With more and more security systems migrating to IP, contention and bandwidth are hot topics amongst IP camera manufacturers. The benefits of using computer networks for CCTV video are clear for everyone to see, but these benefits may be compromised if data speeds suffer due to bandwidth limitations. As such, any new technology that eases the burden on networks is worth investigating.
Enter Panasonic. Their latest technology, Variable Image Quality on Specified area (VIQS in short), claims to cut bandwidth use by streaming different sections of the camera view at different image quality. The benefit is that bandwidth for non-essential areas is reduced, leading to overall smaller bandwidth use.
This makes sense in theory. But we wanted to see for ourselves. We put the system to the test to see if VIQS in IP cameras really makes a difference to bandwidth use and investigated any side effects encountered.
First off, it is worth noting that this system is not compatible with MJPEG compression, relying instead on H.264 only, which itself drops bandwidth by 50%-80% depending on the camera and scene in view. In combination with H.264, Panasonic’s VIQS system should drop the bandwidth even further.
For our test we are using a Panasonic WV-SP509 IP camera with 3.1 megapixel resolution and, of course, the VIQS feature. We set it up in a typical office environment with the camera covering an office entrance whilst also viewing part of the general office space (see image). For this test, our interest was to capture people coming through the door (as marked on the image) and record detailed footage of them whilst the general office area in view was regarded as less important. We configured the camera so that the door entry would provide us with the highest image quality and detail possible whilst the remainder of the office within view was set to medium quality.
With these settings, we first tested the system with the VIQS switched off, both with a static scene and then with a person walking through the door to generate movement, increasing the complexity of the image which in turn should increase the bandwidth use.
Test with VIQS switched off:
Video with no movement: 220Kbps
Video with movement: 330Kbps
Next we repeated the test with VIQS enabled:
Video with no movement: 140Kbps
Video with movement: 300Kbps
The results of this simple test speak for themselves. With little to no movement in a camera’s field of view, the stream’s bandwidth is cut by around 35%. With movement in view, the drop is not quite so dramatic, but still a substantial 10%.
At first glance it is easy to disregard a small saving of 10%, but when installing a large IP camera system, 10% can be a very substantial saving.
We did not notice any adverse side effects of using VIQS.
We consider VIQS to be potentially useful in installations where specific areas in a scene are of high interest whilst the majority of the scene does not need to be captured with the same level of detail. For example, VIQS in IP cameras may work well at a car park entry or exit where number plate capture is crucial but the general scene of the entrance / exit area does not need to be recorded in such high detail. For these types of scenarios VIQS can offer an additional tool that helps to stay in control of bandwidth.