# VectorPaint expression families: keep the character structure while changing features

Language: en
Canonical: [VectorPaint expression families: keep the character structure while changing features](<https://allluckly.cn/en/guides/vectorpaint-character-expression-family/>)
中文：[中文 HTML](<https://allluckly.cn/guides/vectorpaint-character-expression-family/>)
English: [English HTML](<https://allluckly.cn/en/guides/vectorpaint-character-expression-family/>)
Published: 2026-11-03T00:00:00+08:00
Modified: 2026-11-03T09:34:11+08:00
Author: Bison

Create three cartoon expressions from one original head, recording fixed and variable features and checking identity across versions.

English translation of the [Chinese original](<https://allluckly.cn/guides/vectorpaint-character-expression-family/>), translated on 2026-11-03. Original publication dates, attribution, code and version notes are retained.

Three expressions should still read as one character instead of changing head geometry each time. Draw calm, questioning and laughing versions of an original simplified character, focusing on shared structure and local changes. These are cartoon conventions, not assessments of real emotions, and require no automatic face replacement or expression generator.

## List identity features that remain fixed

Reserve three equal display areas on a 1200 by 600 pixel canvas and build one head first. Record fixed proportions, hairstyle, ears and a small identifying accessory, with brows, eyes and mouth as variables. Use an original character rather than a real portrait or brand. Equal display areas are comparison references, not a requirement to distort outlines into identical bounds. Save a base head with accessible features so later versions change locally rather than redraw identity.

## Establish reusable feature references

Mark an eye line, nose reference and mouth area on a separate draft layer. These guide this character rather than prescribe realistic facial proportions. Establish eye size/spacing against the hairstyle before duplicating. Hide ordinary references in delivery. Organise head, hair, accessory and features by purpose and test independent selection so moving a brow or mouth does not disturb identity elements.

## Use restrained marks for the calm baseline

Start with relatively level brows, clear eyes and a slight mouth curve without blush, sweat marks or explanatory text. The relationship among features carries the expression, not the mouth alone. Preserve space from hair and chin. Check the character and accessory at reduced size before judging features. Refine readability without demanding identical perception from everyone, then preserve this baseline and create independent copies for the other expressions.

## Change only relevant features for the questioning version

Raise one brow in the second copy while preserving eye spacing, then test a narrower or slightly offset mouth. Avoid lengthening the head or flipping the hairstyle unintentionally. Keep any question mark on a separate layer and test without it. This convention is not a diagnostic rule for real people. Change brows and mouth separately and compare previews, then verify fixed ears, accessory and head geometry against the baseline.

## Give the laughing mouth appropriate space

Simplify eyes into curves if useful and enlarge the open mouth without crossing the chin. Reassess nose, eyes and lower-face space instead of shrinking the whole face. Add teeth or tongue only when they help and preserve the family’s level of detail and line weight. Inspect mouth/face separation with solid fills. Hide labels and compare all three for a recognisable common character.

## Overlay fixed landmarks to detect identity drift

Overlay heads in a test copy with reduced comparison-layer opacity and inspect fixed contours, ears and accessory. Variable features should differ. Distinguish intentional head movement from accidental drift; this exercise keeps head pose fixed. Do not erase expression differences to force all features into alignment. Restore separate positions and normal opacity afterward, retaining a record of fixed landmarks and permitted changes for later additions.

## Deliver the expression family at real sizes

Retain all editable versions and the base head, exporting consistent canvas sizes, margins and descriptive revision names. Keep small-size simplifications separate. Test hair/accessory edges and filled eye objects on destination backgrounds. Reopen each file to catch missing identity features or overlay/test states. If used for product controls or statuses, design the actual textual explanation separately; faces alone should not be assumed to encode every business meaning.

## Restore fixed structure when identity diverges

For inconsistent identity, inspect fixed head, hair, eye spacing and accessory before adding matching colours. Improve expression through feature relationships rather than exaggerated mouth corners alone. Simplify fragile output details. Diagnose a heavy-looking version through line weight and filled area instead of global opacity. Repeat unlabeled comparison and fixed-landmark overlay as separate checks of identity continuity and expression variation.

## Related tutorials and sources

[VectorPaint geometric portrait: describe features and hair with a few shapes](<https://allluckly.cn/en/guides/vectorpaint-geometric-portrait/>), [VectorPaint layers and groups: make complex work easier to select and edit](<https://allluckly.cn/en/guides/vectorpaint-layers-groups/>).

Sources checked on November 3, 2026: VectorPaint 1.0.8, requiring iOS / iPadOS 17 or later. Capability scope follows the [official App Store description](<https://apps.apple.com/cn/app/id1587272880>) and [Apple public version lookup](<https://itunes.apple.com/lookup?id=1587272880&country=cn>). These are exercises and checks you can perform, without a claim of step-by-step physical-device testing or verification in every target environment. Example content is instructional; follow the controls in your installed version.

