Seed Audio 1.0: Text to Audio
Input a text prompt to generate speech, dialogue, background music, and sound effects in one audio file. Describe voices, emotion, and scene details to create multi-speaker audio up to 2 minutes.
简介
This ComfyUI workflow turns a single text prompt into a complete audio scene—speech, dialogue, background music, and sound effects—in one pass. It centers on the ByteDanceSeedAudio node (Seed Audio 1.0, text-only), which interprets your description to synthesize multi-speaker audio up to roughly two minutes without any reference audio. The MarkdownNote node in the graph provides on-canvas guidance, and SaveAudioAdvanced handles exporting the generated waveform to disk.
你需要提供
- 正在读取工作流
在 ComfyUI 中运行。 此工作流无法在本页面运行。你可以在 Comfy Cloud 中打开（已预先加载所需内容），或下载后在自己的 ComfyUI 中使用。
在 Cloud 中试用常见问答
Write clear speaker turns and brief stage directions in-line. For example: “Narrator (calm, warm): Welcome to the market. Crowd ambience, light acoustic guitar. Vendor (cheerful): Fresh oranges!” Use short, unambiguous sentences and specify emotions, accents, and pacing. This helps the model separate voices and time events more reliably.
This workflow is designed for text-only generations up to roughly two minutes. For longer pieces, break your script into scenes, generate them separately, and stitch the results in a DAW for seamless transitions.
The model mixes elements automatically based on your prompt. You can nudge balance by stating preferences (e.g., “music low, dialogue forward, subtle city ambience”). After generation, use SaveAudioAdvanced output and apply standard post-production tools (compression, EQ, gain) if you need precise level control.
If your ByteDanceSeedAudio node uses a remote API, you’ll need to configure credentials as documented by the node’s repository or settings. If a local backend is supported, follow the node’s installation instructions for required assets. Check the node’s README for the exact setup steps.