Voice Enrollment and Profiles

Voice Enrollment lets FluidAudio use speaker identity to decide who can trigger phrases and actions.

When a profile is enabled, FluidAudio compares live audio against stored or discovered voice characteristics and filters what it accepts as valid triggers.

Only Respond to Enrolled Speaker

The Only Respond to Enrolled Speaker toggle enforces strict identity gating.

• When enabled, the app only responds to commands from an enrolled speaker profile.
• When disabled, actions can trigger from any speaker that passes the phrase model’s normal confidence checks.

Best use:

• Turn it on when you want security and reduce accidental activations.
• Keep it off when you want maximum flexibility in open spaces or testing scenarios.

Profiles List

None

None disables enrollment-based filtering.

• No speaker identity matching is used.
• Triggering is fastest and broadest.
• Good for demos, shared listening, and environments where “anyone can trigger.”

Auto

Auto uses automatically derived speaker context without a manual profile. When starting up a Listen session, auto will take an audio sample for the first several seconds. This profile will then be used for the duration of the Listen session.

• Helps improve consistency without manual setup.
• Useful when you do not want to manage named profiles.
• Better than None when you want a lightweight voice preference but still need broad usability.

User-generated Profile

A user-generated profile is a manually created voice identity.

• Uses your recorded or imported voice sample.
• Best for one primary user, or a known trusted speaker.
• Most selective and stable when Only Respond to Enrolled Speaker is enabled.

Best Practices for Choosing a Profile Option

Use None if:

• You want the fewest setup steps.
• Trigger reliability matters more than speaker gating.
• There is no single trusted speaker.

Use Auto if:

• You want better practical filtering without managing people.
• Speaker preferences can change over time.
• You want a moderate balance of convenience and control.

Use User-generated if:

• A specific person should reliably trigger actions.
• You want stronger protection against accidental commands.
• You need clear ownership for one-on-one workflows.

Creating a New Profiles

Create a Profile

1. Open Settings → Speech
2. Make sure that FluidAudio is selected as the Speech Engine
3. Tap Voice Enrollment
4. Tap New Profile
5. Enter a name for your profile
6. Record a clean sample with clear speech and stable volume or import an existing recording

Voice recordings should be 30 seconds or less.

Deleting Profiles

When a profile is selected, tap Delete Profile at the bottom of the screen to remove a profile when it is no longer needed.