AivisSpeechInstall

Summary

Convert multiple text inputs into a single concatenated speech audio using base and per-text voice parameters.

Inputs

• ベース話者ID: speakerId (required) — Base speaker style ID used when no individual speaker ID is specified for a text.
• ベース音声設定: baseAudioParams — Common base voice parameters to apply to all texts unless individually overridden, including speed, pitch, intonation, volume, silent padding, pause length (ignored), and tempo dynamics.
• バイナリプロパティ名: binaryPropertyName — Name of the binary output property containing the resulting audio WAV data, for referencing in subsequent nodes.
• 入力方式: inputMode (required) — Input mode selection for the text list: 'GUI入力' for manual entry or 'JSON入力' for JSON array input.
• テキスト一覧: texts (required) — When using GUI input, a fixed collection list of texts to be synthesized, each optionally overriding voice parameters.
• テキスト一覧 (JSON): textsJson (required) — When using JSON input, a JSON array of objects containing text (required) and optional voice parameter overrides for each segment.
• ステレオ出力: outputStereo — Whether to output audio in stereo (true) or mono (false, default).

Output shape

a single audio file in WAV format containing all input texts concatenated as one audio track

The output audio is returned as binary data under the configurable binary property name (default 'data'). Each input text's voice parameters are merged from baseAudioParams and per-text overrides before synthesis. Texts are synthesized sequentially, their audio buffers concatenated server-side. Errors during synthesis or concatenation throw node errors. The final WAV audio can be referenced in subsequent nodes via $binary property.

Links

• n8n community nodes installation guide