fix: voice input recording format + ASR logging

- Change recorder format from mp3 (8kHz) to aac (22kHz) for better
  whisper ASR accuracy
- Move recorder.onStop binding before recorder.stop() to prevent
  race condition (onStop may fire before handler is registered)
- Backend: add Logger, log ASR errors instead of silent catch {}
- Backend: change default extension from .mp3 to .aac

backend/src/modules/tts/tts.controller.ts

@@ -1,4 +1,4 @@
-import { Controller, Get, Post, Body, Param, Res, HttpException, HttpStatus, UseGuards, UploadedFile, UseInterceptors } from '@nestjs/common'
+import { Controller, Get, Post, Body, Param, Res, HttpException, HttpStatus, UseGuards, UploadedFile, UseInterceptors, Logger } from '@nestjs/common'
 import { FileInterceptor } from '@nestjs/platform-express'
 import { Response } from 'express'
 import * as fs from 'fs'
@@ -10,6 +10,7 @@ import { Public } from '../../common/decorators/public.decorator'
 
 @Controller('tts')
 export class TtsController {
+  private readonly logger = new Logger(TtsController.name)
   constructor(private ttsService: TtsService) {}
 
   @UseGuards(JwtAuthGuard)
@@ -42,7 +43,7 @@ export class TtsController {
     if (!file) throw new HttpException('请上传音频文件', HttpStatus.BAD_REQUEST)
     const uploadDir = '/tmp/asr_uploads'
     if (!fs.existsSync(uploadDir)) fs.mkdirSync(uploadDir, { recursive: true })
-    const ext = path.extname(file.originalname) || '.mp3'
+    const ext = file.originalname ? path.extname(file.originalname) || '.aac' : '.aac'
     const dest = path.join(uploadDir, file.filename + ext)
     fs.renameSync(file.path, dest)
     try {
@@ -63,7 +64,11 @@ export class TtsController {
       if (whisperResult && whisperResult.trim()) {
         return { text: whisperResult.trim() }
       }
-    } catch {}
+    } catch (e: any) {
+      this.logger.error(`ASR failed: ${e?.message || e}`)
+    }
+    // 清理临时文件
+    try { fs.unlinkSync(dest) } catch {}
     return { text: '' }
   }
 }