V valley · round 11 — valley depth clarify + dot-to-l + coupled overshoot

Three corrections from operator on round 10: (1) valley depth — operator perceived R4-Bhm @ y=32.75 as deeper than R4-Bw, math says R4-Bw is deeper, surface with explicit indicators. (2) Dot-to-l proportional tightening — should match V-to-e shift? (3) V+dot overshoot needs COUPLED testing, not just V in isolation. Round 11 addresses each cleanly.

Section A · Valley depth clarification — R4-Bw vs R4-Bhm @ y=32.75 with depth indicators

Math: R4-Bw curve dips to y=32.35 (deeper). R4-Bhm @ y=32.75 curve dips to y=32.10 (shallower by 0.25 viewBox units). The visual perception of "Bhm deeper" is likely optical illusion from Bhm's narrower inner-V at top (inner top x=21.14 vs Bw's x=21.00 — inside V is 0.28u narrower). The pinch makes Bhm's curve feel deeper even though geometrically it's not.

Valley depth — explicit comparison at viewBox scale

R4-Bw @ y=33 vertex y=33 · curve start y=31.70 · stroke 12 ★ DEEPEST · curve dips to y=32.35

R4-Bhm @ y=32.75 vertex y=32.75 · curve start y=31.45 · stroke 12.14 curve dips to y=32.10 (0.25u shallower)

Why the visual confusion

Both curves have identical SHAPE (1.30 viewBox units of vertical span from curve start to midpoint, both with d=1.5 backward from vertex along inner leg). The only difference is WHERE the curve sits on the y-axis.

Bhm's curve starts higher (y=31.45) than Bw's (y=31.70) because Bhm's valley vertex is higher (y=32.75 vs y=33). When you look at the V, the curve "region" in Bhm reaches less far down (32.10) than Bw's (32.35).

But Bhm's inner-V is narrower — inner top at x=21.14 vs Bw's x=21.00. That 0.28u of narrowing at the top makes Bhm's inside-V more compressed/wedge-like, which can create the visual impression that the valley is "deeper" even though it's not. Optical illusion from leg geometry.

If you want deepest valley, pick R4-Bw. If you want the slightly chunkier legs (stroke 12.14 vs 12.00) with a slightly shallower valley, pick R4-Bhm @ y=32.75. The difference is real but small.

Section B · Dot-to-l proportional tightening — should dot move with V-to-e?

V-to-e went -0.11em → -0.12em (tightened by 0.01em). Should dot margin-left tighten by a matching 0.01em (or half) to keep the wordmark symmetrically compressed? Three test rows at R4-Bhm @ y=32.75. Red baseline indicator.

dot ml calc(0.03em+2px) CURRENT · asymmetric (V/e tight, l/. normal)

ecreal

dot ml calc(0.025em+2px) half-step · -0.005em tighter

ecreal

dot ml calc(0.02em+2px) ★ full proportional · -0.01em tighter · matches V-to-e shift

ecreal

dot ml calc(0.03em+2px) · 140 current

ecreal

dot ml calc(0.02em+2px) · 140 ★ full proportional

ecreal

Section C · Coupled V+dot overshoot — both translateY moving together

Round 10 only adjusted V translateY. The rev-8 "symmetric optical system" intent says V+dot should move proportionally. Three coupled translateY values. Red baseline indicator — see how V apex AND dot bottom shift together.

V+dot translateY 0.015em less overshoot · both shapes lifted

ecreal

V+dot translateY 0.025em ★ CURRENT rev-8 lock · V at baseline, dot 0.02em below

ecreal

V+dot translateY 0.035em more overshoot · both shapes pushed down

ecreal

Spec summary + final decisions to lock

Issue 1 · Valley depth clarification

Math is clear: R4-Bw @ y=33 is the deepest valley. Both finalists have identical curve SHAPE (same 1.30-unit vertical span). R4-Bw's curve sits at y=31.70→32.35; R4-Bhm's sits at y=31.45→32.10 — Bhm's curve is 0.25 units HIGHER on the y-axis (shallower).

The visual perception comes from leg geometry, not the valley curve itself. Bhm's inner-V is 0.28u narrower at top (inner top x=21.14 vs Bw's x=21.00). The narrower pinch can make Bhm's valley feel deeper visually even though it's geometrically shallower.

If "deepest valley" was your priority, R4-Bw is correct. If "slightly chunkier legs + slightly shallower valley" was the read you wanted, R4-Bhm @ y=32.75 is correct. Section A's annotated comparison makes the math visible.

Issue 2 · Dot-to-l proportional tightening

Argument for tightening: V-to-e tightened -0.01em creates asymmetric wordmark compression (left side tight, right side normal). Symmetric tightening keeps the bookend feel — both clay anchors equally close to the text.

Argument against: V-to-e is letter integration (V chevron overlapping into "e"); dot-to-l is punctuation gap (dot sitting after "l"). Different semantic relationships, might not warrant lockstep movement.

My revised verdict: tighten dot margin-left proportionally to calc(0.02em + 2px). The argument for symmetric compression is stronger when the wordmark is designed as a tight bookend system. The dot/l relationship isn't fundamentally different — both clay anchors are in tight relationship with the text. The asymmetry I argued for in round 10 was over-thinking. Lock proportional dot-to-l shift.

Issue 3 · Coupled V+dot overshoot

You're right — should be coupled. The rev-8 "symmetric optical system" intent is that V and dot move together via the same translateY value. Round 10 broke that by testing V in isolation.

With proportional movement, the same analysis still holds: translateY(0.025em) for both — V apex at baseline (R4-Bhm geometry), dot bottom 0.02em below baseline. The asymmetric DESCENT (V at baseline vs dot below) is intrinsic to the shape difference, not a translateY problem. Rev-8 "symmetric" was always about matched translateY value, not matched descent.

Keep translateY 0.025em for both. The rev-8 lock survives the geometry change because both shapes still benefit from the same downward shift (V apex grounds at baseline, dot bottom maintains its overshoot). No change.

Three locks needed:

1. V geometry: R4-Bw @ y=33 (deepest, stroke 12) OR R4-Bhm @ y=32.75 (slightly shallower, stroke 12.14)
2. Dot margin-left: calc(0.03em + 2px) (current) OR calc(0.02em + 2px) (proportional tighten ★ my revised pick)
3. V+dot translateY: 0.025em coupled (keep ★)

Once locked, rev-11 SVG regen runs.

Vecreal · brand workshop · 2026-05-12 · v-valley round 11