admin/Mdeical_Sur_Report
1
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 4 Wiki Activity

Stabilize speech audio capture

Browse Source 
- Keep Web Audio source and processor nodes in refs so microphone processing is not lost during speech recognition.

- Explicitly resume the AudioContext before sending PCM frames to the speech WebSocket proxy.

- Disconnect audio nodes and clear speech state when listening stops or the WebSocket closes.

- Show a clear prompt when audio frames were captured but no recognizable text was returned.

- Update progress documentation for the speech capture hardening.
This commit is contained in:
admin
2026-05-02 06:20:48 +08:00
parent 5d936832da
commit 87ab7d4b9c
2 changed files with 39 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
docs/progress.md
View File

@@ -93,3 +93,4 @@
| 2026-05-02 | 模板 HTML 导出包补充模板字段和字段管理设置，导入时恢复字段库元数据。 |
| 2026-05-02 | 修复报告编辑器加载已有 AI 区域后下拉栏初始显示“无可用 AI 区域”的问题。 |
| 2026-05-02 | 调整抽帧百分比为两位小数保序保存；自动截图按时间顺序执行，自动插入按配置顺序执行。 |
| 2026-05-02 | 加固报告编辑器语音采集，保留 Web Audio 节点引用、显式恢复 AudioContext，并在无识别文本时给出提示。 |