg QR in the Band the photo cut to size, 230 px of white below it, the label at the left of the band, the QR at the right
D — page D1 Three Cards, D2 Room for the Title, D3 Collection Title Stack in a row, a margin with room at the top, the title

The reference run

Section titled “The reference run”

The lookbook page: “Summer in the Courtyard” at the top; three cards side by side, each the same young woman photographed in the same courtyard, on its terracotta tiles in front of an arched wooden door under a bougainvillea, in its soft warm light — in the ochre overshirt open over a white T-shirt with black trousers, a hand on the hip; in the terracotta dress with a tie belt, walking, bare legs and white sneakers; in the Breton sweater with black trousers, standing at an angle — and under each a white band with the name and price and a QR code.

Output Result
set the courtyard, 896 × 1200
look_1 … look_3 896 × 1200, the model in each garment, in the set
card_1 … card_3 900 × 1430, look, label and QR
qr_check_1 … qr_check_3 scanned: true, matches: true — e.g. the dress’s
lookbook 2892 × 1750

20 credits: 5 for the set, 5 per look; cards, codes and page 0.

Adapt it

Section titled “Adapt it”
  • More garments. Each garment is a branch: an input, a label, a code, a Try-On and the seven card steps. For a whole collection, run the per-look part over a list with an iteration and collect the cards with a template (Layouts, batches and codes).
  • Your set. Replace A1 with a photo of your own location as an input; every look will be photographed there. A photo taken for a full-length shoot — the whole floor where people will stand, at eye level — works best.
  • Your models. Use photos of your own models, with their release — never a real person without consent.
  • Motion. A look is the starting frame of a fashion clip: see Subjects, continuations and fashion clips.

Install it

Section titled “Install it”
Terminal window
node madoo-install-example.mjs i04-lookbook --api <your Madoo API URL> --key <key.json> --run --save lookbook

The model, the overshirt, the dress and the sweater are pictures made with Madoo’s Text to Image node; the model is not a real person. The brand, the names, prices and codes are invented. The looks and the set were made by Nano Banana 2, for this example.