Crucible Year Peter Yuill
Archive · full record

T_419

The work as shown. Every value below comes from vae.db, the master database of the Crucible Year.

Recorded

Date
2026.11.03, day 310
Size
97 × 180 cm
Paper
S13, Tianxuan Ge Dot Ink Becomes Gold, semi-cooked (self-sized), aged
Inks
MB6, MB8, MB9, MB10
Phase
P4 S14
Shown on the site
yes (combined score of 5.0 or more)

The combined score

Combined score
6.50
Read half
35
Measured half
30
The rating shown on a work page, out of 10: V23's rating times 5, plus V24's twelve tests of 50, divided by 10. 5.0 on it decides what the site shows, and it is read here at two decimals for that reason. How it is built.

Read · V23 (v23.13)

V23 rating
7
Form
2
Tonal depth
2
Layering
3
Composition
2.4
Material and surface
2
Ink cover
87%
Dark share
63%
Tonal steps
4
Flaws named by the reads
other 9, dead ink 2, water stain 3
The V23 rating runs from 1 to 9, with 5 the middle of this body of work. It is the model's reading held by code's rules, and the read half of the combined score. Where Peter reviewed a run, this is the reviewed figure.

Measured · V24, from the 16-bit TIFF

Mass
6 of 10
Tone
8.4 of 14
Void
3.2 of 8
Flow
7.6 of 10
Balance
4.8 of 8
Measured half
30 of 50
Mass width in mm
rung 3 of 5
Mass width against the sheet
rung 3 of 5
Evenness of that width
rung 3 of 5
Depth of the darks
rung 4 of 5
Black plateau
rung 2 of 5
Step from mid to dark
rung 2 of 5
Evenness inside the darks
rung 2 of 5
Spread across tonal bands
rung 4 of 5
Voids gathered, not scattered
rung 2 of 5
Directions across the sheet
rung 5 of 5
Local gesture and edge
rung 2 of 5
Balance on the sheet
rung 3 of 5
Mass width
118 mm
Depth of the darks (p75 density)
1.72
Largest void, radius
280 mm
Direction alignment
0.04
Left and right asymmetry
0.62
Resolution
3.89 px per mm
Twelve deterministic tests with no model and no votes. They track physical change across the year and cannot see a mark's quality, so they are no verdict on one work. The tests and their limits.

How the number was made

Read by the Visual Analysis Engine, version 23.13, on 2026.11.03: measured by code, then read 5 times by Claude Opus 5.5 against three reference paintings (R1 to R3), each read independent of the others. How the engine works.

WHAT THE CODE SEES

Three views of one sheet.

Left, the photograph. Centre, the tone stretch the model is shown beside it: the darkest 45% of tones spread over the whole range, so flat black, banding and stains show. Right, every pixel as the measuring code classes it: cyan black ink, deep teal translucent grey, stone the palest ink, near black bare paper.

T_419T_419, tone-stretchedT_419, classed by brightness
87%of the sheet carries ink
Black 63%Grey 30% SHARE OF THE INK, NOT OF THE SHEET
TONAL STEPS

4 families of tone.

Every inked pixel's brightness in 64 bins, smoothed; each circled peak holds at least 4% of the ink. Life inside the black: 0.042 (under 0.045 reads as flat).

0.300.751234 BLACK INK TRANSLUCENT GREYS PALEST INK, 0.88 LUMINANCE OF EVERY INKED PIXEL, 64 BINS. PAPER (ABOVE 0.88) IS LEFT OUT.
FIVE READS

Five reads, agreeing on 4 of 5 qualities.

Each quality placed on its ladder by each read, and the mean the engine keeps. Where reads differ, they differ by one rung; close calls resolve down.

READ 1READ 2READ 3READ 4READ 5MEAN
Form2BUILT2BUILT2BUILT2BUILT2BUILT2.0
Tonal depth2MODELLED2MODELLED2MODELLED2MODELLED2MODELLED2.0
Layering3ORDERED3ORDERED3ORDERED3ORDERED3ORDERED3.0
Composition3COMMANDING2WORKING2WORKING3COMMANDING2WORKING2.4
Material and surface2SOUND2SOUND2SOUND2SOUND2SOUND2.0

READ 1, IN ITS OWN WORDS

“Off-centre radial wheel with legible translucent layering and strong rotation out to a dissolving right fan; held back by bulky, near-flat dark wedges and a heavy lower-centre mass. Surface sound, with only minor marks.”

Form · BUILT

Arcs and spokes follow the brush's travel around the eye; walls hold, but top-left and left wedges and the long right slab remain bulky, filled shapes.

WEAKEST Bulky merged dark mass at bottom centre, loosely shaped

STRONGEST Concentric arcs and tapered radial spokes around the central eye

Tonal depth · MODELLED

Arcs and right fan step richly through greys; large darks measure flat (0.042), showing only faint grain when stretched, so depth is not sheet-wide.

WEAKEST Left wedge and horizontal slab read nearly flat in colour

STRONGEST Stepped grey bands in upper arcs and right fan

Layering · ORDERED

Many translucent passes overlap in readable order across the upper arcs and right fan, building depth; only the heaviest darks at lower centre swallow earlier passes.

WEAKEST Passes fuse into near-black at bottom centre

STRONGEST Translucent overlapping bars and arcs across the right fan

Composition · COMMANDING

Off-centre wheel around white eye drives rotation; long slab and fan carry the eye rightward, while the top-right paper forms a strong diagonal counter-shape. Lower centre is heavy.

WEAKEST Heavy lower-centre dark, the densest and least active area

STRONGEST Off-centre wheel spinning out to the dissolving right fan

Material and surface · SOUND

Ink is controlled, with clean edges and dry-brush fan ends; a seam, small blots and mottling in the darks are minor, mainly visible in the stretched view.

WEAKEST Mottled, slightly cloudy interior of the bottom dark mass

STRONGEST Dry-brush stroke ends along the upper-right fan edge

FLAWS

3 distinct places, 0 major.

Each read must name every flaw and where it is, in a ninth of the sheet. Here the reads named flaws in 3 distinct places, 14 mentions in all across 5 reads; the number in each ninth is how many times a flaw was named there. Reads agreeing on a place strengthen the evidence for one flaw; they do not multiply it.

5 × OTHER“Vertical dark seam running down the interior …”5 × DEAD INK, WATER STAIN“Mottled, cloudy near-black interior of the lo…”4 × OTHER“Small dark blots low in the long horizontal s…”WATER STAINDEAD INKOTHER
RULE 1Two tonal steps or fewer

Tonal depth held at rung 1.

4 steps: not applied
RULE 2A major flaw

Material held at rung 1. Mild waves in the unmounted paper are never a flaw.

none named
RULE 3Any quality below 0.5

The rating cannot exceed 4.

lowest 2.0
RULE 4At least two valid reads

A read that breaks the reply format is discarded.

5 of 5 valid
THE ARITHMETIC

V23 rating 7.0.

Engine v23.13, 5 reads. Fixed arithmetic, the same for every painting; the constants were set on a stratified sample of 24 paintings so the median sheet scores 5. The scale is relative to this archive, not to painting in general, and a difference of less than half a point between two paintings is noise.

MEAN RUNG2.28(2.0 + 2.0 + 3.0 + 2.4 + 2.0) ÷ 5
→
RAW5.561 + 2 × 2.28
→
RATING7.05 + 2.086 × (5.56 − 4.60)