6 A CCTV storage calculation starts with one number, the bitrate of each camera, and ends with a capacity figure that can easily run to petabytes for a large estate. Between those two points sit choices about resolution, codec, frame rate, recording mode, retention and protection that can change the answer by an order of magnitude. Getting the calculation right before buying storage avoids running out of retention mid-year or paying for capacity that will never be used. This article explains the formula, the variables that matter most, a worked example and the overheads that are often forgotten. For the broader architecture, see our hub on video surveillance storage. The basic formula Storage per camera depends on how many bits per second it produces and how long it records: Daily storage per camera (GB) = bitrate (Mbps) x seconds recorded per day / 8 / 1,000. For a camera recording continuously, there are 86,400 seconds in a day. That simplifies to: Daily storage (GB) = bitrate (Mbps) x 10.8. So a camera streaming at 4 Mbps continuously produces about 43 GB per day. Over 30 days, that is about 1.3 TB per camera. Multiply by the number of cameras and retention days to get the total. The variables that drive bitrate Resolution Higher resolution means more pixels per frame and generally higher bitrate. Moving from 1080p to 4K roughly quadruples the pixel count, though codec efficiency means bitrate does not always rise proportionally. Codec Compression standards make a large difference. H.265 (HEVC) typically achieves similar quality to H.264 at a substantially lower bitrate, often cited as 30 to 50 percent less, depending on scene and encoder. Some camera vendors add proprietary smart codecs that reduce bitrate further in static scenes. Frame rate Recording at 30 frames per second produces more data than 15 or 10. Many surveillance deployments record at 10 to 15 fps for general coverage and higher rates only where needed, such as cash handling or license plate capture. Scene complexity Busy scenes with lots of motion, trees moving in wind, rain, or nighttime noise produce higher bitrates than static indoor scenes, especially with variable bitrate encoding. A camera that averages 2 Mbps indoors might average far more outdoors at night. Quality settings Compression level and variable versus constant bitrate settings affect bitrate. Constant bitrate is predictable for sizing; variable bitrate saves space on quiet scenes but needs careful averaging. As a very rough guide, 1080p cameras often run in the low single-digit Mbps range and 4K cameras several times higher, but measure your own cameras rather than relying on rules of thumb. Recording mode How often cameras record matters as much as bitrate: Continuous recording captures everything and is simplest to size. Motion or event-based recording writes only when activity is detected, which can cut storage dramatically in quiet areas but is harder to predict. Scheduled recording records at full quality during set hours and reduced quality or not at all otherwise. Dual streaming records a high-resolution stream for evidence and a lower-resolution stream for live viewing or long-term retention. For motion-based recording, estimate the percentage of time each camera will record. A busy entrance may record most of the day; a storeroom only a few percent. Retention Retention multiplies everything. Thirty days of footage requires twice the capacity of fifteen. Retention is usually set by policy and law, and may differ by camera group or site. Some organizations keep recent footage at full quality and older footage at reduced quality or frame rate. A worked example Consider an illustrative campus with 1,000 cameras: 700 cameras at 1080p using H.265 at an average of 3 Mbps, recording continuously. 300 cameras at 4K using H.265 at an average of 8 Mbps, recording continuously. Retention of 30 days for all cameras. Daily storage: 1080p cameras: 3 Mbps x 10.8 = 32.4 GB per camera per day. x 700 = about 22.7 TB per day. 4K cameras: 8 Mbps x 10.8 = 86.4 GB per camera per day. x 300 = about 25.9 TB per day. Total: about 48.6 TB per day. Retention: 48.6 TB x 30 days = about 1.46 PB of footage. Aggregate write bandwidth: 700 x 3 Mbps + 300 x 8 Mbps = 4,500 Mbps, or about 4.5 Gbps of continuous writes. If 40 percent of the 1080p cameras switch to motion-based recording that records half the time, their contribution drops accordingly, reducing total storage by roughly 10 percent. If retention rises to 60 days, capacity doubles to nearly 3 PB. Overheads that are often forgotten Storage protection RAID, erasure coding or replication consumes raw capacity beyond usable capacity. Depending on the design, protection