1. Compensation stitches at pattern boundaries
In a bargello pattern, every stitch unit is specified by two numbers: its height (how many canvas threads it covers vertically) and its horizontal position (which column of canvas holes it enters and exits). The peak row of a standard flame stitch might call for a unit sequence of 2-4-6-4-2-4-6-4-2 repeating across the canvas width, where each number specifies a vertical satin stitch covering that many canvas threads. In the interior of the canvas, every position in that sequence can be stitched with full units because there are enough canvas threads above and below the stitch position to accommodate the full height.
At the canvas edges, this is no longer true. The step sequence that produces the pattern’s peaks and valleys will arrive at the left and right canvas edges in the middle of a rising or falling section. The next position in the sequence would require a stitch unit whose top thread entry is outside the canvas area, or whose bottom thread exit is below the first available canvas thread. Neither is possible.
The solution is a compensation stitch: a shorter unit that fills that edge position by covering only the canvas threads that are available within the canvas boundary. If the pattern sequence calls for a 6-thread unit at a position near the top of the canvas, but only 3 canvas threads exist between that column’s current step height and the canvas top edge, the compensation stitch covers those 3 threads. The stitch enters at the highest available hole in that column and exits at the lowest hole the full unit would have exited — or vice versa if the unit is near the bottom edge.
Compensation stitches are not shorter because the stitcher chose to make them shorter. They are shorter because the canvas boundary has cut the available thread coverage. Their purpose is to bring the stitching right to the canvas edge with no canvas visible between the last full unit and the edge. Without them, there will be a ragged gap of bare canvas around the perimeter of the bargello field.
The same compensation logic applies at any internal boundary — a frame area left unstitched around a pillow, a defined border within a four-way bargello piece, or any line at which the bargello pattern must stop. At an internal boundary, the compensation stitch stops at the boundary line rather than at the canvas edge, but the logic is identical: cover as many canvas threads as are available between the current step height and the boundary.
The practical problem for Patreon creators is that most bargello pattern software and most charting methods generate only full-height units. A charted pattern will show the peak row with perfect symmetrical zigzags that terminate cleanly at the canvas edge — because the software draws the chart to a convenient repeat width rather than to the actual canvas dimensions the patron will use. Patrons who purchase a digital pattern and attempt to stitch it on their chosen canvas will find that the peak sequence does not terminate cleanly at their canvas edge unless their canvas happens to be exactly a multiple of the pattern repeat width.
For Patreon documentation, the practical solution is to stitch the first repeat of the peak row on your own canvas to the actual canvas edge, note which positions required compensation stitches and what their abbreviated heights were, and include that information in the pattern notes. A diagram showing the last several stitch positions before each canvas edge — with the compensation stitches drawn at their actual heights — is clearer than a verbal description. Specify that compensation stitches are necessary for any canvas width and that the heights shown are specific to the canvas mesh count and canvas dimensions used in the sample; patrons using different canvas sizes will need to calculate their own compensation heights using the same principle.
2. Center-out counting: why to start from the middle of the canvas
The most common way to begin a bargello project is to start at one corner of the canvas — pick up the canvas, thread the needle, begin at the upper left corner, and stitch across to the right. This approach feels natural because it follows reading direction and because the canvas is held with a consistent orientation. It is the approach that produces off-center bargello patterns.
Most bargello patterns have a horizontal repeat: the peak zigzag sequence rises to a maximum height, falls to a minimum height, and returns to the starting position over a fixed number of canvas columns — the repeat length. If the canvas width is not exactly a multiple of that repeat length, the last partial repeat at the right edge will not mirror the first partial repeat at the left edge. The left edge will have begun in the middle of a rising section; the right edge will end in the middle of a falling section, or vice versa. The piece will be visually asymmetric.
Center-out counting resolves this. Find the exact center of the canvas width by counting the total number of canvas holes in the horizontal direction and dividing by two. Mark that center column with a removable basting thread or a fine-point water-soluble marker. Determine which position in your pattern sequence corresponds to the visual center of the pattern — for a symmetrical flame stitch, this is typically the apex (highest peak) or nadir (lowest valley) of one complete peak unit. Begin stitching from that center position, working outward toward both canvas edges simultaneously.
