Security Cameras & NVR
How Much Storage Do Security Cameras Need?
Learn how bitrate, camera count, recording schedule, retention, codec, resolution, and activity affect NVR storage planning.
Security-camera storage is a data-rate calculation, not a camera-count guess. Two eight-camera systems can require very different disk capacity because resolution, frame rate, codec, scene activity, image quality, audio, recording schedule, and retention all change the amount of video written.
Key takeaways
- ✓Start with measured or conservatively estimated bitrate per camera.
- ✓Multiply by recording hours, retention days, and camera count.
- ✓Add operating margin instead of sizing disks to the exact theoretical result.
- ✓Confirm usable array capacity, not the drive manufacturers’ raw total.
The basic storage formula
For continuous recording, storage is driven by average bitrate. A convenient estimate is: gigabytes per day ≈ bitrate in megabits per second × 10.8. Multiply that daily value by the number of cameras and retention days.
Example: a camera averaging 4 Mb/s produces about 43.2 GB per day. Eight cameras at that rate produce about 345.6 GB per day, or roughly 10.4 TB over 30 days before filesystem, array, database, export, and safety margins.
Why resolution alone does not determine storage
Resolution is only one input. A busy parking lot, trees moving in wind, rain, low-light noise, or a high-detail warehouse can require more bits than a quiet hallway at the same resolution. Frame rate, compression codec, keyframe interval, quality target, and whether the camera uses constant or variable bitrate also matter.
The best estimate uses representative camera settings and observed bitrate from scenes similar to the installation. When measurements are unavailable, use conservative assumptions and preserve room for change.
- Include audio when cameras record it.
- Account for continuous versus motion- or event-based recording.
- Plan for higher nighttime bitrate when image noise increases.
- Reserve capacity for indexes, thumbnails, clips, and operational headroom.
Raw capacity is not usable retention capacity
Disk labels show raw capacity. RAID or erasure protection, filesystem formatting, recorder databases, reserved free space, and failed-drive rebuild needs reduce what can safely hold video. A recorder also performs better when it is not operated permanently at 100 percent full.
Retention should be verified after installation by comparing actual daily writes with available archive capacity. Seasonal activity or later camera-setting changes can shift the result, so retention is something to monitor rather than promise once and forget.
Plan storage around evidence requirements
Start with the business question: how far back must staff be able to review an incident, and what image detail is required to make that footage useful? Entry identification, cash transactions, parking coverage, overview video, and license-plate capture may need different cameras and recording profiles.
A practical design may use different bitrates and retention priorities by camera instead of treating every stream identically.
Frequently Asked Questions
Direct answers
How many days can an NVR record?
Retention depends on usable storage divided by the actual daily video written by all cameras. Camera count alone is not enough to calculate it.
Does H.265 always cut storage in half?
No fixed percentage is guaranteed. Savings depend on the scene, camera implementation, settings, motion, and the H.264 configuration used for comparison.
Should an NVR be sized from raw drive capacity?
No. Use protected, formatted, operationally available capacity and preserve margin for databases, exports, rebuilds, and bitrate variation.