11 sources, every one cited inline. Your loops are not falling out because your stitches are loose. They are falling out because the weave is too coarse to grip them.
Material matching, loop depth and three printable patterns
Because the backing is too coarse for the needle. Punch needle holds without knots by friction alone: the woven threads close around the yarn and grip it. Oxford specifies 12–14 holes per inch for its Regular point and warns that 6–7-hole cloth will not hold loops. Tightening your stitches cannot replace the grip a correct weave supplies.
Written by Abdelmounim El Moussaddar · published by TopDigitalKnowledge · 35 pages · 4 illustrations
What everybody says
If the loops keep pulling out, punch them deeper or pack the rows more tightly together.
What the evidence says
Packing rows more tightly is the opposite of the fix and causes the second classic fault. Oxford describes rows packed too closely buckling the cloth and crowding neighbouring loops — so filling a visible gap by adding more and more stitches makes the surface worse, not better. Loops stay in because the weave grips them, which is a property of the cloth and the needle together. DMC explains that fabric grip and stitch density are what hold punched yarn without a knot. If the grip is not there, the answer is the right backing and the right yarn for that point size, not more force.
Either face can be displayed, but you have to choose before you trace the design.
Punching forms flat stitches on the side you work from and loops on the opposite side. DMC shows both faces and notes that either can be shown. This book displays the loop face, which means you are working from the back of the finished piece.
That has one practical consequence that catches everyone once. To reproduce a template exactly as drawn on the loop face, you must mirror it before tracing onto the working face — DMC explains this reversal for loop-face designs. The three patterns here contain no lettering, so an unmirrored version is a usable variation rather than a mistake, but with any design that reads as text or asymmetric imagery the reversal is not optional.
Trace boundaries only. Drawing filling rows across the final cloth leaves marks that the loops will not reliably cover.
Both follow from the needle, not from preference.
For the Oxford Regular point this guide uses punch-needle monk’s cloth at 12–14 holes per inch, and Oxford warns that much coarser 6–7-hole cloth will not hold the loops properly. Confirming compatibility with the supplier is worth doing, since “monk’s cloth” is sold in several weaves and only some suit this tool.
Yarn follows the same rule. Oxford distinguishes the rug-weight yarn suited to its Regular point from the finer yarn used with a Fine point, and the guide starts you on smooth, plied wool because a textured or loosely spun yarn hides whether the fault is your technique or your materials.
The perimeter matters too. Secure the raw edge against fraying with an appropriate sewn finish before stretching, as Whole Punching describes, and keep that finished perimeter outside the punching area.
For punching, usually not — it will not hold the tension.
Whole Punching distinguishes no-slip gripper hoops from ordinary wooden hoops, which may need extra measures to keep tension while you punch. Punching is a repeated downward force into a stretched cloth, and a hoop that creeps a little with each stitch produces exactly the loose, uneven surface a beginner blames on their hand.
The workflow here uses a suitable locking or no-slip working hoop for the punching, and transfers the finished panel to a wooden display hoop afterwards. The final mounting adapts a running-stitch gathering method used for hoop-framed embroidery, which is why the patterns include an unpunched clamping border.
That border is a design constraint, not a margin for error: this is a small decorative panel, not a working rug, and the book is explicit about the difference.
Fully, every time, and without lifting between stitches.
With the Oxford channel facing the direction of travel, punch until the handle meets the cloth, then bring the tip back to the surface and move to the next stitch without lifting high. Oxford describes this full-depth, surface-skimming motion, and the two halves matter equally: partial depth gives short loops, and lifting the needle clear between stitches pulls the previous loop back out.
Loop height is set by the tool, not by how hard you push. Oxford supplies stitch-gauge recommendations for the needle and yarn combination, and the guide asks you to take the gauge from your actual tool instructions and confirm it on the practice sample. The two-inch bar printed on the pattern sheets is a printer calibration check, not a stitch gauge — a distinction worth making because the two look similar on the page.
To change direction, stop with the needle inserted and rotate the hoop before continuing, keeping the channel aligned with the new direction of travel.
Usually not. Whole Punching notes that the woven threads of its monk’s cloth can be gently moved back into place after stitches are removed.
What matters is what you do next. The guide’s troubleshooting sequence follows Oxford’s: inspect the failed area, correct one factor, then make a short test before committing to the whole piece again. Changing the yarn, the tension and the depth together teaches you nothing, because a success cannot be attributed and a second failure cannot be diagnosed.
After removing stitches, ease the weave back, re-tension the hoop and punch a small test patch with the corrected setup. If that holds, continue; if it does not, the remaining variable is usually the backing.
Starts and finishes need no knots. DMC explains that fabric grip and stitch density hold the yarn, so for a colour change you cut the yarn, bring the tail through to the loop side and trim it to loop height after tidying.
The measurements in this guide are tied to the Oxford Regular point, because gauge, threading and backing requirements are properties of a particular tool. Another tool can be used, but take its gauge, its threading method and its recommended cloth from its own instructions — a closed-barrel needle, for example, may need a wire threader and should not be forced into the Oxford threading path.
Three printable hoop designs using broad regions rather than fine detail: a pebble arrangement, a banded composition and a texture study. They are drawn at actual size with a calibration bar so you can confirm your printer has not scaled them.
Related but not the same tool or motion, and this guide does not cover rug making. What is built here is a small decorative hoop panel with an unpunched clamping border. The book says plainly that the finish is not suitable for a piece that will be walked on.
Rows packed too closely. Oxford describes this directly: over-packed rows buckle the cloth and crowd neighbouring loops. It is counter-intuitive because the instinct on seeing a gap is to add stitches, which makes it worse. Space the rows to the gauge and re-test.
No. The four interior illustrations and the cover are illustrative images. They show composition and materials rather than documented results from a tested piece.
No. The manufacturer publishes free video demonstrations and the guide links to them as optional external references, but no video is produced by or included with this download.
Eleven, all cited inline and all linked: four Oxford pages covering FAQ, threading, technique and mistakes, four DMC technique guides, two from a monk’s cloth supplier and one hoop-framing tutorial. These are manufacturer and supplier instructions rather than research literature, which is the right authority for figures that describe a specific tool.
The first punch needle project usually fails in one of two ways: the loops pull straight back out, or the cloth buckles into a ridge that no amount of extra stitching flattens. Both are material matching problems, and both are decided before the first stitch. Oxford recommends a backing of 12–14 holes per inch for its Regular point and warns that much coarser 6–7-hole cloth will not hold the loops properly — so a tool, a yarn weight and a weave are one decision rather than three. 35 pages, three printable hoop patterns at actual size, a troubleshooting sequence that changes one factor at a time, and 11 linked references from the tool and thread makers themselves.