# KH-004 "Sri Yantra" (working name) — LOOM NOTES

**SKU:** HRIDAYA-GTC-004 / Masterpiece 004 (working name **Sri Yantra**; the founder has not confirmed the name yet)  
**OpCo:** Khard Hridaya Inc · Toronto  
**Built:** 2026-09-25, 10:55–11:25 PT

> **Status: new 004 design supplied by the founder on 2026-09-25 at 10:53 PT; replaces Palace Ashtamangala.**
> Source: his reference photo `/workspace/004-new-founder-ref.jpg`. It shows the round carpet on a wood floor, seen at an angle. The file is **1408×1408** px, not 1024×1024 as described in the brief. The retired Palace Ashtamangala pack (cream/emerald/gold) was moved, complete, into `retired-palace-ashtamangala/`. Nothing was deleted.

## Files (`/workspace/hridaya/brand/loom/KH-004/`)

| File | Role |
|---|---|
| `HRIDAYA-GTC-004-LOOM-MASTER.svg` | Loom master. viewBox 0 0 1500 1500 = 1500 mm round. Exact geometry, flat fills only, palette in `<metadata>` |
| `HRIDAYA-GTC-004-LOOM-6000.png` | 6000×6000 render of the SVG (4 px/mm), transparent outside the round |
| `HRIDAYA-GTC-004-LOOM.pdf` | Vector PDF, 1500×1500 mm page |
| `KH-004-FACE-2048.png` | The flat face: the SVG rendered at 2048 px, white outside the circle |
| `004-check.png` | Founder photo · new flat face · new Sit |
| `004-triangle-centre-zoom-600.png` | 600 px crop of the 6000 render at 1:1, on the yantra centre |
| `retired-palace-ashtamangala/` | The whole previous 004 pack (faces, loom SVG/6000, Sit files, weaver brief, `site/`), kept |

## How it was built: geometry from his photo (no trace, no AI redraw)

1. **Rectification.** I fitted an ellipse to the navy disc (the carpet was seen at about 22° tilt: 1216 × 1126 px), then affinely normalised it. Next I took the two full gold circles (the one around the yantra and the one at the base of the outer petals), RANSAC-fitted their edge points, and solved a projective correction that makes both **concentric true circles** (residual about 0.3% of radius, about 2 px). That is a perspective correction, not only an affine squash.
2. **Measurement** was done in the rectified frame:
   - circle radii from radial gold profiles;
   - petal count and angles from the harmonic of the petal outline (16 and 8 are the dominant harmonics; phase −0.2°, set to 0);
   - petal shapes from polar outline profiles, folded over all 16 (or 8) petals, median, then mirrored;
   - the yantra's 27 straight sides from the gold-line skeleton, fitted by least squares per triangle.
3. **Rebuild** used perfect circles, identical symmetric petals and straight triangle sides. **Gold outlines are one constant width, 7.0 mm (about 3.4 knots).** Faces are flat fills. Everything was built as exact polygons (shapely), then written as SVG paths.
4. Colours were **sampled from his photo, region by region.** For each region I took the lit-thread colour (upper lightness percentile of the region's own hue, with gold stitch specks excluded), then grouped the samples into named silk colours.

**Geometry (1500 mm round, centre = 0, y down):**
- Carpet edge r 750 mm (plain navy edge, as in the photo; no gold rim added).
- Outer petal tips r 731.3 mm. Outer-petal base circle (gold) r 596.6 mm.
- Inner petal tips touch that circle. Yantra circle (gold) r 482.7 mm.
- Gold line width 7.0 mm everywhere.

## Counts

- **Outer lotus ring: 16 petals**, every 22.5°, one centred at 12 o'clock. Going clockwise from 12 o'clock: violet, periwinkle, orange, olive, violet (3 o'clock), violet, olive, periwinkle, crimson (6 o'clock), olive, teal, saffron, violet (9 o'clock), crimson, olive, crimson.
- **Inner lotus ring: 8 larger petals**, every 45°, one centred at 12 o'clock. Going clockwise from 12 o'clock: periwinkle, saffron, crimson (3 o'clock), saffron, violet (6 o'clock), orange, periwinkle (9 o'clock), orange.
- **Yantra: 9 interlocking triangles, 4 pointing up and 5 pointing down**, each with one horizontal side, apexes on the vertical axis. The photo has exactly 9 horizontal gold lines, one per triangle.
- Inside the yantra circle the lines make **78 visible regions**: **66 coloured** and **12 navy gaps**. **40** of the 78 are plain triangles; the rest are small quadrilaterals and other polygons where lines cross.
- **26 micro-slivers** are left where three lines nearly meet. Each is smaller than the gold line itself (all under 1 knot), so they are simply gold.

| Triangle | Points | Horizontal side y (mm) | Half-width (mm) | Apex y (mm) | Base angle |
|---|---|---|---|---|---|
| U3 | up | +130.0 | 464.8 | -482.7 | 52.8° |
| U2 | up | +233.5 | 321.3 | -343.8 | 60.9° |
| U4 | up | +65.0 | 156.1 | -224.2 | 61.6° |
| U1 | up | +337.0 | 227.9 | -123.8 | 63.7° |
| D3 | down | -123.8 | 466.5 | +482.7 | 52.4° |
| D2 | down | -224.2 | 322.0 | +337.0 | 60.2° |
| D1 | down | -343.8 | 264.3 | +65.0 | 57.1° |
| D4 | down | -59.8 | 158.0 | +244.3 | 62.5° |
| D5 | down | -14.0 | 102.7 | +139.4 | 56.2° |

These are **his proportions**, not textbook Sri Yantra proportions (base angles run from 52° to 64°). They were not "corrected".

