Before you buy disks or upgrade a site's internet line, you need two numbers. How much storage will the recordings need? And how much bandwidth does each site need? This guide shows a simple way to estimate both. The results are estimates. Skyline confirms the real design during the site survey.
Before you start
- Count the cameras you will record at each site.
- Note the resolution of each camera type, such as 2 MP or 4 MP. MP means megapixels: how many million dots make up the picture.
- Decide how many hours per day you record, and how many days you keep.
The words you need
- Bitrate: how much data a camera sends each second.
- Mbps: megabits per second, the unit used for bitrate.
- Codec: the method used to compress video. H.264 is common. H.265 is newer and usually needs less bitrate.
- TB: terabyte, a unit of storage. One TB is 1,000 GB.
Typical starting bitrates
These are typical starting values for one camera. They are estimates you can change, not specifications. Your own cameras may differ.
| Camera resolution | H.264 | H.265 |
|---|---|---|
| 2 MP (1080p) | 4 Mbps | 2 Mbps |
| 4 MP (1440p) | 6 Mbps | 3 Mbps |
| 8 MP (4K) | 12 Mbps | 6 Mbps |
In these starting values, H.265 needs about half the bitrate of H.264.
The storage formula
Terabytes = cameras × bitrate in Mbps × 3,600 × hours per day × days ÷ 8 ÷ 1,000,000
What each part does:
- × 3,600 turns seconds into hours.
- ÷ 8 turns megabits into megabytes.
- ÷ 1,000,000 turns megabytes into terabytes.
Worked example 1
32 cameras at 3 Mbps (4 MP with H.265), recorded 24 hours a day, kept for 30 days:
32 × 3 × 3,600 × 24 × 30 ÷ 8 ÷ 1,000,000 = about 31 TB.
That figure is before RAID and free-space headroom. See the safety margins below.
Worked example 2
16 cameras at 2 Mbps (2 MP with H.265), recorded 24 hours a day, kept for 30 days:
16 × 2 × 3,600 × 24 × 30 ÷ 8 ÷ 1,000,000 = about 10.4 TB.
A quick rule of thumb
1 Mbps recorded all day uses about 10.8 GB per day. That is 1 × 3,600 × 24 ÷ 8 = 10,800 MB.
To use the rule, add up the bitrate of all your cameras. Multiply by 10.8 GB, then by the days you keep. For example 1 again: 32 cameras × 3 Mbps = 96 Mbps. 96 × 10.8 GB is about 1 TB per day. Over 30 days, that is about 31 TB. The two methods agree.
Add the safety margins
- RAID: several disks working together, so one failed disk does not lose the recordings. RAID uses part of the raw disk space.
- Free-space headroom: spare space, because disks should never run completely full.
Agree the size of both margins with your IT team and Skyline. Add them on top of the formula result.
Bandwidth for remote sites
SSMS records centrally. So video from a branch travels over your network to the SSMS server. The branch's line must carry every stream that is recorded centrally.
Branch upload needed = the bitrates of its centrally recorded cameras added together, plus headroom.
Example: a branch has 8 cameras at 3 Mbps. It sends about 24 Mbps, all day and all night. Its line must upload at least 24 Mbps without a break, plus headroom for normal office traffic.
- Check the upload speed, not just the download speed. On many business lines, upload is much lower.
- Live viewing can add load too, depending on how your deployment is set up. Ask Skyline about this during the survey.
Use the estimator
The SSMS page has a storage estimator. It turns cameras, bitrate, hours and days into terabytes for you. Try it at the SSMS storage estimator. For the bigger picture on growing a system, read how many cameras one VMS can run.
Where to go next
- Storage and retention: how long video is kept
- Watching branches and remote sites from one console
- Choosing new cameras for an SSMS site
- Connecting sites and remote support safely
Quick checklist
- I counted the recorded cameras at each site.
- I picked a starting bitrate for each camera type.
- I worked out terabytes with the formula or the estimator.
- I added RAID and free-space headroom on top.
- I checked each branch's upload speed against its total bitrate.
- Skyline will confirm the design during the site survey.
Comments
0 total · 0 threads