A woven fabric and a knitted fabric are not two versions of the same material. They are held together by different geometry, and that geometry decides how the cloth moves. Long before either reaches a sewing machine, it has to be rested, spread and cut. At every one of those steps the two behave differently, and a large share of the sizing and fit problems a buyer finds later were decided there.
The two structures, and why the difference matters
A woven fabric is two sets of yarns crossing at a right angle. The warp runs down the length of the roll and is held under tension on the loom. The weft crosses it. Both sets of yarns are more or less straight, so pulling along either one pulls directly on the yarn itself. The yarn barely extends, so the fabric barely extends. Where a woven does give is on the diagonal, because the diagonal lets the crossing yarns hinge and shear against each other instead of stretching.
A knitted fabric is yarn bent into loops. In a weft knit — jersey, tadyang, magkabit — one yarn travels across the fabric making a row of loops, and the next row hangs from it. Rows across are courses; columns down are wales. When you pull a knit, the yarn does not need to stretch. The loops simply change shape, and yarn moves from one part of a loop into another. That is why knits extend so much further than wovens, and why they extend in every direction rather than mainly on the diagonal.
Warp knits are a different family again. Many yarns run down the fabric at once, each looping into neighbouring wales. Tricot and raschel fabrics are made this way, and they are noticeably more stable and less extensible than weft knits.
One consequence runs through everything below. A woven’s dimensions are held by where the yarns sit and by friction at the crossings, and those are hard to shift. A knit’s dimensions are held by loop shape, and loop shape is easy to change and slow to return.
Why cloth has to rest before it is spread
Fabric does not arrive in its natural state. It leaves dyeing and finishing hot and dry, having been pulled lengthwise through machine after machine, then wound tightly onto a roll. At that moment it is longer and narrower than it wants to be.
Laid out flat in the cutting room, it takes moisture back from the air and the structure relaxes. It shortens in length and widens. This is not damage. It is the fabric returning to the size it will actually hold.
If it is spread and cut before that happens, the panels are cut at the stretched dimension and then contract afterwards. Nothing looks wrong at the cutting table. It appears later as garments that measure short, and as size variation between bundles cut from different rolls that had been sitting for different lengths of time.
So knits are unrolled and left to rest flat or loosely folded before spreading. A minimum of one full day is ordinary practice, and fabrics containing elastane are usually given longer. Fabric that has been tightly rolled for a long time, or shipped in a container, holds more tension and needs more time. Wovens benefit from the same rest, but they hold less of the tension in the first place and give it back faster, so the penalty for rushing is smaller.
Spreading and cutting: tension turns into a size problem
Whatever tension is present in the lay is stored in the fabric and released after the knife has passed. A lay spread tight contracts, and the panels come out smaller than the marker. A lay spread slack sits in ridges and bubbles, and the panels come out larger. Both are sizing faults, in opposite directions, and neither is visible while it is happening.
Worse, the tension can differ from ply to ply. That produces panels of different sizes inside a single cut, which is far harder to trace than a whole cut running small. This is why knits are spread as close to tension-free as the equipment allows, and why the spread lay is commonly left to settle on the table before cutting.
Lay height is also a live question for knits in a way it is not for a stable poplin. Knit plies are springy and slip against each other easily, and the taller the lay, the more scope there is for plies to shift and for the blade to wander away from vertical.
Two more knit-specific nuisances show up here. Single jersey curls at its edges, because the face and the back of the loop are pulling unequally and the structure is not balanced through its thickness. Curled edges reduce the usable width and drag the top ply out of position. And knit width varies from roll to roll, and within a roll, more than woven width does — so the marker has to be planned to the narrowest usable width. Where a knit has been finished in tube form, the folded edges can carry crease lines that never fully come out, and panels cannot be placed across them.
After cutting, the edges behave differently too. A woven frays. A knit does not fray, but it curls, and a broken loop can run down a wale.
Spirality: the knit fault that arrives after the wash
In a well-behaved knit the wales run at a right angle to the courses. In plain single jersey they very often do not. They lean. That lean is spiralidad.
The cause is torque in the yarn. A singles yarn is twisted, and that twist is always trying to unwind. In a balanced structure the loops hold each other square. Plain single jersey is not balanced, so the loops give way and settle at an angle. The direction of the lean is related to the direction of the yarn’s twist and the direction the machine ran.
The reason it matters commercially is the delay. The fabric looks acceptable on the table, the panels are cut straight, the garment passes inspection. Then it is washed, the fabric relaxes further, the lean increases, and the side seam of a t-shirt walks round the body towards the front.
Cutting cannot fix it. A straight panel cut from a fabric whose wales lean is a straight panel with leaning wales. It is reduced upstream instead: plied yarns and yarns with less lively twist lean less, balanced structures such as rib and interlock lean less than plain jersey, and setting with heat or steam can square the wales — though that holds far better on thermoplastic synthetic fibres than on cotton.
