Glyph vs Character: What Changes When You Make a Font?

Type a lowercase a, then switch between two installed fonts. The text remains the same character even though its appearance changes. One font might use a single-storey a and another a double-storey design. Those are different glyph designs representing the same character in this example.

Uppercase A and lowercase a, however, are different characters. Drawing them alike does not merge their identities. This matters when a handmade alphabet has similar-looking capitals, lowercase letters, or digits.

The Unicode character and combining-mark FAQ↗ (opens in a new tab) explains distinctions between abstract characters and their visual representation. Think of the text as the instruction and the glyph as the shape the font supplies for that instruction.

A useful font project needs a drawing, a character assignment, and spacing information. The mapping tells the application which glyph to use for a code point. The OpenType cmap specification↗ (opens in a new tab) describes the character-to-glyph mapping table used by the font format.

For a practical example, draw a distinctive digit zero with a slash. Assign it to 0, not to uppercase O. Then type both characters in a test line. Their similar silhouettes make this a good early check of whether you are editing pictures alone or building a correctly mapped font.

A single drawing per character is a useful starting point, but it does not reproduce every behavior of natural handwriting. A person may draw the same letter differently at the beginning and end of a word. Sophisticated fonts can use alternate glyphs and shaping rules, but those features require explicit support and design.

Do not assume that an AI-generated alphabet automatically includes contextual alternates, cursive joining, or every accented form. In MakeFont, inspect the selected character set and the actual preview instead of inferring capabilities from the sample artwork.

The same caution applies to an accented letter. A visibly correct example in one word does not prove that every character sequence used by an application will render identically.

Create a small specimen containing Aa, O0, Il1, and the accents your language requires. Copy the specimen into a plain text field so that you know which characters it contains. Compare the preview with the downloaded font in another application.

If the right character has an awkward shape, edit its outline. If a different character appears, inspect the assignment. If nothing appears, inspect coverage. Keeping those questions separate makes troubleshooting faster and prevents redesigning a perfectly good glyph to solve a mapping problem.

• Unicode character and combining-mark FAQ↗ (opens in a new tab)
• OpenType cmap specification↗ (opens in a new tab)

These references explain the relevant formats, software behavior, or concepts. They do not endorse MakeFont or guarantee the quality of a generated font.