Explainers

Patreon for Hardanger embroidery creators: how Kloster blocks function as structural anchors that must completely surround every fabric square before any threads are withdrawn, why the cutting sequence runs between Kloster blocks and never through them, the structural difference between woven bars and wrapped bars in the drawn thread grid, how dove’s eyes are constructed at open grid intersections, the evenweave counting requirement, and the Apple Tax on iOS-heavy traditional needlework Patreon audiences from November 2026

2026-09-17 · ~5,900 words

Hardanger embroidery tutorial videos show Kloster blocks being stitched and threads being withdrawn, but they rarely pause to explain why completing every surrounding Kloster block before cutting is not just a procedural preference but a structural requirement — or what exactly the distinction is between a cut between two blocks and a cut through a block. Kloster blocks are the structural anchors that hold cut fabric thread ends in place; without a complete ring of blocks around a square, cut threads fray and the drawn area distorts irreversibly. The correct cut runs between two facing Kloster blocks along the drawn thread channel, not under the block’s satin stitches. After withdrawal, woven bars and wrapped bars are structurally different in how the working thread passes through the remaining thread groups, producing different surface textures in the finished grid. Dove’s eyes are ring-shaped filling stitches worked by looping from bar to bar around an open square’s interior, which is a different construction from the spoke-and-spiral structure of a spider web filling. This post documents the mechanical and structural layer that most Hardanger instruction compresses, the sequence constraints that determine whether the drawn thread area holds or fails, and the Apple Tax that iOS-heavy traditional needlework audiences will impose on creator revenue from November 2026.

Kloster blocks as structural anchors: the five-over-four ratio, the block completion sequence, and why the drawn thread area fails when blocks are incomplete before cutting

Kloster blocks are the element in Hardanger embroidery that distinguishes drawn thread work from free embroidery. In free embroidery, the fabric is intact throughout the work and each stitch is independent of the fabric structure beneath it. In Hardanger, the fabric structure is deliberately cut and removed in designated areas, and the surrounding embroidery must be in place to prevent the cut edges from fraying before that removal happens. Kloster blocks are that surrounding embroidery. They are satin stitch groups — five parallel satin stitches, each passing over four adjacent fabric threads — positioned at the edges of every square that will be drawn.

The structural logic of a Kloster block is straightforward once you understand what happens to a cut fabric thread. When a fabric thread is cut at a point within the weave, the severed end has no inherent resistance to fraying. The interlocking of warp and weft threads keeps the weave intact as long as both threads are continuous, but a cut removes the continuity of one thread and allows adjacent threads to shift. If a pearl cotton satin stitch lies over the cut end of the fabric thread — passing over it, holding it flat against the fabric surface — the satin stitch physically grips the cut end and prevents it from migrating. The block holds the cut thread end against the fabric even after the thread in the drawn area is removed.

This holding function only works if the block is complete. A block that is half-worked — two or three satin stitches rather than all five — covers only a portion of the fabric threads at the block location and leaves uncovered fabric threads at the cut edge free to shift. A block entirely absent at one side of a square leaves the cut ends of the withdrawn threads without any anchor on that side. The drawn area distorts toward the absent block as adjacent fabric threads pull inward to fill the structural gap. This distortion is visible as an asymmetric, ragged appearance in the drawn thread channels after withdrawal and cannot be corrected by subsequent bar construction or pressing.

The five-over-four ratio that defines every Kloster block is not an aesthetic preference. It derives from the fabric count requirement. Hardanger fabric is a 22-count evenweave: 22 threads per inch in both the warp and weft directions. Four threads at 22 count equal approximately 4/22 of an inch — roughly 4.6 mm. A satin stitch of pearl cotton number 5 thread has a diameter of approximately 0.9 mm, and five adjacent satin stitches without spacing produce a total block width of approximately 4.5 mm. At 22-count Hardanger fabric, therefore, a block of five satin stitches over four threads is approximately square — its length along the stitching direction (four thread widths) matches its width perpendicular to the stitching direction (five stitch widths). The square block is the design unit that makes the Hardanger grid work: adjacent blocks that share corners produce perfect right-angle transitions and symmetrical drawn thread channels. If the fabric count is different — 25-count or 28-count linen — the same five-over-four block will not be square, and the visual effect of the grid will be rectangular rather than square, distorting the pattern geometry.

