VocalStack Logo

Documentation

获取转录数据

从待处理或已完成的转录中获取数据。 这包括转录时间线、关键字、摘要和段落段。

从麦克风或 LiveStream 转录

从麦克风或直播流转录实时演讲。 与 Polyglot 集成,为用户可以用任何语言阅读的转录创建公共可共享链接。

转录会议

通过会话监控和管理转录状态。 使用会话,您可以重新连接到之前创建的异步连接。

翻译一个转录

从 URL 转录音频

将 URL 中预先录制的音频转录为纯文本。 支持主要文件格式,包括 MP3、WAV、FLAC 和 OGG。

客户端身份验证令牌

为客户端请求创建临时身份验证令牌。 在 Web 浏览器中安全实现 API 请求,而不会暴露您的 API 密钥。

转录请求和回复

所有转录操作的常见请求选项和响应。 使用选项配置转录设置。

转录和演示多语种会议

创建一个会话,可用于通过公共可共享链接广播实时转录。 用户可以用他们首选的语言阅读实时转录,甚至在您的会话处于闲置状态时,也可以阅读过去的转录。

翻译一个转录

浏览文档
将转录的文本翻译成另一种语言。 这可以对任何转录进行,包括预录制的转录、现场转录或多语种会话转录。.
You can use the VocalStack API to translate any transcription, whether it is finished or still processing. If you have not yet started transcribing audio, you can do so with any of the following methods:
从 URL 转录音频
将 URL 中预先录制的音频转录为纯文本。 支持主要文件格式,包括 MP3、WAV、FLAC 和 OGG。.
从麦克风或 LiveStream 转录
从麦克风或直播流转录实时演讲。 与 Polyglot 集成,为用户可以用任何语言阅读的转录创建公共可共享链接。
转录和演示多语种会议
创建一个会话,可用于通过公共可共享链接广播实时转录。 用户可以用他们首选的语言阅读实时转录,甚至在您的会话处于闲置状态时,也可以阅读过去的转录。.

Async Method

If you need to load your translations as soon as they become available then you will want to listen to the translation request asynchronously:
JavaScript
import { Transcriptions } from '@vocalstack/js-sdk';

const sdk = new Transcriptions({ apiKey: 'YOUR-API-KEY' });

const translation = await sdk.addTranslationAsync({
  id: 'TRANSCRIPTION-ID',
  language: 'de',
});

translation.onData((response) => {
  // 'waiting', 'processing', 'done', or 'error'
  console.log(response.status);

  // the translated timeline
  console.log(response.data?.timeline);
});

Fire and Forget Method

This method for translating simply sends the translation request, but does not wait for a response. This method can be useful if:
  • You don't need the translation right now, but you want to cache it for quick access in the future. (see Performance Caching)
  • There is a transcription in a live Polyglot session that you're already monitoring in another process, and you simply want to add a new translation to this transcription.
JavaScript
import { Transcriptions } from '@vocalstack/js-sdk';

const sdk = new Transcriptions({ apiKey: 'YOUR-API-KEY' });

// This is a synchronous request, so we cannot listen for the response
sdk.addTranslation({ id: 'TRANSCRIPTION-ID', language: 'de' });

Performance Caching

As translations execute on the backend, the persisted transcription data is updated to include the new translations. The persisted transcriptions get updated with each timeline segment translated, so API calls requesting transcription data will always return timeline objects with the most recent translations available.
This also means that you will only need to issue one translation request per language. (Additional requests will have no effect as the translations are already persisted.)
Scroll Up