The confusion of reading specifications on the food labels at the grocery store and understanding what they’re going to do to your waistline is similar to understanding what an IP camera will do to computer network speed. And, an improperly spec’d IP camera can slow your computer network as much as too many bags of “natural” potato chips can ruin your finish time in a marathon.
The first step in determining proper IP camera network speed is finding out where the customer will be viewing his or her IP cameras. Will the system be viewed on the internal network or remotely? If you have a corporate or homeowner client, then the main method of viewing will probably be internal, between 100 and 1,000 Mbps. When over 100 Mbps, I recommend Cat- 6 wiring and connectors to allow an increased amount of traffic in the system. Call on your client’s office IT department to understand the number of users and amount of Mbps data flowing through his company network switches.
If you’re dealing with a retail store, the owner will probably be viewing the cameras offsite and, thus, the bandwidth will be asymmetrical; the download Mbps speed is greater than the upload speed. It is important to confirm Mbps speed information by running an online test (www. speedtest.net). The difference in camera refresh speeds must be discussed up front, so the end user understands the difference between viewing the internal network at 100 Mbps versus over the internet with an upload speed at 2-10 Mbps.
Because of our need for speed and quality, below is a quick comparison to show the difference in network usage from a low-quality image compared to a high-quality image:
|HD 1080P/2MP||30 fps||H.264||Expect 4 Mbps per camera|
|CIF/0.07MP||5 fps||H.264||Expect 125 Kbps per camera|
|HD 1080P/2MP||30 fps||MJPEG||Expect 12 Mbps per camera|
This shows you the 100x variance in the camera!
Let’s assume that you work with the recommended levels to use the megapixel camera at a good performance level of HD 1080p/2MP at 30fps and H.264, expect 4 Mbps per camera. The key point to take into consideration is the 4 Mbps– if you are controlling it–so you don’t upset the IT staff. Consider adding 16 cameras to the system: 16 cameras x 4 Mbps = 64 Mbps, which leaves you 36 Mbps for computers, video downloads, etc. This is definitely going to affect system speed. A quick fix is to recommend separating the network switches into a parallel network.
Saving bandwidth is one of the keys to a successful system design. Compression loss can be one effective option, depending on the preference of viewers or the details of the scene being captured. Increasing compression also can result in great cost saving on hard drive, cloud, switch, and server reductions. This approach can help reduce bandwidth/storage without significantly impacting visible image quality.
Frames per second (fps) rate selection has a major impact on bandwidth consumption for surveillance cameras. While most cameras have a maximum frame rate of 30 frames per second, users typically choose significantly lower frame rates, driven by a desire to minimize storage consumption. The most direct way to reduce storage is to use a lower bit rate stream for each camera with lower frame rates.
As the scene becomes more complex (night time, noisy scene), bandwidth increases linearly as frame rates increase. Try running in variable bit rate (VBR) mode first to get a baseline bandwidth level, and use that as your CBR setting.Written By Darrel Hauk, CEO
This article was reposted with permission from Residential Systems magazine. To read more, visit www.residentialsystems.com