Working outward toward both edges simultaneously means: stitch the first unit at the center position, then stitch the next unit to the right, then the corresponding unit to the left (the mirror image of the right unit across the center axis), then the next unit to the right again, and so on. This is slower than stitching one direction continuously, but it guarantees that any partial repeat that appears at the right edge will be the exact mirror of the partial repeat at the left edge. The piece will read as intentionally composed from the center outward.
Before starting any color rows, stitch the first complete peak row and verify that it closes correctly. “Closes correctly” means: at the point where you have completed exactly one full repeat of the pattern measured from the center position, the next stitch position in the sequence is the same as the starting position. In other words, if you started at the center at a height of 4 canvas threads from the bottom of the stitch unit, and your repeat length is 16 columns, then the stitch at column 16 counting from center should also be at height 4. If it is not, there is a counting error in one of the units between center and column 16, and it must be found and corrected before adding color rows — once color rows are in, finding a miscounted unit in the peak row requires unworking all the color rows above and below it.
The verification step is the most important discipline in bargello construction. An off-by-one error in a stitch height at column 8 of 16 will produce a peak row that “looks right” locally but that is one canvas thread too high or too low at the return-to-starting-position check. That one thread of error causes the first color row to be one thread off at the same column, and every subsequent color row compounds the error. The finished piece will have a subtle jog in the zigzag pattern at that column position that appears in every row but is nearly impossible to identify after the fact without unworking the piece.
For Patreon documentation, specify the center-out approach explicitly in the getting-started section of every bargello pattern, not as optional advice but as the required method. Include a note identifying which position in the peak row sequence corresponds to the visual center of the design — for complex patterns with multiple peak heights and step intervals, the center position is not always obvious from the chart.
3. Thread twist in long stitches and the drop-the-needle technique
All embroidery thread is spun with a twist direction: the individual fibers are aligned and then spun either clockwise (Z-twist, named for the diagonal of the letter Z) or counterclockwise (S-twist). Plied threads have the plies spun together in the opposite direction from the individual strand twist — S-plied strands over Z-twisted fibers, or Z-plied strands over S-twisted fibers — which produces a balanced thread that lies flat without curling. Paternayan Persian wool uses Z-twisted single strands plied S-wise into three strands; Appleton crewel wool is a 2-ply with a similar balanced construction.
The problem is that pulling thread through canvas holes is not a neutral action with respect to twist. Each pass of the needle through a canvas hole produces a small rotational torque on the thread as it slides through the friction of the canvas structure. The direction of this torque depends on the angle of approach, the direction of travel, and the canvas mesh type. In practice, stitching left-to-right tends to add Z-twist to the working thread (tightening it), and stitching right-to-left tends to add S-twist or remove Z-twist (loosening it).
In short tent stitches or small cross stitches, the thread passes through canvas holes many times per inch of finished work, and the accumulated twist after each pass is small relative to the total thread length consumed. In bargello, the situation is different: each stitch covers four, six, or eight canvas threads in one long pull from front to back. A single 6-thread unit on 18-mesh canvas consumes 6/18 = one third of an inch of vertical canvas in one pass. Over the course of a row, the thread is pulled through far fewer canvas hole pairs per inch of coverage than in tent stitch, which means far fewer opportunities for twist to cancel as the direction of pull changes. The accumulated twist from even ten or fifteen bargello stitches can be enough to visibly tighten or loosen the thread plies.
When the thread is too tightly twisted relative to its natural resting state, it becomes narrow, stiff, and springy. It covers less canvas width per stitch (the thread forms a tighter cylinder) and the stitch surface appears slightly ridged — the plies are visible as a spiral texture rather than lying flat. When the thread is too loosely twisted (over-unspun), the plies separate slightly and the thread becomes fuzzy, covering more canvas width but with reduced color saturation and a textured surface that reads as slightly lighter in value than the intended color.
