dialogue-audio

To generate multi-speaker dialogue audio for a podcast, this Skill uses the Dia TTS model to synthesize up to two distinct speaker voices from tagged text inputs. It structures conversation patterns and controls pacing through punctuation to produce realistic conversational flows.

Yes, you can control emotion and pacing in text-to-speech voiceover generation by using specific punctuation, non-speech cues, and sentence structuring within the Dia TTS model. These text formatting techniques fine-tune the delivery and emotional tone of the synthesized speech.

The maximum number of speakers supported for audio generation is two. The Dia TTS model synthesizes distinct voices for up to two speakers using specific speaker tags within the text input to differentiate the dialogue.

To normalize volume and merge audio segments after TTS generation, this Skill provides post-production guidance for integrating background music and combining multiple dialogue parts. These instructions help achieve consistent audio quality across the final conversation output.

Yes, the Dia TTS model supports non-speech sounds in dialogue synthesis by interpreting specific text-based cues. This feature allows the generation of realistic audio that includes natural conversational elements alongside the spoken voiceover lines.