Variable Components
A component normally follows the corresponding source drawing for its host layer. A variable component can instead request a named source layer or a calculated designspace location. Element placement itself interpolates, as established in Interpolation and Compatibility. Use a variable component when the reused source design also needs a selected location or different proportions from its host.
Reuse across Masters
A variable component retains the identity of its source glyph and adds an explicit source-layer or instance choice. A plain layer name selects that drawing. A location such as #wt=650,wd=100 requests an interpolated source design at those axis values.
This is different from an element reference, which shares drawing content. A reference does not acquire an axis-location selector simply because the font has several masters. Use a variable component when the construction needs to ask a source glyph for a particular location; use references when shared content itself is the intended relationship.
Choose the component's source instance
- Select the component in the Glyph window.
- Click Select component instance in its property controls.
- Adjust the source-axis sliders and inspect the result.
- Check the component's placement and transform after choosing the source shape.
For exact or repeatable entry, expand the component properties in the Elements panel and edit Base layer name. Enter a source layer name or a #-prefixed location using the font's axis codes. Copy that field's value to another component when both should request the same source instance.
Reset to current layer restores the ordinary corresponding-layer relationship. Clearing an explicit choice and manually entering a particular layer name are not equivalent: the named choice continues to request that layer.
Build a superior figure
Suppose ordinary figures have compatible weight masters and you need a superior figure:
- Add a component of the source figure to the superior glyph.
- Choose a heavier source instance to compensate for the loss of apparent weight when reduced.
- Scale and position the component for the superior design.
- Adjust its width with an element transform or a source width axis when available.
- Check the small figure at its intended size and alongside the other superior figures.
The same method can help with subscripts or small caps. It supplies adjustable material, not automatic optical correction. Reduced figures often need different width, spacing, or detail from a uniformly scaled original.
After changing source drawings, use Font > Update Glyphs when dependent variable components need refreshing, then inspect the hosts. Do not judge a construction from a stale preview or assume that changing the source's current layer changes an explicitly selected instance.
Preserve the construction in the source
Variable components are a FontLab source feature. An output format may receive resolved contours rather than the editable instance relationship. UFO/DesignSpace and ordinary OpenType output can require decomposition of these components; exchange to .glyphs can lose the explicit location and fall back to ordinary source-layer components.
Keep a native source and inspect representative exported or exchanged glyphs; per-format handling is covered in OpenType, Variable, and Web Fonts and Interchange and Legacy Formats. Decomposition should be assessed by the resulting shape, while editable exchange must also preserve the intended source relationship. Do not infer round-trip support merely because the destination format supports some kind of component.
Glyph-Specific Axes
A source glyph can have local variation parameters that are not font-wide axes. For example, a serif library glyph might use le for length and th for thickness in its layer locations. Those glyph-specific axes control the reusable part without adding those sliders to the entire family's public designspace.
For a two-axis serif source, create four compatible layers and enter their locations in Layers & Masters > Location, for example:
| Source drawing | Location |
|---|---|
| Short, thin serif | le=30,th=40 |
| Long, thin serif | le=60,th=40 |
| Short, thick serif | le=30,th=60 |
| Long, thick serif | le=60,th=60 |
The numbers locate the designs; they do not draw or resize the serif automatically. Make the corresponding shape changes in those source layers. Then place the source as a component in a host and try #le=45,th=50 in Base layer name to request an intermediate part. Check its shape and fit before using it elsewhere.
Use the variable-component instance selector to choose the local values in each host. The selector can expose the source glyph's additional axes alongside the font-wide ones. Exact locations can also be entered through the component's #-prefixed field.
A component’s location control can select a position along its available axis. The shown Weight value belongs to this example.
Use distinct short codes and inspect the resulting axis names. Avoid collisions with the font's codes or with another local parameter; similar descriptive names can otherwise resolve to the same abbreviated identity.
For a modular part used only as construction material, turn off the source glyph's export. The hosts still need its drawing resolved correctly in the final output. Test a representative host before building a large library around the relationship.
This local model is useful when many glyphs need related but individually tuned serifs, terminals, or stems. A global edit to the source changes the family of available parts, while each component retains its requested location and placement.
Auto-Layers
Automatic construction belongs to each glyph layer. In a multi-master font, an auto layer can rebuild from the source designs appropriate to that master. The same glyph can therefore have a live construction in several masters without requiring identical resulting outlines.
Copying a recipe can regenerate target layers, so first check for deliberately independent constructions that you need to preserve. Use Copy recipe to masters in the Layers & Masters recipe controls when the construction rule should be shared:
- Check the current master's recipe and its generated result.
- Use Copy recipe to masters to copy the current recipe to all other master layers of this glyph; this is not a per-target selection control.
- Inspect their automatic-layer state as well as the recipe text.
- Check source coverage, anchors, component locations, and metrics in every target.
The operation propagates construction settings as well as recipe text.
A common recipe does not guarantee a compatible result. Missing source layers, different source structures, or different component/anchor decisions can still make the generated masters diverge. Check interpolation after updating the source glyphs or copying recipes, and keep local exceptions explicit.