The seams that survive start before the machine: with a pattern that is actually your size. Print at true size and prove it, the app's pattern resizer prints a calibration sheet (a one-inch square next to its 25.4 mm twin) as its own first sheet; measure it with a real ruler before you trust a single tile, because a printer's "Shrink to Fit" quietly scales down every future seam line. The tiled pages carry overlap borders for gluing or taping, so joins stay registered across the floor.
A stitch line is the design; the allowance is the room to repair, alter and press it. The app's soft-goods tools bake this in as millimeters: the circle-skirt and cloak tools open at a 10 mm waist seam allowance and a 20 mm hem allowance. Change the allowance and the tool re-cuts the math, the allowance is an input, not a habit.
One hard rule the calculator enforces: a waist seam allowance must stay smaller than the waist radius. Ask for more than the radius and the tool refuses to run, the allowance outgrew the garment.
The circle tool takes waist or neck circumference, finished length, and a fullness in circles, the app's choices are ¼, ½, ¾, 1 or 2 circles. The stitching radius is the circumference over two-pi times the fullness; the inner cutting line sits one waist-seam allowance inside it, and the outer cut adds the finished length plus the hem allowance. Then The plain question: does that outer arc fit across your usable bolt width (the tool defaults to 1,500 mm)? If yes, you get one conservative circle layout in meters; if no, the tool says so plainly, use panels or a planned layout. It calls its own layout conservative on purpose: it is not a nesting plan.
Gore panels are the paneled answer. The planner takes 1–100 panels and keeps every panel's sector at or below 90°, for a full circle of fullness, that means four gores or more, and bakes the seam allowances into every panel's cut size. Add the optional hood and the planner returns two hood side blanks, cut two, but read the fine print it ships with them: the hood dimensions come from a tested pattern, so match the neckline to the cloak and test a toile before cutting good cloth. Yardage is bounding rectangles plus a 10% shrinkage-and-layout reserve; add lining separately, and directional prints may need more.
The app's pleat tool plans three constructions, and each has its own honest intake, fullness between 2× and 4× the finished width is the fabric already spoken for:
Work by count or by target spacing (spacing rounds to whole repeats and re-fits the panel), keep the edge allowances outside the repeats, and the tool's own advice is the craft: make one paper or fabric repeat first. Bulk and spring-back change the result.
A 600 mm finished skirt band in knife pleats. The intake decides the fabric before the machine runs: at the tool's honest 2× fullness the band eats 1,200 mm of fabric to show 600, the fabric already spoken for. Worked by count, twelve repeats across the band puts each finished repeat at 50 mm, and the knife construction reserves twice the return depth per repeat as hidden fold, deeper folds ride on more fabric, the tool's line. The accordion variant carries its own honesty: the plan's width is the ideal open zigzag, so a fabric with real thickness settles narrower than the plan — press the sample, and trust the sample. The gate before any cloth: one paper repeat (a 50 mm finished knife fold in paper) shows the bulk and the spring-back the machine will fight, then one fabric repeat settles the real depth. That single repeat is the difference between a panel that works and 1.2 metres of learned lessons.
Continuous bias tape is one formula: the starting square's side is the square root of (tape length × unfolded strip width × one-plus-waste). The tool defaults to a 15% allowance for seams and trimming, takes the cut width before folding (not the finished tape width), and tells you to round up before cutting, joins and trimming always eat usable yield, so check the assembly method you actually use.
On a hybrid build, some of a garment's worst stress isn't in the garment, it's where armor bolts onto cloth. The wall's sewing shelf keeps Kamui Cosplay's armor-attachment clip for exactly this; the foam side of the bargain lives in the first armor build guide (2 mm foam is strap territory, and a seam that cracks at a flex point wants fabric or a thin foam strip behind it). The con planner's packing list keeps repair-kit items grouped by element, when a bodysuit seam pops on a con floor, the fix travels with that costume, in the bag, on purpose.
Fabric weathering is its own craft, paint behaves differently on cloth than on foam or resin, and the shelf's distressing clip is the visual teacher. The app's soft-goods math is silent on distressing (no invented numbers here), but the wall's sewing shelf carries the how-to, and the paint side of aging, washes, dry-brush, chips, is the prop paint guide's job. Capes and robe bands that need weaving rather than sewing are next door at Fiber Bench, where the family keeps the loom crafts.
| Symptom | Cause | Fix |
|---|---|---|
| The pleat panel cuts far wider than the finished piece | Fullness is the intake, between 2× and 4× the finished width | Make one paper or fabric repeat first; bulk and spring-back change the result |
| Accordion pleats sit narrower than the plan | The plan's width is the open zigzag; compression is fabric thickness and pressing | Press a sample; fold depth is ideal-only, so trust the sample |
| The hood won't meet the cloak's neckline | Hood numbers are bounding blanks, not a fitted pattern | Test a toile, match the neckline, then cut good cloth |
| Bias tape comes up short | Joins and trimming eat the yield | Round up before cutting (the 15% default is a start) and check your assembly method |
| Fabric ran out mid-build | The yardage is bounding rectangles plus the 10% reserve, grain, lining and directional prints are extra | Add lining separately; reserve more for directional prints |
| The calculator refuses to run | The waist seam allowance grew past the waist radius | Shrink the allowance, the tool is protecting the garment |
| Every tile came out slightly small | The printer's "Shrink to Fit" scaled the tiles | Reprint with scaling off, then remeasure the calibration sheet — the one-inch square against its 25.4 mm twin — before any seam line |
| Mirrored pieces cut as one hand | The flip is a review step, not automatic | Check the mirrored parts list before cutting; cut one, flip the paper, cut two |