Screen 2.4: Voice Calibration

Prototype

Does the captured output actually sound like Aaron?

A/B Testing: Side-by-Side Compare

Aaron is quality checking. Show captured vs known samples. Let him tune. Critical: If output doesn't sound like him, trust erodes.

A: Side-by-side reviewB: Quick A/B picksC: Slider controls
Journey Phase: Cautious Exploration (Phase 2) • Aaron's State: Quality checking • Success Signal: "That actually sounds like me."
🎚️

Does this sound like you?

I've been capturing how you talk. Let's make sure I'm getting it right.

Design Process

What you said:

“The thick marker thing is about commitment. A fine-point pen lets you noodle around forever. A fat marker? You've got to make a decision.”

How Lex captured it:

“I use thick markers because they force commitment. You can't be precious with a fat marker — every stroke counts. It's about making decisions, not decorating.”

Version A Design Notes:

  • Side-by-side comparison — original vs captured
  • Three options: approve, reject, or tweak
  • Sequential through all samples
  • Pros: Comprehensive calibration
  • Cons: Can feel tedious

Version A

Detailed comparison — approve, reject, or tweak each.

Risk: Tedious

Version B

Quick binary choices — faster calibration.

Risk: Less precise

Version C

Direct slider control — maximum user agency.

Risk: Overwhelming