Skew and bow: the woven equivalents, and they are not the same thing
Both describe the weft sitting somewhere it should not, relative to the warp. They are different geometries and they are routinely confused.
- Skew is angular. The weft runs in a straight line, but at an angle instead of square, so one selvedge is ahead of the other.
- Bow is curved. The weft sags or arcs across the width, typically because the centre of the fabric and its edges moved at different speeds or under different tension through finishing.
On a check, a stripe or a placed print, both are obvious at once — the pattern visibly refuses to sit straight. On a plain dyed fabric neither is visible on the cutting table at all. That is the dangerous case, because the pattern piece carries a grain line that assumes the weft is square to the warp. If it is not, a panel cut carefully “on grain” is off grain in the actual yarns, and it will behave like an off-grain panel no matter how accurate the cutting was.
Grain and bias on a woven
Straight grain follows the warp. Cross grain follows the weft. Bias is the diagonal, and true bias, at forty-five degrees to both, is where a woven has the most extension and the least willingness to come back.
The warp is usually the stronger and more stable direction — it was made to survive the loom. The weft gives more. So a panel cut a few degrees off grain has one edge sitting closer to the bias than the other. Under the garment’s own weight, and after a wash, those two edges extend and relax by different amounts. The garment twists on the body and the hem drops unevenly on one side.
This is the classic delayed defect. It is not visible flat, it often survives final inspection, and it appears on the customer. Bias cutting done deliberately, for drape, is a separate matter — but then it is a design decision, it has to be marked on the pattern, and mirrored pieces have to be handled so that both sides hang the same way.
Growth and recovery are two different properties
A single “mag-inat” figure hides the two questions that actually matter.
- Recovery asks how much of the extension the fabric gives back once the load is removed, and how quickly.
- Growth asks how much it never gives back — the permanent extension left behind after the fabric has been held stretched for a period and then released.
These come apart in practice. A fabric can snap back well from a quick pull and still take a permanent set after being held extended for a long time. A fabric can stretch a long way and recover poorly. Knowing only how far it stretches tells you nothing about either.
Both show up before sewing. A knit with weak recovery that gets stretched during spreading does not come back, and the panels are permanently oversized. And in wear, growth is what produces bagged knees and dropped elbows on a garment that measured correctly when it shipped.
Why knit tolerances are wider
A woven can be held to a tight measurement because the thing being measured barely moves. A knit moves under the tape. How it was laid out, and how much the person measuring leaned on it, changes the reading.
So knit garments are normally given wider tolerance at each point of measure than wovens. As a rough shape of the difference: a stable woven shirting might be held to about half a centimetre at a point of measure, while a jersey version of the same garment is more often given something nearer a centimetre and a half.
Length usually gets more room than width, because knits generally shrink more along their length than across it — the loops shorten as the fabric relaxes and widens. It is worth setting length and width tolerances separately rather than applying one figure everywhere.
A wider tolerance is not a slacker standard. It is an honest statement of what the material can hold. A tight tolerance written onto a knit specification does not make the garment more consistent; it makes the inspection report less useful, because the failures it generates are noise from the material rather than signal about the work.
What to specify differently
For a knit
- Fabric weight with a stated tolerance, since weight is the primary handle on a knit’s construction.
- Shrinkage in length and width as separate figures, each tied to a named wash test method and a stated number of cycles.
- A maximum spirality measured pagkatapos paglalaba, not on the roll. Spirality measured on the roll answers the wrong question.
- Mag-stretch, recovery and growth as three separate requirements. One stretch figure is not a specification.
- Whether the fabric is required open width or may be supplied in tube form.
- Relaxation before cutting, written into the tech pack rather than assumed.
- Point-of-measure tolerances wide enough to be real, and wider in length than in width.
For a woven
- A grain line marked on every pattern piece, including facings, pockets and small parts, where it is most often left off.
- Maximum bow and maximum skew stated as two separate limits, because they are two separate faults with two separate corrections.
- Warp and weft shrinkage separately, again against a named test method.
- Konstruksyon — ends, picks and yarn count — not weight alone. Two fabrics of identical weight can behave nothing alike.
- Any piece intended to be cut on the bias called out explicitly, with the mirroring specified.
For either construction, the same rule applies to every number in the specification: a shrinkage, spirality or growth figure with no test method attached is not a requirement. It is a hope, and it cannot be argued either way when the shipment is inspected.
Where this ends
Everything above is settled before the first stitch is formed. What happens at the machine — matching thread, needle and stitch to the fabric in front of it — is a separate set of decisions with its own failure modes, covered in Bakit Nabigo ang mga tahi: Katugmang Thread, Karayom at Tusok sa Tela.
The point worth carrying away is that most of these faults are invisible at the moment they are created. Tension in a lay, a leaning wale, a bowed weft, a panel a few degrees off grain — none of them look like anything on the cutting table. They surface in a fitting room or after a wash, by which time the cause is several weeks upstream. That is why the specification, rather than the inspection, is where they are actually controlled.
