How-To · Tech Packs & Renders

Turn Tech Pack Sketches Into On-Model Catalog Shots

For the first few months of a garment’s life, the only visual you own is a drawing. Here’s how to photograph it anyway.

July 25, 2026 · 10 min read

Top row — 2D technical flat sketches of a raglan tee in forest green, blush pink and navy. Bottom row — the MODA AI on-model outputs of those same three flats, on one model in one clean studio

Above: the tech pack drawings. Below: the same three, photographed on a model in one studio. No sample was ever cut.

8–16

weeks from design to a production-ready garment, before any shoot

Apex Fashion Lab

30–60%

fewer physical samples already, thanks to 3D virtual sampling

Apex Fashion Lab

~$1

per 10-shot set, generated straight from the drawing

The Months When All You Have Is a Drawing

Every garment starts as a technical flat: the line drawing in your tech pack that tells the factory where the seams go, how the sleeves are cut, and where the logo sits. It is a precise document and a terrible marketing asset. You cannot put a CAD sketch on a product page.

Getting from that drawing to something you can photograph takes months. Apex Fashion Lab puts the full run — tech pack, sourcing, proto sample, fit corrections, pre-production sample, approval — at eight to sixteen weeks. The shoot happens after that. So for a quarter of the year the design exists, the costing exists, the factory slot exists, and the only picture you own is a drawing.

The industry already decided digital assets are good enough to make decisions from: 3D virtual sampling in tools like CLO3D and Browzwear has cut physical samples by 30 to 60 percent and saves $500 to $5,000 a collection. But a flat still reads as a flat, and a render still reads as a render. Neither looks like a photograph of a person wearing your product. That last gap is the one MODA AI closes.

The Input

A Flat Sketch Is Enough to Start

The garment below is a raglan tee from an activewear label: white body, contrast sleeves, a wordmark across the chest, and nothing on the back. Front and back flats are all that went in. No sample was cut, no fabric was sourced, nobody stood in front of a camera.

Input technical flat — front view CAD sketch of a raglan tee, white body with forest green sleeves and a wordmark across the chest

Front flat

Input technical flat — back view CAD sketch of the same raglan tee, blank back with raglan sleeve seams marked

Back flat

The tech pack drawing, exactly as the designer drew it. This is the whole input.

2D flats are the common case, but the same works for CAD line drawings, colored-up flats, and 3D garment renders exported from CLO3D or Browzwear. What matters is that the construction reads clearly in the drawing: seam lines, sleeve cut, color blocking, and where the graphic sits. Whatever is legible in the flat is what carries through.

Fidelity

The Logo Holds Its Place, Its Scale, and Its Shape

This is the part that decides whether a drawing-to-photo workflow is usable or a novelty. A wordmark has to land in the right spot, at the right size relative to the body, and then behave like it is printed on fabric rather than pasted on glass. Across the frames below the mark sits where the flat put it, foreshortens as the model turns, and curves with the chest when she sits down.

MODA AI output — raglan tee front view, chest wordmark straight-on at correct placement and scale

Straight on

MODA AI output — raglan tee three-quarter view, the chest wordmark foreshortening naturally with the turn of the body

Turned

MODA AI output — model seated in the raglan tee, the chest wordmark curving with the fabric across the chest

Seated

MODA AI output — raglan tee back view, back left blank exactly as the back technical flat specifies, raglan seams visible

Back

Placement, scale, foreshortening — and a blank back, because the back flat is blank.

The back frame is the one worth staring at. The back flat has no graphic on it, and the output has no graphic on it. A tool that guesses would have wrapped the logo around to the back or invented a neck-line print; this one read the spec. The raglan seam from neck to underarm and the sleeve color blocking come through the same way. For a deeper look at how prints and logos survive generation, and where tools fail them, see the logo reproduction guide.

The Set

One Sketch, a Full Catalog Shot List

A product page needs more than a hero image. From the single flat, the run returns the catalog shot list a PDP is built from — front, three-quarter, turned, back, seated, and side profile, all on the same model in the same studio. Here is the navy colorway’s complete set. This brand asked for six frames instead of the full ten, because six covered what their product page needed.

MODA AI output — the raglan tee in navy on a model, straight-on front, full length, clean studio

Front

MODA AI output — the raglan tee in navy, three-quarter front with hand on hip, chest wordmark foreshortening with the body

Three-quarter

MODA AI output — the raglan tee in navy, turned standing pose, full length with black leggings and trainers

Turned

MODA AI output — the raglan tee in navy, back view, blank back with the raglan sleeve seams visible

Back

MODA AI output — the raglan tee in navy, seated on a stool, wordmark curving with the chest

Seated

MODA AI output — the raglan tee in navy, side profile, full length

Side

Six frames from the navy colorway. A standard run returns ten across 16+ pose angles.

The Range Board

The Whole Colorway Range, Before the Whole Range Exists

Most brands sample one colorway and rely on it to sell the rest. Because every colorway runs through the same face and background reference, the full range comes back looking like one shoot. Below is the range as a development board: the flat that went in, the shot that came out, and three more angles behind it, for each colorway in the range.

