GSM is a proxy: what weight does and does not tell you

Grams per square metre is popular because it is comparable. Every sheet has it, it is a single number, and bigger feels sturdier. It is also the number a supplier can most easily move without changing anything you care about.

What does gsm actually tell you?

Mass per unit area, measured under a published method — ASTM D-3776 on these sheets. That is all. It is the product of fibre density, yarn thickness and how tightly the structure is packed, and those three can trade against each other freely.

It correlates loosely with drape, with opacity and with cost, because more material usually means more of all three. It correlates much less well with strength, and barely at all with dimensional stability.

Is a heavier fabric stronger?

Not reliably. Epic's range, sorted by weight:

Fabric	Weight	Thickness	Breaking strength (L / W)
Rio	175 g/m²	0.22 mm	670 N / 662 N
Halo	180 g/m²	—	710 N / 465 N
Lumex	195 g/m²	0.23 mm	921 N / 984 N
Tesa	230 g/m²	0.55 mm	717 N / 1050 N
Aureline	260 g/m²	0.57 mm	388 N / 1107 N
Eos	260 g/m²	0.43 mm	407 N / 804 N

Lumex is the third lightest, the thinnest of the five that publish a thickness, and the strongest in the length direction of all six — 921 N, where Aureline at 260 g/m² and more than twice the thickness manages 388 N. Aureline is not a weak fabric; it is a fabric whose strength is overwhelmingly in the width direction, at 1107 N. But if you picked it over Lumex on the basis that 260 beats 195, and the load in your application runs along the roll, you chose a material less than half as strong for the job.

That is the whole argument against gsm in one row.

What should I compare instead?

Whichever two or three properties your application actually turns on:

• Opacity — compare the light shading or transmission figure directly. Weight is a guess at this; the sheet gives you the measurement.
• Strength, in the right axis — breaking and tearing strength, along the axis that will take the load. The two figures are often very different.
• Dimensional stability — shrinkage and its tolerance band, if the graphic goes into a frame.
• Thickness — for a SEG bead, for how the material sits in a groove, and for diffusion in a shallow light box. A separate number from weight, and more relevant to both.
• Hand and drape — which no number captures, and which is the one thing you genuinely need a sample for.

When does gsm genuinely matter?

Three cases.

Shipping and handling. A 50 m roll at 260 g/m² against one at 175 g/m² is a real difference in freight cost and in what one person can carry. On a large order this is not trivial.

Hanging weight. A large suspended graphic's own mass is a structural load, and the rigging has limits. Here gsm is the number you want, directly.

Perceived quality. Worth saying plainly: a client who handles a sample will read heavier as better, whatever the test data says. If the job is sold on the feel of the material in a meeting, that is a real constraint even though it is not a technical one.

Why is gsm quoted with a tolerance?

Because it varies, and the tolerance tells you by how much. Most of these are held to ±5 g against the target.

Watch where the result sits inside the band rather than only whether it passes. Lumex is specified at 200 g/m² ±5 g and measures 195 — on spec, at the bottom edge. Its thickness sits at the bottom edge of its band too, 0.23 mm against 0.28 mm ±0.05. Both are legitimate pass results, and together they tell you something the pass/fail column does not: this is a material running light and thin within its specification, which is consistent with the strength-to-weight it delivers and worth knowing if you are matching it to an earlier lot.

Every figure here is from our own measured test data with the method named. Full sheets are on each product page, and the comparison table on this site puts all six side by side on the properties that matter rather than on weight alone. Ask for sample yardage for the part that no number covers.