Understanding Video Quality

To some, video quality means high resolution, to others it means a faster frame-rate. By simulating the most common kind of events you’re trying to detect and recording them at different settings, you can quickly determine what’s best for you.

Frame-rate

Higher video frame-rates mean higher bitrates, which translates to the need for higher network bandwidth and more storage space. Three frames per second (FPS) is generally sufficient for everyday surveillance. Seven FPS picks up fine movement and is often used for alarm video. Fifteen is the next best thing to full-motion video, which is 30 FPS. It’s important to note that the consequence of moving from three to seven, or seven to 15, or 15 to 30 FPS is that your bit-rate increases by around 70 percent each time. That’s 70 percent more bandwidth and 70 percent more storage costs.

Resolution

Image size (which corresponds to resolution) is measured in terms of CIF. One of the greatest modern CCTV myths is that increasing the resolution and raising the frame-rate will automatically result in superior image quality. This is only true if the bit-rate available for the video is also raised accordingly. If the bit-rate is not raised, then the compression engine will severely compress the video – and 4CIF video can actually look worse than CIF video. When soliciting quotes for a project, make sure you understand any and all assumptions made by the supplier. Bosch always declares the conservative bit-rates used, and from this you are assured high-quality video, even under stressful conditions like low-light (high Automatic Gain Control) and busy scene activity or PTZ motion.

Compression Algorithms

Uncompressed digital video is unusable for CCTV because it uses over 200 Mbps. Many different compression standards exist, but for CCTV, they fall into three broad categories: JPEG, Wavelet, and MPEG. JPEG and Wavelet are both very efficient at low frame-rates, but only JPEGs are interoperable between different manufacturers. However, they both use up to six times more bandwidth than MPEG-4 at higher frame-rates. Most modern CCTV systems are based on MPEG-4, which is an excellent compromise of video quality and bandwidth demands – and significantly more efficient than its broadcast quality predecessor, MPEG-2.

	Good	Better	Excellent Compromise	Best
CIF Resolution	CIF	2CIF	2/3 D1	4CIF Full D1
Width x Height (pixels)	352 x 240	704 x 240	464 x 480	704 x 480

It is important to note the Frame per Second rate (fps rate) in regards to DVR claims. So for example a DVR that does full D1 on each channel at 30 fps is not the same as a DVR that claims to do Full D1, since it may be Full D1 for all cameras combines at 30fps.