The drop-the-needle technique corrects accumulated twist. After every ten to fifteen bargello stitches — or whenever the thread begins to feel resistant to pulling through, or the stitch surface looks uneven — simply release the needle and let it dangle freely from the working thread. Do not hold the thread; let it hang under only the weight of the needle. The needle will spin in the direction that unwinds the accumulated twist until the thread reaches its natural balanced state, then stop. This typically takes five to twenty seconds depending on the amount of accumulated twist. Resume stitching.
Two supplementary techniques help: first, work with shorter thread lengths. Thread cut to 14 or 15 inches consumes its length in fewer passes and reaches the end of the thread before twist accumulates to a problematic level. This costs more thread tail waste at each end, but the surface consistency improvement is worth it on fine canvas or visible display work. Second, maintain a consistent stitching direction within each color row where possible — not always possible in bargello where some rows stitch partly left-to-right and partly right-to-left within the same zigzag unit, but on straight horizontal color rows stitching always in the same direction minimizes twist accumulation variability between rows.
For Patreon tutorial content, include the drop-the-needle habit in the materials setup section rather than the troubleshooting section. Position it as a routine practice rather than a correction for errors. Patrons who see it in the troubleshooting section will only use it after they already have visible twist problems; patrons who see it in setup will use it proactively and avoid the problem.
4. Canvas directional distortion and wet-blocking
Needlepoint canvas is woven from stiff interlocked threads: typically cotton, linen, or coated poly in mono canvas, and paired threads of the same materials in Penelope double-canvas. The woven structure holds the canvas square while it is unstretched and unstitched, but stitches pull the warp and weft threads in directions that the weave structure was not designed to resist.
In bargello, the pull is directional. All stitches are vertical: every needle entry and exit pulls the vertical (warp) canvas threads slightly toward the center of the stitch, and every stitch passes over horizontal (weft) canvas threads, pressing them laterally. Across an entire worked piece, the cumulative warp-thread pull in the vertical direction slightly compresses the canvas vertically (the finished piece is shorter top-to-bottom than the original design dimensions) and the lateral spreading of weft threads slightly expands it horizontally (the finished piece is wider left-to-right than the original dimensions). The net result is a bargello piece that is shorter and wider than it should be.
This distortion is not a mistake and is not correctable by changing stitch technique. It is an inherent consequence of the stitch structure. Wet-blocking corrects it after the piece is complete.
Wet-blocking protocol:
Prepare a blocking board large enough to hold the finished piece flat: a piece of dense foam, a corkboard, or a purpose-made needlepoint blocking board. Using a permanent marker, draw the original design dimensions on the blocking board as a rectangle aligned with the board grid. If your board has a printed grid (one-inch or two-centimeter squares), align the rectangle to the grid lines.
Mist the completed piece with clean water from a spray bottle, wetting both the front and back of the canvas until the canvas threads are pliable and the wool slightly darkens from moisture. The piece should be damp throughout — not soaking wet. Excess water can cause wool dyes to bleed if you have used threads from different dye lots or from brands with less-than-colorfast processing. If you are uncertain about colorfastness, test by wetting a small corner of the unused canvas selvedge and blotting it onto white paper to check for dye transfer before wetting the entire piece.
Place the piece face-down on the blocking board with the canvas selvedge aligned to the marked rectangle. Face-down blocking prevents the T-pin heads from leaving impressions in the stitched surface. Gently stretch the canvas to the marked rectangle dimensions, pulling on the canvas margin (the unstitched selvedge area, typically one inch from the stitched area to the canvas edge) rather than on the stitched surface itself. Pin the first corner with a T-pin driven through the canvas margin into the blocking board. Pin the opposite corner, stretching the canvas to the correct dimension. Pin the remaining two corners. Then add pins at one-inch intervals along each edge, working outward from the corners on both sides alternately — pin two inches from the top-left corner on the top edge, then two inches from the top-right corner, continuing inward until the entire perimeter is pinned. This alternating approach distributes the blocking tension evenly rather than pulling the canvas toward one side.
