How Nighttime Video May Crash Your Network

Author: Ethan Ace, Published on May 02, 2012

Nighttime video may crash your network. During the day, under ideal conditions, bandwidth remains relatively stable. However, when the lights go down, bandwidth changes can become severe, creating problems like lost video, degraded quality and network instability. In this note, we look at the causes and negative effects of these low-light issues, and give our recommendations for overcoming them.

Bandwidth Planning

Most surveillance networks are designed using average numbers (typically from daytime video) or manufacturer camera calculators. Both of these methods are, if not best case, optimistic, and do not account for nighttime spikes. In other video industries, such as broadcast and A/V, constant bitrate encoding is generally used, to provide more stable, predictable bandwidth consumption. In surveillance, this is not the case, as our poll shows most prefer VBR to CBR, due to the storage savings it provides. However, it creates the potential for spikes and variations which CBR does not.

The Causes of Nighttime Bandwidth Issues

VBR encoding, along with noise created when automatic gain controls turn on, are the typical causes of this issue. Under normal circumstances, video is relatively clear and noise-free. As light levels lower, however, gain controls activate to digitally increase the brightness of the image. This digital correction introduces noise, however, which results in increased video bandwidth.

The following video illustrates surging bandwidth when using an Axis Q1604:

The increase from ~100 kBps (800 kbps) to over 2800 (over 22 Mbps) illustrates just how potentially dangerous surges are.

Network Issues

A nearly 28x increase in bandwidth, as shown above, may easily overload the network. For example, assuming five cameras and a server are connected to a 10/100 switch, bandwidth surges at this level would quickly overwhelm the server's connection to the network, as it totals over 100 Mbps. This could result in a dropped connection, lost video, and other unexpected issues.

The issue is even more critical when using limited-bandwidth connections, such as wireless, 3G/4G, or cable/DSL. Even a moderate surge in bandwidth on a single camera may exceed available bandwidth, dropping the connection.

This is also dangerous when sharing the network with other services. If designers are accounting for 4 Mbps bandwidth on all cameras sharing a LAN with voice and data applications, these spikes may result in interruption to other services, potentially overwhelming switch interconnects and backbone connections.

Storage Issues

Surges may greatly reduce expected storage durations, as well. If storage calculations are based on cameras using average bandwidth, even moderate surges could cut days or even weeks off of retention periods. This is especially dangerous if the organization must meet state or federal regulations, such as the gaming industry, or in critical use systems, such as municipal or government security.

Solutions

There are two ways to account for these nighttime bandwidth spikes: setting bitrate caps on VBR streams, or using constant bitrate streams. We cover both of these in detail in our VBR vs. CBR streaming report:

  • Use VBR with a cap: The preferred way to handle these spikes is to set a maximum bitrate when using VBR streaming. This allows bandwidth to remain low most of the day, but increase to reasonable levels in low-light situations.
  • Use CBR: The second, though less preferred method for dealing with variations in bandwidth is to use constant bit rate streaming. This allows designers to account for variations, as bandwidth should generally not exceed the set rate. However, the CBR bitrate must be set high enough that degradation does not occur during bandwidth surges. This results in higher overall bandwidth and storage usage, where VBR would be lower during other times.

Using gigabit switches may also alleviate minor or moderate network issues, given their higher capacity. However, in larger systems, using 24- or 48-port switches, even gigabit links may not provide enough headroom.

Finally, gain management should be carefully considered. Most often, installers leave cameras set to default gain settings, which typically results in very aggressive levels, and resulting increased noise and bandwidth. Gain should be set to the lowest possible levels which still produce quality images, to avoid this. Users may see our automatic gain control test results for more information on setting gain, as well as comparisons of differing gain levels.

Comments : PRO Members only. Login. or Join.

