Make transcript speakers distinguishable without colour

Given [AUDIO TRANSCRIPT] and [KNOWN SPEAKERS], format a readable transcript with stable speaker labels, turn breaks, relevant non-speech sounds, and uncertainty markers for unidentified voices. Provide a legend that works in monochrome and screen readers. Do not assign identities from voice alone or invent inaudible words. If speakers overlap or a turn is ambiguous, retain the uncertainty and request audio verification rather than forcing attribution. Check that every turn has an attributable or explicitly unknown speaker and that formatting preserves chronological order.