Allow the piece to dry completely before removing pins. Complete drying takes 24 to 48 hours depending on canvas weight and ambient humidity. Test dryness by touching the center of the canvas back: it should be at room temperature and show no coolness (remaining moisture feels cool against the hand). Remove pins only after the canvas is fully dry. A piece removed while still slightly damp will spring back toward its unblocked shape as the remaining moisture evaporates.
After removal from the blocking board, the blocked dimensions should hold as long as the piece is stored flat or rolled (not folded) and kept away from moisture. A finished pillow or framed piece will not require re-blocking unless it gets wet again.
For Patreon documentation, include the blocking board setup in the required-materials list at the start of every bargello pattern. Patrons who do not have a blocking board may attempt to substitute an ironing board and steam pressing. Steam pressing bargello wool is possible but riskier: excess steam can full the wool slightly, and pressing with the iron surface on the stitched front (rather than on the back through a damp cloth) can flatten the stitch texture permanently. Specify blocking on a board as the preferred method and steam pressing on the reverse through a damp cloth as an acceptable alternative only for pieces that have dried in a slightly distorted shape after the initial block.
5. Color gradient value stepping: how many steps for a smooth gradient
The visual power of bargello is the gradient: bands of color stacked in horizontal rows that shade from one value to another across the depth of the piece. Whether those bands read as a smooth continuous shading or as a set of distinct stripes depends almost entirely on value — the lightness or darkness of each band relative to the bands above and below it.
Value is not the same as hue. Hue is the color family (red, blue, green, yellow). Value is how much white or black is mixed into a hue: light pink and dark crimson are the same hue (red family) but very different values. A bargello gradient that changes hue without changing value — moving from medium-value blue to medium-value green with no light-to-dark progression — will read as flat color stripes, not a gradient, because the eye interprets value difference as depth.
The practical specification: five to seven distinct value steps produce smooth gradients in standard 4-canvas-thread bargello row height. At five steps, each step spans a value jump of approximately 20% of the total light-to-dark range (assuming the gradient travels from near-white to near-black or from the lightest to darkest available version of one hue). This is near the limit of smooth perception — a viewer looking at the piece from normal viewing distance (two to three feet for a pillow or framed work) will perceive a continuous shading. At four steps, the individual jumps are 25% each, and the bands are visible as discrete stripes with distinct edges, especially in raking light.
At seven steps, each step is approximately 14% of the total range, and the gradient reads as smooth at all but close inspection distances. Beyond seven steps, the individual rows become narrow enough (at standard 4-canvas-thread height on 18-mesh canvas, each row is approximately 5.6mm tall) that the total gradient depth at seven rows is 39mm — about 1.5 inches. An eighth row at the same step interval would add another 5.6mm but would likely be perceptually indistinguishable from its neighbors, making the additional row invisible and wasting thread on a step the viewer cannot see.
For gradient selection in practice: the Appleton crewel wool shade series and the Paternayan Persian wool shade families are organized around this scale. The Appleton 940–948 series (pale pink to deep rose) provides nine shades that span the value range evenly. Using shades 940, 942, 944, 946, 948 — every other step — produces a 5-step gradient; using 940, 941, 943, 945, 947, 948 produces a 6-step gradient covering nearly the full range. The Paternayan families use first-digit grouping: the 900 series (creams) and 910 series (tans) each provide six to eight values within one hue family.
DMC tapestry wool (the 7000 series) provides less predictable value stepping within families — the numbering within a family does not always correspond to monotonic light-to-dark progression, and some families have five members while others have only two or three. When using DMC tapestry wool for gradient bargello, select candidate threads by color number and then arrange them physically in a light-to-dark sequence to confirm the actual visual value order before purchasing in quantity. Photograph the sequence under daylight-balanced light (not incandescent, which shifts warm colors toward orange and cool colors toward gray) and check the grayscale version of the photograph to verify value order without the hue influence.
Document thread color selection by brand, series name, and color number (not color name) in every Patreon pattern. Color names are not standardized across brands, and brands occasionally rename colors between dye lots. A patron purchasing thread in 2027 using color names from a 2026 pattern may receive threads from a renamed or reformulated series that does not match the original value progression.
