Skip to main content

← Back to Blue Iris

Blue Iris – Tuning Guide

🔹 Overview​

Tuning is the process of refining a working Blue Iris system — not fixing a broken one.

This page focuses on how to think about tuning, not just what settings to use.

🎯 Objective​

  • Improve system stability
  • Reduce false alerts
  • Maintain usable video evidence
  • Keep storage growth predictable
  • Balance quality against performance

🧠 Key Principle​

Tune in this order:

  1. Camera / Image
  2. Motion
  3. Recording
  4. AI
  5. Alerts

Skipping this order creates instability and poor results.

🛠️ Tuning Workflow​

Step 1 — Image & Stream Quality​

  • Confirm camera is producing a clean, stable image
  • Validate resolution, FPS, and stream type
  • Avoid unnecessary re-encoding

Step 2 — Motion Cleanup​

  • Define zones intentionally
  • Reduce noise (trees, shadows, insects, reflections)
  • Tune object size and contrast

👉 Goal: clean, reliable triggers before AI

Step 3 — Recording Behavior​

  • Confirm direct-to-disk where appropriate
  • Validate pre-trigger and post-trigger timing
  • Ensure clips reflect real events

Step 4 — AI Confirmation​

  • Enable AI only after motion is stable
  • Use moderate confidence thresholds
  • Confirm AI is analyzing a useful image

👉 AI improves signal — it does not fix bad input

Step 5 — Alerts​

  • Start simple
  • Validate event quality first
  • Add notifications only after system is stable

🛠️ Core Tuning Areas (Reference)​

Stream Strategy​

  • Main vs sub-stream usage should match the goal
  • Avoid unnecessary processing

Codec Choice​

  • H.264 is the safest baseline
  • H.265 may increase complexity and instability

Frame Rate​

  • ~15 FPS is a strong baseline
  • Increase only where justified

Resolution​

  • Match resolution to the scene
  • Placement often matters more than megapixels

Storage​

  • Validate retention rules early
  • Monitor growth after any major change

Performance​

  • Monitor CPU, GPU, storage, and memory together
  • Scale gradually — test after each change

🧠 System Insight​

Better image → better motion → better AI → better alerts

Most problems originate earlier in the chain.

⚠️ Common Mistakes​

  • Tuning AI before motion works
  • Changing multiple variables at once
  • Increasing FPS or resolution without purpose
  • Enabling alerts too early
  • Ignoring system performance impact

📊 Related Pages​

✅ Result​

A structured tuning approach that keeps Blue Iris stable, predictable, and scalable in real-world deployments.