For Patreon documentation, the correct instructions are explicit about the block-first sequence. Every pattern should state: before cutting any fabric threads in any drawn square, verify that all four Kloster blocks sharing edges with that square are complete with all five satin stitches in place. For a pattern with multiple adjacent drawn squares, all Kloster blocks for all drawn squares should be complete before any withdrawal begins. The temptation to work one square to completion (blocks, then cut, then bars, then fillings) before moving to the next is natural but structurally incorrect — the threads that frame two adjacent drawn squares share a Kloster block, and cutting one square before the adjacent square’s surrounding blocks are complete can remove threads that the adjacent square will later need intact.

The cutting and withdrawal sequence: identifying the correct cut location between Kloster blocks, distinguishing block stitches from fabric threads, and the blunt-needle withdrawal technique

Once all Kloster blocks surrounding the drawn areas are complete, the thread withdrawal sequence begins. The physical act of cutting a fabric thread in Hardanger is brief but irreversible, and the correct location for the cut must be understood precisely before scissors or a seam ripper contacts the fabric.

Two Kloster blocks that face each other across a drawn square share an empty channel between them. This channel is the path through which the to-be-withdrawn fabric threads run. The threads in this channel — the threads that will be cut and removed — run in one direction. The Kloster block’s satin stitches run perpendicular to the to-be-withdrawn threads. Visualizing the relationship: if two Kloster blocks face north and south across a drawn square, the satin stitches of each block run east-west, and the to-be-withdrawn threads run north-south, passing underneath each block’s satin stitches at each block location before entering the open drawn channel between the blocks.

The correct cut is made at the inner edge of each Kloster block — the edge of the block that faces the drawn channel. The scissors blade, angled to contact the fabric thread at that inner edge, cuts the fabric thread at the precise point where it exits from beneath the last satin stitch of the block into the open drawn channel. The cut severs the fabric thread immediately beside the innermost satin stitch of the block without touching that satin stitch. Because the innermost satin stitch of the block passes over and holds the fabric thread at the cut point, the cut end of the fabric thread is held flat immediately after the cut.

The destructive cut is one that goes underneath the block’s satin stitches — placing the scissors blade under the pearl cotton threads and cutting through both the fabric thread and the satin stitch together. This cuts the block’s embroidery thread, severing the satin stitch at that position and removing the holding capacity of the block at the cut edge. A block with severed satin stitches will not hold the cut fabric thread end reliably; the stitches will loosen from the cut point outward and the block may eventually unravel if the cut is at a structurally critical position.

Distinguishing block satin stitches from fabric threads requires practice and magnification at first. Pearl cotton number 5 has a slight sheen and a smooth, plied surface; it lies over the fabric threads in a regular parallel array at each Kloster block position. The underlying fabric threads are thinner, have a matte weave texture, and run in continuous straight lines through the fabric in both directions. Under a 3× or 4× loupe, the distinction is clear: the pearl cotton satin stitches are visibly thicker, shinier, and parallel to each other; the fabric threads are thinner, set at right angles to the satin stitches, and run continuously past the block into the drawn channel. Beginners are advised to make the first several cuts with magnification until the hand reliably finds the gap between the innermost satin stitch and the first fabric thread of the drawn channel.

After cutting fabric threads in one direction at both block edges of a drawn square, the process is repeated for the perpendicular direction: the two Kloster blocks facing each other on the other pair of sides are each cut at their inner edges. With all four sets of cuts complete, the fabric threads within the square have been severed from the continuous fabric in both directions.

Withdrawal uses a blunt tapestry needle, not the sharp crewel or chenille needle used for Kloster blocks. A blunt needle tip can be slipped under individual fabric threads in the drawn area and used to ease them out of the weave without risk of piercing adjacent threads. Each cut fabric thread is teased free individually: the needle tip slides under the thread at one cut end, the thread is lifted slightly from the weave surface, and then drawn out along its own axis — parallel to the thread direction, not at an angle to it — by pulling with the needle tip or with fine tweezers. Drawing perpendicular to the thread axis attempts to pull it through the weave structure rather than out of it, producing resistance and potential fabric distortion. Withdrawn threads that resist despite correct direction indicate an incomplete cut at one end: the thread is still attached to the fabric beyond the cut point. Return to the cut location with the scissors rather than forcing the withdrawal.