6. Pattern repeat calculation: confirming the repeat closes before adding color rows
A bargello pattern repeat is the minimum sequence of stitch heights and horizontal positions that, when repeated end-to-end across the canvas width, produces the full design. For a simple two-peak flame stitch with a 2-thread step interval and stitch heights of 4-6-4-2-4-6-4-2 (eight units per repeat), the repeat length is 8 columns. For a complex Pomegranate bargello with multiple height levels and a step interval that varies within the row, the repeat length might be 24 to 40 columns.
The repeat is defined as “closed” if the stitch at position (repeat length + 1) is exactly the same as the stitch at position 1. That is: after stitching one complete repeat, the next stitch is the same as the first stitch. This closure property is what allows the repeat to tile across the canvas width: each copy of the repeat picks up exactly where the previous copy ended.
Before committing to color rows, stitch the first peak row across one complete repeat length and verify closure. The verification: the last stitch of the repeat ends at a canvas thread height H; the first stitch of the repeat started at the same canvas thread height H; and the next column position in the sequence (which would begin the second repeat) calls for a stitch starting at height H. All three must agree.
If the last stitch of the alleged repeat is one canvas thread too high or too low, the error is not at the last stitch: it is somewhere within the repeat, where one unit’s height was miscounted by one thread. Finding that error requires comparing each unit in the stitched row to the chart, height by height. Working with a ruler or thread gauge alongside the chart is faster than counting threads on the stitched canvas by eye.
After confirming closure over one repeat, verify that the repeat length divides into the canvas width in a way that produces acceptable partial repeats at the edges. The calculation: available canvas width in holes ÷ repeat length in columns = number of full repeats + remainder. If the remainder is zero, full repeats fill the canvas exactly and no partial repeats are needed. If the remainder is nonzero, divide it by two (centering the pattern means equal partial repeats on each side of center). If the remainder is odd, the two partial repeats will differ by one column — one will have the extra column. Decide whether this asymmetry is acceptable or whether the design canvas dimension should be adjusted by one canvas hole (changing from a 100-hole-wide canvas to a 99-hole-wide canvas, for example) to produce an even remainder.
For Patreon pattern documentation, include the repeat length and closure verification as an explicit step in the pattern instructions, not as background information in a footnote. Patrons who skip the closure verification and discover a miscounted unit after stitching four complete color rows will face the choice between unworking those rows or living with the jog — and will blame the pattern, not the skipped verification step.
7. Penelope double-canvas versus mono canvas
Standard needlepoint canvas — mono canvas — is woven with single threads in both the warp and weft directions. Each intersection of one warp thread and one weft thread creates one canvas hole, and each stitch unit enters and exits canvas holes as individual thread pairs. Penelope double-canvas (also called duo canvas) is woven with paired threads in both directions: each “hole” in the weave is actually formed by two closely spaced parallel warp threads crossing two closely spaced parallel weft threads. Visually, Penelope double-canvas looks like a grid of paired lines rather than single lines.
For standard bargello, Penelope double-canvas is stitched the same way as mono canvas: each stitch unit enters and exits holes treating each thread pair as one counting unit. A 4-canvas-thread stitch on 14-mesh Penelope double-canvas covers 4 counting units — which are actually 8 individual threads — but is stitched and counted the same as a 4-canvas-thread stitch on 14-mesh mono canvas. The visual scale of the finished stitch is the same; the mechanical difference is that the doubled threads make the canvas stiffer and more resistant to distortion during stitching. Penelope double-canvas was historically preferred for bargello precisely because its greater stiffness reduces the directional distortion described in the blocking section, making the piece easier to manage during stitching without a frame or stretcher bars.
