One of the industry's top growing myths is that IR illumination massively reduces storage and bandwidth costs. No one has done more to perpetuate that myth than UK manufacturer Raytec. In this report, we break down the fallacy and provide clear recommendations on how you can minimize storage and bandwidth costs without buying any IR illuminators.

What's the Claim?

Raytec claims that adding IR illuminators can reduce storage and bandwidth costs by up to 90% - a key component of their PR and marketing releases. The most substantive Raytec document on this topic is their 'white paper' on lighting for network cameras where their test results found no less than 74% bandwidth and storage reduction when using IR illuminators.

The claim is quite important to their value proposition because if IR illuminators really are the only way to reduce such costs, it could result in hundreds of dollars of savings, greatly offsetting the purchase of an IR illuminator.

Unfortunately, their claims, configuration and testing is riddled with fundamental errors.

Error #1: Confusing VBR and CBR

Raytec erroneously assumes that all cameras use Variable Bit Rate streaming. Many leading cameras use Constant Bit Rate by default (e.g., Sony and Panasonic). Those cameras never have problems with bandwidth in low light. Bandwidth savings will be 0% for those cameras. Even cameras, like Axis, easily can be switched into CBR mode also eliminating any low light bandwidth spikes.

UPDATE 12/2011: Axis's CBR mode acts as bandwidth ceiling allowing for minimal bandwidth consumption during the day or low motion scenes but stopping bandwidth spikes at night without needing to add in an IR illuminator. It is a 30 second configuration change.

Bottom line: you do not need to buy Raytec's products to reduce bandwidth and storage costs. Raytec is either grossly incompetent or purposefully manipulative not to explain CBR or bandwidth caps in their 'educational' materials.

Error #2: Confusing Gain Control and Low Light Impacts

It is a myth that low light increases bandwidth consumption. Increasing gain levels is the real cause. Raytec is confusing correlation with causality. Indeed, we have tested increasing gain levels even when there is plenty of light. Bandwidth still spikes. Why? High levels of gain drive bandwidth consumption not insufficient light.

This has a very important impact on camera selection and bandwidth problems. Since camera manufacturers implement gain (and encoding) differently, some camera manufacturers are far worse in pumping up gain and in handling noise from gain control. Raytec picked Arecont cameras for its tests. Unfortunately, Arecont is one of the worst low light bandwidth performers in the dozens of cameras we have tested (see test results of Arecont's low light bandwidth issues here, here, here and here).

Understand the particular camera you are using and how bad its low light bandwidth consumption is. If you care about reducing bandwidth and storage consumption and want/need to use VBR, consider avoiding bad performers like Arecont.

Error #3: More Bandwidth is Useful In Low Light

There is almost no benefit of increasing bandwidth consumption in low light, high noise environments. We have ran various tests with the same low light conditions but at sharply different bit rates. Image quality is essentially the same.

Using CBR or even VBR with a bandwidth cap will eliminate any spikes in bandwidth without reducing image quality. No Raytec illuminator required.

Error #4: Unrealistic Test Scenes

Most of Raytec's tests are done in very small areas such as a cardboard box. Here's an example of one Raytec test environment:

If your real world surveillance cameras are monitoring the insides of a cardboard box and are using VBR, Raytec illuminators will definitely massively reduce bandwidth consumption. However, this is pointless. It is easy to illuminate small areas but those areas do not reflect real world conditions.

Unfortunately only one of Raytec's test results are outdoors (the most common use of IR). However, again it is in a relatively small backyard scene. 

To get the 74% bandwidth savings they claim, they had to use a VBR camera with very poor low light bandwidth performance (Arecont) and flood a small scene with IR light (indeed if anyone was standing in the left center, the subject would be totally overexposed - you can see how white hot that area is).

Even with VBR, bandwidth reduction is heavily dependent on saturating the entire area with illumination which can be unrealistic outdoors and can make image quality actually worse (e.g., washing out subjects who are overexposed).

Contrast to our test results in outdoor areas using both IR LEDs and White Light LEDs at nights with various cameras. Even with VBR, bandwidth reduction is modest to negligible.

Conclusion

Lighting can be useful for quality surveillance but adding lighting to reduce bandwidth and storage is a waste of money. While Raytec is certainly entitled to promote the sale of its product, it uses a combination of unrealistic test conditions and ignorant camera settings to deliver fundamentally misinformed and erroneous conclusions. Try using the recommendations in this report to eliminate bandwidth and storage costs without having to buy additional equipment.