Full Canvas Is a Bet on the Cloth

I am not going to tell you which shop this was, because what matters is what happened on their fitting room floor. A jacket came back to me last spring that should never have gone out as full canvas. The cloth was a floppy, high-twist tropical wool, barely two hundred grams, the kind that folds beautifully over a hanger and does almost nothing once a body is inside it. Somewhere upstream, a full canvas had gone in because that is what the retailer told the client he was paying extra for. The finished jacket did not drape any better than a half canvas version would have. It draped like the cloth, because cloth that light drapes like itself no matter what you sew inside it. The retailer had already explained, in his own fitting room, why the extra money was worth it. Then the client asked why the jacket did not feel any different from the one hanging next to it, and there was no good answer standing there.

That is the failure this piece is about, and it has almost nothing to do with anyone's stitching. The canvas was cut correctly and pad-stitched correctly. The mistake was deciding to use it at all.

What the floating canvas does

A layer of stiffened horsehair and wool sits between the cloth and the lining, hand pad-stitched rather than glued, and in a full canvas jacket it floats free through the whole front instead of being fixed to the cloth at every point. What matters is what happens after the first wearing. That floating layer moves slightly against the cloth every time the jacket goes on, and over months, with body heat, the front settles into the shape of the person wearing it, which is what people mean when they say a good suit gets better with age. A fused, glued-down front cannot do this. Nothing moves, so nothing settles.

If you want the full anatomy of all three constructions, the comparison table and the ten-second pinch test to identify any jacket on a rail, we wrote that up separately in the construction guide. What follows here is the question that page does not answer: when the extra hand work is worth paying for.

The bet is on the cloth

This gets skipped in most conversations about construction, including some we have had with our own retail partners before either side knew better. Full canvas does not sit above half canvas the way a bigger engine sits above a smaller one. It is a bet placed on one specific bolt of cloth, and the cloth decides whether you win it.

The settling I described only happens if the cloth has something to settle against. Horsehair canvas has body of its own and wants to hold a shape. If the cloth on top also has weight, and a weave tight enough to hold its own structure, the two argue with each other in exactly the way that produces a natural roll and a chest that keeps its line for years. If the cloth is limp, the canvas is arguing with nothing. It rides along inside a shell that falls the way light cloth falls, and the years of settling that are supposed to justify the price never happen, because there was never enough resistance for anything to settle against.

Weight in grams per metre tells you which side of that line a bolt sits on, more than fibre content or the name on the bunch. Of the 822 fabrics in our own books with a stated weight, 597 sit between 240 and 300 grams, and that band is roughly where full canvas starts earning its keep: enough body for the canvas to press against, not so much that the jacket feels like armour. Below 240, running down to our lightest tropical wools and linens near 98 grams, a floating canvas has nothing firm to work with, and the roll you are paying hand labour for will not show up however well it is stitched. Above 340 the case flips. That is where fusing struggles to bond cleanly through a dense weave, and a badly fused heavy jacket goes dead and board-like, sometimes delaminating outright. Canvas sidesteps the problem by never depending on glue at all.

So the question a buyer should be asking is never full canvas or half canvas. It is what the cloth weighs and whether it has anything for the canvas to hold onto, which is the same discipline as asking for the bunch instead of the country: a specific, checkable number instead of a reassuring word.

The thirty-second test

You do not need a laboratory for this, and you should not need to trust a swatch card's adjectives either. Ask for the weight in grams per metre first, always. Then put your hands on the cloth. Let a cut length fall across your forearm. Cloth with enough body pushes back and holds a soft curve. Cloth without it collapses flat and stays flat. That one motion tells you more about whether a bolt deserves full canvas than any description on a swatch card will.

A tight, dense worsted or a flannel with real heft usually passes both tests, the number and the hand feel agreeing. A loose, airy plain weave, the kind chosen deliberately for hot climates, will often fail both, and there is nothing wrong with the cloth when it does. It is doing exactly what it was woven to do, which is breathe and stay light, and that job pulls in a different direction from full canvas's job.

Our own range spans that gap on purpose, from featherweight tropical wools through to heavier flannels and worsteds with real body, and the weight of every line in it is published rather than implied, so you can run this test against a swatch card before committing a retail price to it.