Range Board — Raglan Tee

3 colorways · generated from tech pack flats

Illustration only

Range board spec — technical flat of the raglan tee in the forest colorway

Range board output — the raglan tee in the forest colorway on the shared model and studio
The raglan tee in forest, additional catalog angle 1The raglan tee in forest, additional catalog angle 2The raglan tee in forest, additional catalog angle 3

Forest

RGL-01
FABRIC ——
SIZE RUN ——

Range board spec — technical flat of the raglan tee in the blush colorway

Range board output — the raglan tee in the blush colorway on the shared model and studio
The raglan tee in blush, additional catalog angle 1The raglan tee in blush, additional catalog angle 2The raglan tee in blush, additional catalog angle 3

Blush

RGL-02
FABRIC ——
SIZE RUN ——

Range board spec — technical flat of the raglan tee in the navy colorway

Range board output — the raglan tee in the navy colorway on the shared model and studio
The raglan tee in navy, additional catalog angle 1The raglan tee in navy, additional catalog angle 2The raglan tee in navy, additional catalog angle 3

Navy

RGL-03
FABRIC ——
SIZE RUN ——

Illustration only. MODA AI generates the imagery; the greyed fields — fabric, size run, style data — stay yours to fill in.

Pre-Production

What the Weeks You Get Back Are Worth

Having catalog imagery months early changes the order you can do things in. Brands use it to open pre-orders and take deposits before stock lands, to put real-looking product in a buyer or retail deck instead of a line drawing, and to start building content while the factory is still cutting. It also settles arguments: seeing a colorway on a body is a faster way to decide whether it earns a place in the range than staring at a swatch.

The colorway case is the practical one. Sampling every color in the range means a set of fabric and a sew-down for each one, then a shoot. One sample plus the flats gets you the full range on the site, and the colorway that actually sells can be the one you commit fabric to.

What this does not do

It does not replace your sampling process. A generated image cannot tell you that the jersey feels like sandpaper, that the raglan seam rubs, or that the size 12 pulls across the shoulder. Fit, drape, and hand-feel are decided on a physical proto and a pre-production sample, and you should still touch the garment before bulk. This buys you the imagery early, not the sign-off.

Waiting for the Sample vs. Shooting the Flat

The sampling process runs either way. The question is whether your imagery waits for it.

 Wait for the sampleShoot the flat
Imagery availableAfter sampling and a shootFrom the tech pack, in minutes
What you needFabric, a sewn sample, a studioThe front and back flat
Colorways shownThe ones you sampledEvery colorway in the range
Pre-orders and decksLine drawings until stock landsCatalog imagery from day one
Fit and hand-feelJudged on the physical sampleStill judged on the physical sample
Cost per styleSample, model, photographer, studioAbout $1 per 10-shot set

The Range

The Rest of the Range, Same Studio

The other colorways, generated the same way from the same set of flats. Different colors, one model, one setting, one look.

MODA AI output — the raglan tee in blush pink, straight-on front, same model and studio as the rest of the rangeMODA AI output — the raglan tee in blush pink, three-quarter front with hand on hipMODA AI output — the raglan tee in blush pink, turned standing pose, full lengthMODA AI output — the raglan tee in blush pink, back view, blank back with raglan seams visibleMODA AI output — the raglan tee in blush pink, seated on a stoolMODA AI output — the raglan tee in blush pink, side profile, full length

The blush colorway, all six frames. Every one of them started as the same line drawing.

Frequently Asked Questions

What kind of drawing or render can I use?

A technical flat from your tech pack, front and back. CAD line drawings, colored-up flats, and 2D or 3D renders from CLO3D or Browzwear all work too. The clearer the seams, sleeve cut, and color blocking read in the drawing, the more accurately they carry into the on-model shots.

Does the logo survive the jump from sketch to photo?

Yes. In this example the chest wordmark keeps its placement and scale, foreshortens in the three-quarter frames, and curves with the chest when the model sits. The back stays blank because the back flat is blank — it followed the spec rather than guessing.

Does this replace making a physical sample?

No, and it shouldn't. This produces imagery, not a fit or quality decision. You still need the proto and pre-production sample to judge fit, drape, and hand-feel before bulk. What it replaces is the wait for imagery — you can market and pre-sell while sampling runs in parallel.

Can I use these before the product is manufactured?

Brands do, for pre-orders, buyer decks, and campaign planning — but be straight with customers about it. These are AI-generated images of a design, not photographs of a manufactured unit, so what ships has to match what was shown. Our AI model disclosure guide covers how brands are labelling this and what the current rules require.

How many angles do I get from one sketch?

A run returns ten on-model images across 16+ pose angles: front, three-quarter, back, side, detail, and seated. The brand in this post asked for six rather than the full ten, because six covered what their product page needed.

Will every colorway look consistent?

Yes, if you set one face reference and one background reference and run each colorway through the same two. The colorways here came back on the same model, in the same studio, styled the same way, so the range reads as one collection.

Photograph the design, not the wait.

Install MODA AI from the Shopify App Store. Two free credits, no card. Upload your flat and see the range on a model.

Get Started Free

Or read what Shopify merchants say — reviews on the Shopify App Store →

More from MODA AI