Woven bars and wrapped bars: construction mechanics, structural differences, and the visual distinction in the finished drawn thread grid

After threads are withdrawn from the drawn squares, the remaining fabric threads — those running in the direction that was not withdrawn — form loose groups within the drawn channels. In a standard Hardanger drawn square where threads are withdrawn in both directions, the four remaining fabric threads in each channel run parallel to each other, loosely connected to the fabric only at their ends where they enter the Kloster block at each side. These thread groups must be stabilized into bars before any filling stitches can be worked. Two bar-making methods are standard: woven bars and wrapped bars.

A woven bar is made by working the embroidery thread — pearl cotton number 8 at this stage, finer than the number 5 used for Kloster blocks — in a true weaving motion through the remaining thread group. The pearl cotton thread approaches the bar group from one side and the needle passes over the first two threads of the group and under the second two threads on the first pass, then back under the first two and over the second two on the return pass. This alternating over-under passage creates an interlaced structure: the working thread passes between the individual fabric threads of the group, not merely around the outside of the group. Each successive weaving pass is placed snugly against the previous pass, building up an interlaced woven bar from one end to the other. The resulting structure has the working thread genuinely interlocked with the individual threads of the group.

The visual result of a woven bar is a flat, firm bar with a woven texture visible on its surface under moderate magnification. When viewed from above, the working thread appears to alternately cross over and under the thread group, producing the characteristic over-under shadow pattern of woven structure. Woven bars are slightly flatter and more compact than wrapped bars because the interlacement compresses the thread group more efficiently than external wrapping does. The bar’s width stays consistent from end to end because the weaving motion distributes the working thread evenly through the group cross-section.

A wrapped bar is made by passing the working thread around the exterior of the entire thread group in successive wraps. The needle approaches the group from one side, passes under the entire group (not between individual threads), returns over the top of the group to the starting side, then passes under again for the next wrap. The working thread makes a helical circuit around the four-thread bundle, gripping the exterior of the bundle without entering between any individual threads. Each successive wrap sits immediately beside the previous wrap.

The visual result of a wrapped bar is a rounded bar with a smooth exterior showing the helical compression of successive wraps. In raking light, the gentle diagonal shadow of overlapping wraps is visible along the bar length. Wrapped bars are slightly rounder in cross-section than woven bars because the wrapping motion allows the thread group to maintain its natural rounded bundle form rather than being compressed by an interlaced working thread. Wrapped bars are also slightly more flexible along their length because external wrapping does not lock the individual threads together as firmly as interlacement does.

For Patreon documentation, the bar type must be specified for each bar in the design — not just a single note at the start of the drawn thread work section. Some Hardanger designs mix woven and wrapped bars in the same piece to produce contrasting visual textures. The specification must also indicate the thread size: pearl cotton number 8 is standard for bars; pearl cotton number 12 is used in some designs for very fine filling stitches in supplementary surface detail. A diagram showing the weaving path for woven bars versus the circuit path for wrapped bars, illustrated with the four-thread group visible as individual thread lines, is the most effective Patreon documentation for this distinction because verbal description alone is insufficient for students who have not previously worked either technique.

Dove’s eye filling stitches and spider web filling stitches: ring-versus-wheel construction, the loop-to-bar attachment sequence, and tension management in the open grid

Dove’s eyes and spider webs are the two primary filling stitches worked in the open drawn squares of a Hardanger grid after bars are complete. Both are decorative elements that add visual interest to the open areas. They are constructed very differently and produce very different visual results: a dove’s eye is a ring, and a spider web is a wheel.

