Orpheus Voice Lab

← dream

Three-way Ghost voice comparison: Gemini TTS global style (A) · Gemini TTS experimental (B) · Orpheus TTS phrase-level emotion tags (C). Wear headphones.

“The resonance here is ancient. Let yourself be absorbed by it.”

AGemini TTS · global style direction
no audio yet
BGemini TTS · experimental style
no audio yet
COrpheus TTS · phrase-level emotion tags
no audio yet

Which voice best captures the Ghost character for this scene?

Orpheus emotion tags (column C only):

<reverent> <whispers> <sad> <fearful> <happy> <excited> <surprised> <disgusted>

Wrap a word or phrase: The <reverent>resonance</reverent> here is ancient.

Gemini (A/B) applies style globally — the whole sentence gets the same character. Orpheus (C) shifts tone word-by-word: one word can whisper while the next is fearful.

Gemini TTS · fal-ai/gemini-tts · Orpheus TTS · fal-ai/orpheus-tts · FAL_KEY · ~$0.01–0.02/rowdesign notes: src/app/dream/61-orpheus-voice/README.md
The dream lab — Karel Barnoski