## Knot context (150 knots/in²)

A 1.5 m round is 59.1 in across, about 2,740 in², so **about 410,900 knots** and **about 723 knots across**. One knot is about 2.07 mm, and the 6000 render is about 8.3 px per knot. The minimum shape is 3×3 knots (about 38.7 mm²).

- **Every region is well above 3×3 knots.** The smallest coloured region is about 96 knots (a teal triangle near the centre).
- Gold lines are 7 mm wide, **about 3.4 knots** (the requirement is at least about 2).
- Checked on the 6000 render: only 16 isolated single-pixel anti-alias specks exist at seams, and they are not shapes.

## Palette (10 silk colours)

"Share" is measured on the 6000 render inside the round. Knots ≈ share × 410,900. "Regions" counts separate areas of that colour.

| Yarn | Hex | Share | ≈ Knots | Regions | Smallest region (knots) |
|---|---|---|---|---|---|
| Deep Navy | `#161C33` | 41.2% | ~169,000 | 21 | 1,529.6 |
| Antique Gold | `#C19C70` | 12.0% | ~49,000 | 1 | 48,217.0 |
| Crimson Red | `#863138` | 6.1% | ~25,000 | 8 | 1,117.0 |
| Violet | `#875D81` | 8.2% | ~34,000 | 17 | 136.9 |
| Burnt Orange | `#C2692C` | 5.9% | ~24,000 | 11 | 312.2 |
| Teal | `#266A7C` | 7.4% | ~30,000 | 21 | 96.2 |
| Royal Blue | `#0A315F` | 3.9% | ~16,000 | 16 | 128.7 |
| Periwinkle | `#435383` | 4.8% | ~20,000 | 4 | 4,535.1 |
| Olive Green | `#5E6C48` | 6.8% | ~28,000 | 10 | 971.1 |
| Saffron Yellow | `#D69A43` | 3.7% | ~15,000 | 3 | 4,538.1 |

**SVG:** 0.73 MB · **32 paths + 2 circles** (one path per petal, one per yantra colour, one gold-line path; the disc and the clip circle) · **10 colours**. Only palette colours are used (checked).

## Where his photo had to be interpreted (all small; followed the photo)

1. **Concentric layout.** In the photo, the navy edge of the rug sits about 12 mm off-centre relative to the gold circles, and the outer petal tips vary ±2% in radius with it. That comes from the rug's thickness and the camera angle. I made everything concentric: the petal tips are at their median radius and the edge at the fitted radius.
2. **Petals** were averaged into one symmetric outer shape and one symmetric inner shape, placed at exact 22.5° and 45° steps. Individual hand-drawn petals in the photo differ by a few mm.
3. **Yantra axis and symmetry.**
   - The yantra's axis in the photo was 6 mm right of the circle centre. It is now centred.
   - Each triangle was fitted mirror-symmetric. A few sides in the photo bow or lean by roughly 5–10 mm (notably the right sides of the second-smallest up triangle and the lower down triangle). These are straight now.
4. **Vertices that touch a line in the photo were snapped exactly onto that line** (all ≤ 5.4 mm): D1 apex onto U4 (2.3 mm); D2 apex onto U1 (5.3 mm); D4 L onto U2 (4.8 mm); U1 apex onto D3 (0.7 mm); U2 apex onto D1 (3.1 mm); U4 apex onto D2 (3.4 mm); U4 L onto D2 (0.9 mm); D4 L onto U2 (0.8 mm). The largest up and down triangles have their apexes and base corners on the enclosing circle, as in the photo.
5. **Colours.**
   - The photo is lit from the left, so the same teal reads lighter on the left than on the right or centre. I used **one Teal**, and the same for the two outer **blue** petals (the lower-right one is in shadow).
   - The **top inner petal** sits between violet and periwinkle. It reads blue-violet, so it is **Periwinkle**, like the left inner petal.
   - One tiny face that lies under a gold crossing near the centre takes the teal of its mirror face.
   - No navy was allowed inside the inner star: the photo shows every central cell coloured.
   - Navy was nudged very slightly bluer (sample `#161A29` → `#161C33`) so it reads as the "deep navy" field and not black.
6. **Gold line width.** The embroidered lines in the photo are about 7–9 mm, and slightly thicker where the stitching doubles. They are one constant 7 mm here.
7. Photo texture, sheen, stitch direction and shadows are **not** part of the design and were not carried over (flat fills).

## Fidelity

This is the same design as his photo: navy field, 16-petal outer lotus and 8-petal inner lotus in the same colour order, two gold circles, and the same 9-triangle yantra with the same line layout and the same colour in each cell. The colours were checked against the rectified photo; the cells that needed a judgement call are listed above. Overlaying the rebuilt lines on the rectified photo, they sit on his gold lines to within about a line width everywhere.

No weaver or artisan names. Not a medical device; no health claims.