A dove’s eye is worked by carrying the working thread from bar to bar around the four sides of an open square’s interior, looping under each bar as it goes, and forming a ring in the center as the thread closes back to the starting bar. The procedure: the working thread is brought up from the fabric back at a point near one end of the first bar, typically at the bar’s center or at approximately one-third of the way along its length. The thread is then carried diagonally across the open square to an adjacent bar — not the opposite bar, but one of the two bars at 90 degrees to the first. At the adjacent bar, the needle is passed under the bar thread from one side to the other, picking up the bar thread without piercing the fabric, and the working thread is drawn through to form a loop around the bar. This loop-under-bar action is repeated at the next adjacent bar, and the next, proceeding around the square’s perimeter until all four bars have been looped.

When the working thread returns to the starting bar after looping the fourth bar, the needle passes through or under the starting loop — the first loop formed at the starting bar — and is drawn back through the fabric. Drawing the thread through the starting loop closes the ring: the four loop segments between bars contract to a circle or oval in the center of the open square, balanced by the equal tension at each of the four attachment points on the surrounding bars. The dove’s eye is complete. The finished form is a ring of thread floating in the center of the drawn square, attached at four points to the surrounding bars, with the ring diameter approximately half the width of the open square.

Tension is the variable that most affects dove’s eye quality. The working thread must be drawn through the starting loop with enough tension to close the ring flat but not so much tension that it pulls the bars inward. The test for correct tension is visual: after closing, the ring should be visible as a smooth loop floating freely in the open area, the bars should remain straight and parallel to their original positions, and the four attachment points on the bars should be at approximately equal distances from each bar’s midpoint. An over-tightened dove’s eye pulls the bars into a concave shape toward the ring center; an under-tightened dove’s eye lies in a loose, irregular loop that touches the fabric or one of the bars rather than floating freely in the center.

For Patreon documentation, the dove’s eye starting position matters because it determines the visual centering of the ring in the open square. Beginning at the center of a bar (rather than at one-third) produces a ring that is centered in the square and symmetrical from all four sides. Beginning off-center shifts the ring and changes the visible crossing angles of the diagonal carrying threads. Traditional Hardanger documentation specifies the starting position as a fraction of bar length (center, one-third, end), and this specification must be included in any Patreon pattern where the visual centering of the dove’s eye is part of the design effect.

A spider web filling stitch differs fundamentally in structure. Before the web can be worked, spoke threads must be laid across the open square from bar to bar to create the radial framework. Typically five spokes are used, arranged so that their center intersection point is approximately the center of the open square. The odd spoke count is required for the over-under weaving motion to work correctly: with an even number of spokes, the weaving thread would return to the same over-under relationship on each circuit, producing a striped structure rather than an interlaced one. With an odd number, the over-under sequence shifts by one position on each circuit, producing true interlacement.

After spokes are in place, the working thread is brought up as close to the spoke intersection as possible and begins spiraling outward, passing over one spoke and under the next on each circuit. The spiraling motion builds concentric weft rings over the spoke framework from the center outward, producing a dense, wheel-like filling that covers the spokes and fills the square center with a solid web of thread. The finished spider web is a compact disk of thread with visible radial spoke lines beneath the concentric weft surface. It fills the open square visually in a way that a dove’s eye, which is a ring, does not: a dove’s eye leaves most of the open square visibly open; a spider web covers the center of the square with a solid thread mass.

For Patreon patterns, the distinction between dove’s eye and spider web must be communicated with photographs or diagrams rather than only with text. Many students working from written Hardanger instructions have encountered only one filling type in their earlier practice and will not have a visual reference for the other. A single close-up photograph showing one dove’s eye and one spider web in adjacent open squares of the same piece communicates the visual contrast more efficiently than any written description.

The evenweave fabric requirement: thread count equality in both directions, the effect on Kloster block proportions, and thread size selection for each stage of the work

Hardanger embroidery is a counted needlework discipline. Every stitch in the work is positioned by counting fabric threads, not by measuring distance or by following a transferred design line. This makes the fabric structure the essential design medium: the pattern exists in the relationship between fabric threads and embroidery thread count. The counting requirement determines the fabric specification: Hardanger must be worked on evenweave fabric, and the most important property of evenweave is that the thread count is equal in both the warp and weft directions.

