AI Audio Troubleshooting

This guide covers common issues with the AI Audio Talkdown feature and their solutions.

Quick Diagnostics

Run these checks first:

Check	Command/Action	Expected
Audio Enabled	Anava Console → Device → Features	"Audio Talkdown: Enabled"
Speaker Working	Camera web UI → Audio → Test	Hear test tone
MQTT Connected	Device status in Console	"Connected"
ACAP Version	Device info in Console	v3.0 or later

Common Issues

Audio Not Playing

Symptoms:

• Events trigger but no audio plays
• Console shows "Audio sent" but camera is silent

Solutions:

1. Check speaker configuration

   Camera Web UI → Audio → Output
   - Verify "Audio output" is enabled
   - Check volume is not 0
   - Test with built-in audio test

2. Verify ACAP audio permissions

   Camera Web UI → Apps → Anava Agent → Settings
   - "Audio Output" permission must be granted

3. Check audio format compatibility

   • Some older cameras require μ-law encoding
   • Contact support if camera model is pre-2020

Audio Delayed (more than 3 seconds)

Symptoms:

• Audio plays but significantly after the event
• Noticeable lag between detection and announcement

Causes & Solutions:

Cause	Solution
Network latency	Check camera's network connection (ping < 100ms to gateway)
MQTT backlog	Verify MQTT connection is stable, check for reconnects in logs
Cloud processing	Typically < 1.5s; if consistently slow, contact support
Jitter buffer	Expected 200-400ms buffering for smooth playback

Audio Cuts Off or Stutters

Symptoms:

• Announcement starts but doesn't complete
• Audio has gaps or repeats

Solutions:

1. Network bandwidth

   • Audio requires ~32 kbps sustained
   • Check for network congestion during events
   • Verify camera isn't on a saturated WiFi network

2. Increase jitter buffer (via ACAP config)

   {
     "audio": {
       "jitterBufferMs": 400
     }
   }

3. Check for packet loss

   • View device logs for "audio chunk timeout" messages
   • Consider wired connection if on WiFi

Wrong Language or Voice

Symptoms:

• Audio plays in unexpected language
• Voice doesn't match configuration

Solutions:

1. Verify group settings

   Console → Device Groups → [Your Group] → Audio Settings
   - Language: [Should match expected]
   - Voice: [Should match expected]

2. Check device override

   • Individual devices can override group settings
   • Console → Devices → [Device] → Audio Settings

3. Clear audio cache (if recently changed settings)

   • Restart ACAP via camera web UI
   • Or wait 5 minutes for settings to propagate

No Audio After Firmware Update

Symptoms:

• Audio was working, stopped after camera firmware update
• ACAP shows running but audio features fail

Solutions:

1. Reinstall ACAP

   • Camera firmware updates can reset app permissions
   • Download latest ACAP from Console
   • Reinstall and grant audio permissions

2. Check VAPIX API availability

   GET http://{camera}/axis-cgi/audio/info.cgi

   Should return audio capabilities, not 404

Audio Too Quiet or Too Loud

Symptoms:

• Audio plays but can barely hear it
• Audio is distorted due to volume

Solutions:

1. Adjust in Console

   Console → Device Groups → Audio Settings → Volume: 1-100

2. Adjust camera output gain

   Camera Web UI → Audio → Output Gain

3. Check speaker hardware

   • External speakers may have their own volume control
   • Verify speaker is properly connected

Diagnostic Commands

View Audio Events (Console)

Navigate to: Console → Events → Filter: Audio

Shows:

• Audio generation requests
• Delivery confirmations
• Playback status

Device Logs

Access via Console → Devices → [Device] → Logs, or directly:

Camera Web UI → System → Logs → Application Log
Filter: "anava" or "audio"

Look for:

[INFO] TTS: received audio chunk 1/8
[INFO] TTS: playback started
[INFO] TTS: playback complete (3.2s)
[ERROR] TTS: chunk timeout, using fallback

Test Audio Manually

From Console → Devices → [Device] → Actions → Test Audio

This sends a test announcement to verify the full pipeline.

Error Codes

Code	Meaning	Action
AUDIO_001	Speaker not available	Enable audio output in camera settings
AUDIO_002	Generation timeout	Retry; if persistent, check cloud status
AUDIO_003	Delivery failed	Check MQTT connection
AUDIO_004	Playback failed	Verify VAPIX audio API working
AUDIO_005	Rate limited	Wait 30 seconds between audio events
AUDIO_006	Feature not enabled	Enable Audio Talkdown in group settings

Performance Baseline

Use these as reference for normal operation:

Metric	Normal Range	Investigate If
Generation time	800-1500ms	> 2000ms consistently
Delivery time	100-300ms	> 500ms consistently
Playback start	1.2-2.0s total	> 3.0s consistently
Chunk loss	0-1%	> 5%

Getting Support

If issues persist after troubleshooting:

1. Collect diagnostics

   • Console → Devices → [Device] → Download Diagnostics
   • Include last 24 hours of logs

2. Note the following

   • Camera model and firmware version
   • ACAP version
   • Time and date of issues
   • Network configuration (wired/WiFi)

3. Contact support

   • Email: support@anava.ai
   • Include diagnostic bundle and notes

Related Documentation

• Overview - Feature introduction
• How It Works - Technical architecture

---

Last updated: December 2025