design.vecreal.com / Workshop / V overshoot revisit

Brand atom workshop · wordmark · round 7b (V overshoot revisit at display sizes)

V overshoot revisit — does 0.0325em read too low at marketing scale?

Operator flag after rev-7 W-0.23 lock: the V at translateY(0.0325em) (rev-5 V2a) feels "a touch low" at 72-96 px display sizes. Em-proportional overshoot is correct in principle — but the picked VALUE was calibrated at workshop sizes (32 px ≈ 1 px descent). At marketing scale (96 px ≈ 3 px descent), the same em-value reads heavier. This workshop tests 5 overshoot values at the larger display sizes specifically.

Pixel descent by value × display size

The em-overshoot descent in actual pixels at each size:

OVERSHOOT	32 PX	48 PX	72 PX	96 PX	128 PX
0em (zero)	0	0	0	0	0
0.020em	0.64	0.96	1.44	1.92	2.56
0.025em (matches dot)	0.80	1.20	1.80	2.40	3.20
0.030em	0.96	1.44	2.16	2.88	3.84
0.0325em (CURRENT rev-5)	1.04	1.56	2.34	3.12	4.16

Pixel-descent threshold from "subtle anchoring" to "visibly low" sits around 2-2.5 px for most viewers. Current 0.0325em crosses that threshold around 72 px and clearly sits below the line at 96 px+.

V-ZERO

No overshoot. V tip exactly at baseline.

V overshoot 0em
descent @ 96 px 0 px
reads as V geometrically anchored

Eliminates overshoot entirely. May read as "V tip floating above baseline" at small sizes (the original rev-5 problem).

ecreal.32 px

ecreal.48 px

ecreal.72 px

ecreal.96 px

ecreal.128 px

V-020

Light overshoot. ~2 px descent at 96 px.

V overshoot 0.020em
descent @ 96 px 1.92 px
reads as subtle anchor, still grounded

Less than dot (0.025em). Anchors the V without sinking it at marketing scale.

ecreal.32 px

ecreal.48 px

ecreal.72 px

ecreal.96 px

ecreal.128 px

V-025 · MATCHES DOT ★

V overshoot = dot overshoot. Symmetric optical system.

V overshoot 0.025em
descent @ 96 px 2.40 px
reads as consistent with dot, less sink at large

The rev-5 workshop tested this as V2. Operator picked V2a (0.0325em) for "stronger pointed-shape compensation" — but didn't review at 96 px scale.

ecreal.32 px

ecreal.48 px

ecreal.72 px

ecreal.96 px

ecreal.128 px

V-030

Between dot and current. ~3 px descent at 96 px.

V overshoot 0.030em
descent @ 96 px 2.88 px
reads as slight pointed-shape compensation

Halfway between V-025 (matches dot) and V-current (rev-5 0.0325em). Still some pointed-shape bonus descent.

ecreal.32 px

ecreal.48 px

ecreal.72 px

ecreal.96 px

ecreal.128 px

V-CURRENT (rev-5 V2a)

Locked rev-5. What the operator flagged as "too low" at 72-96 px.

V overshoot 0.0325em
descent @ 96 px 3.12 px
reads as visibly below baseline at marketing scale

Reference. Kept for direct comparison.

ecreal.32 px

ecreal.48 px

ecreal.72 px

ecreal.96 px

ecreal.128 px

Stacked at 96 px for direct A/B (the size where the issue manifests)

V-ZERO ecreal.

V-020 ecreal.

V-025 ★ ecreal.

V-030 ecreal.

CURRENT ecreal.

Spec deltas at a glance · V geometry + dot + overshoot all locked from rev-7 except V translateY

VARIANT	V translateY	DESCENT @ 32 px	DESCENT @ 96 px	NOTE
V-zero	0em	0 px	0 px	no overshoot — risk of "floating tip" at small sizes
V-020	0.020em	0.64 px	1.92 px	light anchor, less than dot
V-025 ★	0.025em	0.80 px	2.40 px	matches dot — symmetric optical system
V-030	0.030em	0.96 px	2.88 px	slight pointed-shape compensation
V-current	0.0325em	1.04 px	3.12 px	rev-5 V2a locked — flagged "too low" at 96 px

Reply with your pick — "V-zero" through "V-current". My read: V-025 (matches dot exactly) is the simplest fix — makes the dot and V optical-overshoot system symmetric (one rule, two atoms) and reduces 96 px descent from 3.12 px to 2.40 px, which crosses the visibility threshold back into "subtle anchoring." V-020 is the next defensible pick if V-025 still feels too low at marketing scale. Once picked, propagates to wordmark.svg, all consumer HTMLs, design-tokens.json, and gets a rev-8 CHANGELOG entry. V-current = no change (keep rev-5 lock).