Equal thread count in both directions means that a block of four fabric threads in the warp direction occupies the same width as a block of four fabric threads in the weft direction. When the count is equal, a Kloster block of five satin stitches over four threads is the same length whether the satin stitches run east-west or north-south. Two Kloster blocks that meet at a corner — one running east-west and one running north-south — produce a perfect right-angle intersection, and the drawn square they enclose with two other blocks is a true square. The entire visual logic of the Hardanger grid depends on this right-angle precision: the geometric regularity of the pattern is the aesthetic content, and geometric regularity requires equal count.

If the fabric count is unequal — more threads per inch in the warp than the weft, which is common in many dress linens and tableware linens — the Kloster blocks in the two directions will be different sizes. Horizontal blocks will be taller than they are wide; vertical blocks will be wider than they are tall. The drawn squares will be rectangles rather than squares. The visual effect in the finished piece is that the pattern looks compressed in one direction and stretched in the other, and the geometric regularity that defines traditional Hardanger is absent.

Traditional Hardanger fabric is a 22-count cotton or cotton-linen blend woven specifically for Hardanger embroidery, with equal warp and weft count and equal thread diameter in both directions. It is available from specialty needlework suppliers in white, antique white, and a limited range of solid colors. The 22-count specification is the starting point for all standard Hardanger patterns. For Patreon documentation, the fabric specification must include thread count and confirm evenweave; a fabric listed only as “linen” or “cotton” without count and evenweave confirmation is insufficient for a Hardanger pattern and will produce student failures if the selected fabric turns out to have unequal count.

Thread size selection in Hardanger follows a two-size convention that matches thread diameter to the structural function at each stage. Pearl cotton number 5 is the standard thread for Kloster blocks and any surface embroidery (satin stitch motifs outside the drawn areas, buttonhole stitch borders). Pearl cotton number 5 has a diameter that fills the satin stitch coverage over four fabric threads without gapping between adjacent stitches at 22 count. Pearl cotton number 8 is the standard thread for drawn thread work: bars (woven or wrapped), dove’s eyes, spider webs, and any other filling stitches worked in the open grid. Pearl cotton number 8 is finer than number 5, appropriate for the intricate needle movement required in bar construction and filling stitch work in a relatively small open area.

The thread sizes are not interchangeable. Using number 8 for Kloster blocks produces blocks that are too fine to cover four fabric threads without gapping; each satin stitch leaves visible fabric thread between the stitches, and the block loses its structural compactness. Using number 5 for bars and fillings produces bars too thick to maneuver through the relatively narrow spaces in the drawn thread channels; the bar becomes stiff and the filling stitches in the open squares are visually heavy relative to the delicate character of the drawn thread work. Patreon patterns must specify thread sizes for each technique section separately: number 5 for all Kloster block and surface embroidery sections, number 8 for all drawn thread work sections.

Patreon documentation requirements for Hardanger: pattern grid format, stitch count specifications, and the sequencing problem in multi-drawn-square designs

Hardanger is among the most documentation-demanding embroidery disciplines for Patreon creators, because the irreversibility of thread withdrawal means that errors in documentation that lead students to work out of sequence cannot be corrected by unpicking stitches. Unlike surface embroidery errors, which are generally fixable by removing stitches and reworking them, a drawn thread area worked without complete surrounding Kloster blocks produces structural damage that requires cutting out the drawn area and re-starting on fresh fabric. Patreon patterns for Hardanger therefore need a higher level of explicitness about procedure sequence than patterns for most other embroidery disciplines.

The standard documentation format for Hardanger uses a grid chart where each grid square represents one fabric thread crossing. Kloster blocks appear on the chart as solid colored rectangles spanning five grid squares in one direction and four grid squares in the perpendicular direction, oriented at right angles to the adjacent drawn square. The chart clearly shows which blocks share edges with each drawn square, which blocks are shared between two adjacent drawn squares, and which positions are drawn (empty on the chart) versus surface-embroidered (marked with a stitch symbol). The chart must be supplemented by a numbered sequence guide: a list that specifies the order in which Kloster blocks should be worked, with a note indicating when all blocks for a given zone are complete and thread withdrawal can begin for that zone.