The additional use of Penelope double-canvas is tramé work: a preparatory technique in which individual long horizontal stitches are laid across the canvas surface, between the paired warp threads, to create a padded underlayer that the vertical bargello stitches then cover. To separate the paired Penelope warp threads for tramé laying, insert a laying tool (a blunt stiletto or the tip of a closed pair of scissors) between the two warp threads at each column and gently push them apart, creating an enlarged gap between the pair. Thread a tapestry needle with the tramé thread (typically the same wool or a thinner underlayer thread in a matching color) and pass it horizontally through the gap between the separated warp threads, from left to right or right to left across the full width of one color area. Release the warp threads; they spring back together, trapping the tramé thread between them. Work one tramé row for each horizontal bargello color row. The covering bargello stitches are then worked over the tramé underlayer.
Tramé serves two purposes in bargello: it increases the apparent height (three-dimensionality) of the finished stitch surface, producing a softer, fuller texture than stitches on bare canvas; and it masks any gaps between adjacent bargello stitch units that might otherwise show canvas threads at the edges of units. Tramé is optional — the majority of contemporary bargello Patreon content does not use it — but it is the traditional finishing technique for Florentine work on historical pieces and relevant content for creators producing traditional-style work.
For four-way bargello specifically, Penelope double-canvas has an advantage: the paired thread structure makes counting canvas threads accurately across the quadrant boundary easier. In four-way bargello, stitches in one quadrant run vertically and stitches in the adjacent quadrant run horizontally. The point where these two stitch directions meet — the quadrant boundary — requires careful counting to ensure that the last vertical stitch unit and the first horizontal stitch unit meet cleanly without a gap or overlap. The Penelope paired thread structure provides a slightly more visible reference line for counting at the boundary than single-thread mono canvas.
When specifying canvas type in Patreon patterns, always state the canvas type by name (mono canvas or Penelope double-canvas), mesh count (holes per inch for the counting unit, not the individual thread count), and whether the stated mesh count refers to single threads or paired-thread counting units. A 14-mesh Penelope double-canvas has 14 counting-unit holes per inch but 28 individual threads per inch. A patron who counts individual threads rather than counting units will produce a piece at half the intended scale.
8. Apple Tax: what November 2026 costs bargello needlepoint creators on Patreon
Bargello and needlepoint tutorial content is concentrated on visual platforms where the color-gradient aesthetics and close-up canvas photography translate well. Pinterest needlework and fiber arts boards reach 68–80% iOS users. Instagram needlepoint accounts reach 70–82% iOS. YouTube needlepoint and canvas work channels reach 55–70% iOS. TikTok needlework content reaches 72–84% iOS.
Starting November 1, 2026, Patreon applies Apple’s 30% billing fee to all iOS-originated subscriptions — any subscription purchased or renewed through the Patreon iOS app rather than through a web browser. The fee is applied to the gross subscription amount before Patreon’s own platform fee (5–12% depending on plan) and before payment processing.
At $12/month with 70% iOS and 50 subscribers: Apple’s 30% on iOS-billed subscriptions = $12 × 0.70 × 0.30 = $2.52 per subscriber per month × 50 = $126/month ($1,512/year).
At $20/month with 74% iOS and 60 subscribers: $20 × 0.74 × 0.30 = $4.44 per subscriber per month × 60 = $266.40/month ($3,196.80/year).
At $40/month with 78% iOS and 35 subscribers: $40 × 0.78 × 0.30 = $9.36 per subscriber per month × 35 = $327.60/month ($3,931.20/year).
These amounts go to Apple, not to the creator. Calculate your specific fee delta at keeptier.com.
The mitigation path: update all discovery platform profiles (Instagram bio, Pinterest links, YouTube description) to link directly to your Patreon web URL rather than to the app store landing page. Enable web-only billing on the tiers you want protected. Web-only billing disables iOS in-app purchase for those tiers, routing all new subscriber transactions through web billing regardless of how the patron originally found your page. Existing iOS-billed subscribers will need to cancel and resubscribe via web when their current billing period ends. The Patreon help center walkthrough for enabling web-only billing is accessible from your creator settings under Earnings > Billing settings. Pair this with a brief patron communication explaining that you are switching to web-only billing to keep the fee within Patreon and Stripe rather than passing it to Apple. Most patrons who understand where the money goes will resubscribe via web without complaint.