A Series of Avatars in One Style: For a Team and for Yourself

6 min

A coherent set of avatars solves team recognisability. How to produce a set that looks like one piece of work.

A set of avatars in one style appears in two scenarios: a team wanting a recognisable visual language, and one person needing variants for different platforms. In both cases the task is the same — coherence.

What makes a set cohere

  1. The same background: colour, tone, character.
  2. The same light: direction and softness.
  3. The same framing: face scale within the circle, head position.
  4. The same processing: contrast, saturation, grain.
  5. Coordinated clothing: not a uniform, but a shared register and palette.

Of these five the most conspicuous is the background. Different backgrounds destroy the sense of a set faster than any other divergence.

A practical scheme for a team

A workable sequence requiring no studio.

  1. Write a shooting brief for the sources: daylight from a window, facing it, phone on a support at eye level, chest-up framing, a burst of ten frames.
  2. Collect sources from everyone.
  3. Select one best frame per person.
  4. Bring them to a common look: replace the background, align light and colour, crop identically.
  5. Check the whole set side by side at small size.

Bringing them to a common look

Manually processing a dozen portraits into exact agreement of background and light is laborious. Generative alignment does it faster: the source supplies the face, the scene description supplies background, light and clothing.

A condition for correctness: the scene description must be word-for-word identical for everyone. Any variation in wording produces divergence in the result.

A second condition: every face must remain recognisable. Alignment makes a set coherent; it does not make people resemble one another.

Colour coding

A device that works in team sets: one background with different shades of a single colour. The set looks unified while each avatar remains distinguishable in peripheral vision.

In practice: define a palette of five or six close shades and assign them. When new people join, the palette extends in the same direction.

A set for one person

The second scenario: several avatar variants for different platforms — business, neutral, informal. The task is the inverse of the team one: the variants should differ in register while being recognisable as one person.

The same rule of facial consistency applies: shoot every variant in one session, change background, clothing and processing, but not appearance.

What spoils a set

  • Avatars added later outside the common scheme.
  • Different degrees of retouching: one person's skin smoothed, another's not.
  • A mix of photographs and stylised images.
  • Different framing: one face large, another showing shoulders.
  • Different shooting periods, if hairstyles and styling changed in between.

Maintaining a set

Teams change, and a set needs upkeep. It is practical to record everything needed to reproduce it: the shooting brief, the exact scene description, processing parameters, the background palette.

Without that record, new avatars six months later will differ and the set gradually falls apart — the same mechanism as with corporate portraits.

Frequently asked

What matters most for a coherent avatar set?

The same background. Different backgrounds break the sense of a set faster than anything else — it is the first thing noticed when viewing them in sequence.

How do I build a set for a distributed team?

Issue one shooting brief, collect the sources, and bring them to a common look with a word-for-word identical scene description while preserving faces.

What should I record for adding new people later?

The shooting brief, the exact scene description, processing parameters and the background palette. Without it, new avatars six months later will not match.

  • #аватар
  • #команда
  • #серия

Try it in the studio

Upload a photo and pick a style — one step from prompt to result.

Open the studio

Read next

Try it on this topicAI Avatar Generator

AI avatarsAll articles