Some Hardanger designs include Kloster blocks that share sides with three or four drawn squares simultaneously — for example, a central block in a grid of nine drawn squares surrounded on all four sides by drawn area. These shared blocks require the sequence guide to be particularly explicit: the shared block must be completed before any of the surrounding drawn squares are cut, even if only one of those squares is in the current working zone. A partial guide that says “complete all blocks before cutting” without specifying shared-block completion is insufficient for a student working a complex multi-square design and may lead to premature cutting adjacent to an incomplete shared block.

The bar type for each drawn thread channel should be marked on the chart or sequence guide. For designs that use both woven and wrapped bars in different positions, this specification is essential: bars cannot be changed after they are worked without re-doing the entire channel. Filling stitch placement — which open squares receive dove’s eyes, which receive spider webs, and which are left open — should be marked on a separate filling chart rather than combined with the Kloster block chart, because the two charts require different reading strategies.

Apple Tax impact on Hardanger embroidery creator Patreons: iOS audience proportions, November 2026 revenue calculations, and the web-checkout route that avoids the commission

Hardanger embroidery has an active but specialized creator community that publishes on YouTube, Instagram, and Pinterest. The technique’s Norwegian heritage gives it a particular following in Scandinavian craft communities and among needlework practitioners who seek historically grounded techniques. The audience skews toward experienced adult crafters with an interest in counted needlework precision, and this demographic uses iOS devices at rates comparable to other counted embroidery communities.

YouTube hosts the primary long-form tutorial content for Hardanger: video sequences showing the complete process from Kloster block placement through thread withdrawal, bar construction, and filling stitches. The procedural, multi-stage nature of the technique requires the demonstration duration that YouTube format allows. Short-form content on Instagram serves the finished-piece showcase and close-up detail photography markets: images of completed drawn thread grids at close range, with dove’s eyes visible in sharp focus, perform strongly in the embroidery and needlework communities on that platform. Pinterest supports pattern discovery and historical reference: traditional Norwegian Hardanger pattern books, museum textile photographs, and contemporary design interpretations all generate search-driven traffic from users building their technique and project libraries.

The iOS proportion for Hardanger and traditional needlework audiences by platform: YouTube Hardanger and traditional needlework tutorial channels 64 to 76 percent iOS; Instagram Hardanger finished-piece and technique detail accounts 70 to 82 percent iOS; Pinterest Hardanger pattern and traditional Norwegian needlework boards 73 to 85 percent iOS; Facebook Hardanger and Scandinavian embroidery groups 62 to 74 percent iOS.

From November 1, 2026, Patreon passes Apple’s 30 percent App Store commission directly to creators for every iOS subscription processed through the Patreon app. A Hardanger creator whose patron base is 72 percent iOS subscribers will see 72 percent of monthly revenue subject to the 30 percent pass-through before the creator receives any payment.

The mechanism that allows the Apple Tax to be avoided is web-only checkout. Apple’s in-app purchase requirement applies only when a subscription is processed through a native iOS app using Apple’s payment infrastructure. A patron who opens a mobile browser on their iPhone and subscribes through a web payment form uses Stripe’s payment infrastructure instead. The Stripe transaction carries no Apple commission regardless of the device type used to complete it. The patron can then access Patreon content through the Patreon iOS app without triggering any further commission charge: content access through the app does not constitute a new in-app purchase.

For Hardanger creators, the Pinterest discovery pathway is worth noting. Pinterest image search for Hardanger patterns and traditional needlework brings users to a creator’s page through the mobile browser — users tap a pin from their feed, and Pinterest opens the linked content in the default mobile browser, not in a dedicated app purchase flow. A patron arriving at a Patreon membership page through a Pinterest link is already in a web browser context. A web-checkout membership page captures this patron at full rates without any Apple commission.

KeepTier provides a hosted web-only membership page that routes all subscriptions through Stripe in a web browser regardless of the patron’s device and operating system. The November 1, 2026 deadline is public, fixed, and announced by Patreon.

START WITH THE NUMBER

Before you read about the tax, measure yours. Two inputs, one button, zero email capture.