Physical and digital patterns in luxury: why fitting is not done on screen

Published 2026-07-15 · Teresita Piacentini, industrial pattern maker

The starting point: what digital does well

Before defending the physical pattern, it is worth acknowledging what pattern-making CAD software has genuinely improved in a studio's daily work. Grading is the clearest case: on an outerwear piece in 7 sizes, manual grading on paper took days of work; in CAD, a well-set-up grading runs in a few hours with repeatable, on-screen verifiable results. Marker making (nesting) is another case: optimisation software reduces fabric waste in ways no human with a pencil could replicate at scale. The digital file also has logistical value: it can be emailed to a manufacturer in any country without risk of deterioration or loss of the physical pattern.

These advantages are real and irreversible. No modern pattern-making studio works without CAD for the development and production phases.

What digital does not capture: fitting on real material

The problem arises when you try to move the initial construction and fitting phase to digital, particularly on technical garments. A down-filled outerwear piece with 16 channels, 90 grams per square metre of fill, a ripstop nylon outer layer and an internal thermal lining behaves in a way no drape-simulation software reproduces faithfully. The reasons are physical:

• The volume of the fill. A genuinely thick down jacket changes the relationships between parts in ways a 2D simulation cannot predict. The sleeve, once filled, sits forward compared with the flat projection; the shoulder sets differently; the real waistline of the closed garment differs from the flat pattern's perimeter.
• The behaviour of technical fabrics. Stretch fabrics, soft-shells and fabrics with water-repellent treatments have elongation and stiffness that vary with grain direction. CAD uses average parameters that approximate, but do not replicate, behaviour on a living body.
• Fill compression in fitting. When the garment is worn, the fill compresses at precise points (armpit, neck curve, waist). These compressions affect fit and require adjustments that only emerge on the mannequin or a fit model.

Physical fitting on a prototype sewn in the real material (or a substitute of equal weight and behaviour) remains the only way to validate a technical outerwear pattern before proceeding with grading.

The "approved simulation" trap

In recent years, some large maisons have experimented with partially digital approval processes, where the garment is simulated in 3D before the first physical sample is sewn. In some segments — knitwear, jersey, unstructured garments — the result is accurate enough to reduce the number of physical samples needed. In the technical outerwear segment, especially down jackets and multi-layers, the practical experience our clients report is different: the physical sample always corrects something the simulation had not predicted.

The risk of relying too much on simulation in this segment is that problems emerge late — mid-collection, or worse, in production. A timely physical fitting is economically less risky than a production run to correct. Whether a studio fits on a physical sample or only in simulation is one of the criteria to ask about at first contact.

The hybrid workflow that actually works

In pattern-making studios working on technical outerwear for the luxury segment, the workflow that produces the best results is this:

1. Base construction in CAD: the pattern starts in software, with the base size measurements and the structure of the garment's parts. This phase is fast and reversible.
2. Print and flat check: before sewing, the pattern is printed at 1:1 scale and checked on the table: do the assembly curves match? Are the symmetries correct? Are the allowances where they should be?
3. Prototype in substitute fabric: a first prototype is sewn in a fabric approximating the weight and behaviour of the final material. The final fabric is not needed — the right mass and stiffness are.
4. Fitting on mannequin or fit model: the prototype is worn and analysed. This is where the problems CAD had not predicted emerge: the falling shoulder, the tight armhole, the gaping collar.
5. Correction in CAD and repetition: corrections are carried back into the digital file and the prototype is cut again. For technical outerwear, this cycle typically repeats 2-3 times before final approval.
6. Grading and marker making in CAD: only after the base size is approved do grading and nesting proceed, both fully digital.

In this workflow, digital and physical are not in competition: they are different tools for different phases. Removing the physical phase to speed up the process tends to push problems forward in time, where they become more expensive to fix.

A note on technical heritage

There is an aspect of the physical pattern that is often overlooked: its value as an archive. Many maisons have drawers of historical patterns representing decades of fine-tuning a garment. These physical archives are hard to digitise (they require careful scanning and file clean-up) but contain technical information that the CAD files of later versions do not always preserve. Some maisons bring us these archives for re-edition projects or heritage capsules: the starting point is the physical pattern, and digital comes after.

Frequently asked questions

Can a down jacket be fitted in 3D?

Not for technical garments. A down-filled piece with 16 channels, 90 g/m² of fill, a ripstop nylon shell and a thermal lining behaves in ways no drape-simulation software reproduces faithfully: the volume of the fill changes the relationships between parts, technical fabrics vary in stretch and stiffness by grain direction, and fill compression only emerges on the mannequin or on a fit model. Physical fitting on a sewn prototype remains the only way to validate the pattern before grading.

What is CAD actually good for in pattern making?

For development and production. On a 7-size outerwear piece, manual grading on paper took days; in CAD a well-set-up grading runs in a few hours with repeatable, verifiable results. Marker making software reduces fabric waste in ways no one could replicate by hand at scale. The digital file can also be emailed to a manufacturer in any country without risk of deterioration or loss of the physical pattern.