overhead might be 20 to 50 percent or more. Free space and headroom Storage should not run at 100 percent. Keep headroom for rebuilds, growth and bursts, commonly 10 to 20 percent. Evidence and exports Footage exported or protected for investigations is kept longer than normal retention. Estimate how much footage is typically preserved and for how long. Secondary copies If footage is replicated to another site or archived to a second tier, count that capacity too. VMS overhead Databases, indexes, thumbnails and metadata use additional space, usually modest but not zero. Growth New cameras, upgrades to higher resolution and new sites add capacity over time. Plan at least three to five years ahead. Write performance is part of the calculation Capacity is only half the story. The storage must sustain the aggregate write bandwidth continuously, including during drive or node failures and rebuilds. In the example, that means 4.5 Gbps of steady writes, plus headroom, plus any read activity from investigators. Under-provisioned write performance causes dropped frames, which is worse than running out of capacity because the gap may not be noticed until footage is needed. Reducing storage without losing evidence Once the baseline is known, several levers can reduce capacity while keeping the footage that matters: Tune frame rate by purpose. General overview cameras rarely need 30 fps; reserving higher rates for cameras that capture fast movement or fine detail saves capacity. Use smart codecs carefully. Variable and scene-adaptive encoding can cut bitrate in static scenes, but test them at night and in bad weather. Separate retention by camera group. Cameras covering high-risk areas may justify longer retention than those covering corridors or car parks. Lower quality for older footage, where policy allows, by keeping full resolution for recent days and a reduced stream afterward. Protect only what is needed. Mark footage for long retention when an incident is reported, rather than extending retention for everything. Each change should be agreed with security operations and, where relevant, legal or privacy teams, since reduced quality or shorter retention may affect investigations. Tools and validation Many camera and VMS vendors provide storage calculators. They are useful starting points, but validate their assumptions: Measure actual bitrates from a sample of cameras in each environment, day and night. Check how the VMS stores data and what overhead it adds. Confirm recording modes are configured as assumed. Revisit calculations after deployment using real usage data. Checklist: CCTV storage calculation Inventory cameras by resolution, codec, frame rate and location. Measure real average bitrates, including at night. Define recording mode and duty cycle for each camera group. Apply retention per group. Calculate daily and total storage, and aggregate write bandwidth. Add protection overhead, headroom, evidence, copies and VMS overhead. Plan growth for new cameras and upgrades. Validate vendor calculators against measured data. Recalculate annually or after major changes. Putting it together A CCTV storage calculation is straightforward once you know each camera’s real bitrate: multiply by 10.8 for daily gigabytes, by retention days for capacity and by camera count for the estate. The hard part is getting realistic inputs and remembering the overheads: protection, headroom, evidence, copies and growth. Measure cameras in their real environments, size for both capacity and sustained write performance and choose storage that grows in small steps as the camera estate evolves. For architecture options, see NVRs or centralized storage for thousands of cameras. Frequently asked questions How much storage does one CCTV camera need? At 4 Mbps continuous recording, about 43 GB per day or 1.3 TB for 30 days. Actual needs depend on bitrate, recording mode and retention. How do I convert bitrate to storage? Multiply the bitrate in Mbps by 10.8 to get gigabytes per day for continuous recording. Does H.265 reduce CCTV storage? Yes. H.265 typically delivers similar quality at a substantially lower bitrate than H.264, often 30 to 50 percent less depending on the scene. How much does motion recording save? It depends on activity. Quiet areas may record only a small fraction of the day, while busy areas may record almost continuously. What overhead should I add to CCTV storage estimates? Add storage protection overhead, 10 to 20 percent headroom, evidence retention, secondary copies and growth. Further reading Video surveillance storage CCTV footage retention VMS archive to S3 object storage NVRs vs centralized storage Body-worn camera storage