That test also hands you the sentence to say to a client, which is most of why I am writing this. "This cloth has the weight to earn full canvas" is true on some of what you sell. "Save your money on this one, half canvas will look and move the same on something this light" is true on the rest. A client trusts a shop that says the second sentence out loud far more than one that only ever says the first.

The cost of getting it backwards

Go back to the jacket I opened with. The stitching was clean and the canvas well shaped, so the only mistake was upstream, in the decision to spend hand labour on a cloth that could never repay it, and nobody caught it until the client already had the jacket on their back.

Fixing it meant a full remake, not an alteration, because you cannot unpick a floating canvas without damaging the shell around it. That meant recutting, resewing and shipping the garment back to production and the replacement back out again, stacked on top of whatever the original 14 to 20 working days had already cost. Get the construction decision wrong and that clock runs twice, and both times a courier or a container is doing the running while the client waits, and mentions it to somebody else.

Where I will argue against it

The direct version, which a supplier trying to sell you more construction has no reason to say: if lightweight summer suiting is most of what leaves your shop, do not pay extra for full canvas, and do not let anyone talk you into charging for it either. On a cloth that light, full canvas sells your client a word rather than extra performance, and the hand labour behind that word is wasted on cloth that cannot use it.

We make fused, half canvas and full canvas ourselves, under the same roof, so telling you this costs us nothing on the sale in front of you. It is why I can say it flat out instead of finding a softer way to keep a client from walking away feeling short-changed.

Our own full canvas starts at 254 euros ex-works, and I would rather you did not compare that number to anyone else's price sheet until you have the cloth's weight sitting next to it. A cheap full canvas quote on a cloth that cannot use it is the same mistake with better economics for whoever sold it to you. There is no minimum order beyond a single piece, so you never have to guess at this across a whole rail: run the weight test above on one jacket, in the cloth you sell the most of, before deciding whether full canvas belongs on your price list. It is the same discipline as picking the one number worth building the decision on instead of the percentage on the tag; here that number is grams per metre.

Scaling full canvas

The question we get asked most about full canvas is whether it can be produced at real volume without turning into a lottery. Yes, but the difficulty sits somewhere most people do not expect. Pad stitching a floating canvas is old, well understood work, and any competent workshop can teach the sequence in an afternoon. The difficulty is that it stays a hand's work indefinitely, and hands are not identical.

Give the same cut canvas and the same block to two well trained people and you will get two lapels with a slightly different roll, because the angle of the needle, the spacing between rows, and the tension pulled into each stitch are all small human decisions repeated hundreds of times per jacket. None of those decisions is wrong on its own. Stacked across a production run, they are what separate a range that reads as one consistent product from a range that reads as several different tailors having an argument.

More inspection at the end does not fix it, because by then the labour is already spent. Consistency in hand work like this comes from keeping it with the same trained hands rather than rotating it, and from checking new work against a fixed reference before it ships. Any workshop attempting this at volume has to do the same; it is what the craft demands, not a shortcut particular to us. Our own cutting and canvas work happen in our own atelier in Shanghai, by our own tailors, in cloth that is European or Australian wool.

If you are weighing whether to build this capability behind your own shop or buy it from somebody who already has, that is the comparison laid out in the guide to production for tailor shops: what a floor like this costs to run yourself, against what it costs to buy it from a workshop that already does this work daily and has already absorbed the cost of getting consistent at it.

One thing to get in writing

One more thing belongs here, and it is about what happens when any construction decision goes wrong, full canvas or otherwise. Before you promise a client a remake or a return on anything, get your supplier's terms in writing, not a verbal reassurance in a showroom. What counts as a construction fault versus a fit issue. Who pays freight both ways. How long a remake takes once queued behind everyone else's order. Every answer differs supplier to supplier, and the moment you need one is the worst moment to discover you never asked.

This belongs on the same list as the other questions worth putting to any factory before money moves, which we wrote out fully elsewhere. I am not going to promise you our own terms in the middle of an article about somebody else's mistake. Ask us directly, ask it in writing, and judge the answer the way you would judge anyone else's, by how specific it is prepared to be rather than how reassuring it sounds.

---

Send us the composition and the weight of whatever is on your own shelf and we will tell you straight whether it earns full canvas, including when the answer is to save the upgrade for a heavier bolt. Talk to the atelier and we will look at the cloth itself, not the label on it.