Related Reports

Dahua 12MP Fisheye Camera Tested (NK8BR4) on May 16, 2018
Continuing our coverage of 12MP sensor fisheye cameras, we bought and tested the Dahua NK8BR4, examining: Default vs. Optimized...
Top Benefits Of Attending Trade Shows (Statistics) on May 15, 2018
150 integrators told IPVM: What are the top benefits of going to trade shows? The clear top 2 responses in order: (1) Networking (2) New...
Axis 12MP Stereographic Camera Tested (M3058-LVE) on May 10, 2018
Axis has released the M3058-PLVE, a 12MP sensor, stereographic panoramic camera and Axis' first with integrated IR claiming images "sharp to the...
March Networks Targets Cannabis Market on May 10, 2018
Will the next March Networks customer appreciation event be held a steakhouse or at a Taco Bell at 2 am? Can March sell the types of systems to the...
Vivotek 12MP Fisheye Camera Tested (FE9391-EV) on May 08, 2018
Next in our 12MP fisheye camera evaluation, we bought and tested Vivotek's latest generation FE9391-EV, a new model claiming improved smart IR...
Last Chance - May 2018 Camera Course on May 03, 2018
This is the last chance to register as the course starts next week. This is the only independent surveillance camera course, based on in-depth...
Avigilon 12MP Fisheye Camera Tested on May 03, 2018
12MP fisheye cameras have become mainstream, with nearly every major manufacturer offering their take on this segment. To see how these new higher...
IP Network Hardware for Surveillance Guide on May 02, 2018
Video surveillance systems depend on IP networking equipment. In this guide, we explain the key pieces of equipment and features, explaining where...
Hikvision DarkfighterX Vs Darkfighter PTZ Tested on Apr 26, 2018
Hikvision has focused on improving low-light performance for PTZs, an area that has traditionally been a problem, even more so than fixed cameras,...
April 2018 IP Networking Course on Apr 19, 2018
This is the last chance to register for our IP Networking course. Register now. NEW - 2 sessions per class, 'day' and 'night' to give you double...

Most Recent Industry Reports

Dahua Intrusion Analytics And VMD Tested on May 21, 2018
Dahua ships basic analytics on practically all their cameras, ranging from low cost to high end. To see how these analytics work in real world...
Exacq Improving Technical Support, Responding To Integrator Complaints on May 21, 2018
Exacq had been a long-term favorite of integrators, but since their 2014 Tyco acquisition, Exacq has fallen in IPVM integrator studies (though...
Best Manufacturer Technical Support 2018 on May 21, 2018
While 5 manufacturers made the worst technical support 2018 list, only 3 stood out as providing the best technical support to 190+ integrators in...
Stealth / UCIT - Remote Video Monitoring Provider Profile on May 18, 2018
Can 2 remote video monitoring companies, Stealth Monitoring from the US and UCIT from Canada combine to impact the market and compete in a changing...
Cybersecurity for IP Video Surveillance Guide on May 18, 2018
Keeping surveillance networks secure can be a daunting task, but there are several methods that can greatly reduce risk, especially when used in...
Forced Entry / Duress Access Tutorial on May 17, 2018
Even though access control normally keeps people safe, tragedies have revealed a significant issue. If users are forced to unlock doors for...
ADT Stock Drops 50% Since IPO on May 17, 2018
It has been a brutal 4 months for ADT. They first expected to IPO at ~$18. They IPOed at $14, dropping immediately to $12.39 And now, not even...
Dahua 12MP Fisheye Camera Tested (NK8BR4) on May 16, 2018
Continuing our coverage of 12MP sensor fisheye cameras, we bought and tested the Dahua NK8BR4, examining: Default vs. Optimized...
Worst Manufacturer Technical Support 2018 on May 16, 2018
5 manufacturers stood out as providing the worst technical support to 190+ integrators in new IPVM results. These integrators answered: In the...
Installing Cameras in Plenums Tutorial on May 15, 2018
There is often confusion about plenum ceilings, with misinformation about what is required when running cables through them and mounting cameras...

The world's leading video surveillance information source, IPVM provides the best reporting, testing and training for 10,000+ members globally. Dedicated to independent and objective information, we uniquely refuse any and all advertisements, sponsorship and consulting from manufacturers.

About | FAQ | Contact