Streaming
Streaming real-time video over a congested pipeline requires compressing the images to a point where a balance is struck between acceptable levels of dropped frames and image quality. There is no sense in streaming video when frames are constantly lost in transit during critical moments, intervals between images are pointlessly long or images are too small to be useful.
One should definitely look at how independent the streams are, Christensson said. “Most products offer a set of multiple streams, but many put restrictions on having to use the same H.264 profile or only support M-JPEG as a secondary stream. A key feature is to be able to change settings on-the-fly for one stream without influencing the others, such as changing frame rate or quality.”
A key consideration for streaming is interoperability, Golston said. “There is no point in introducing new codec functionality until a large part of the infrastructure is able to decode the video.”
My Phone Has a Better Camera!
Customers expect the same level of image quality from their video feeds as they get from HD TVs or full-HD home theaters, and yet current solutions do not provide the type of crystal clear and silky smooth video as consumer electronics do. For security applications, there are certain aspects to capturing, transmitting and storing video that have little to no tolerance for the occasional quirks common in consumer products.
One of the barriers is the silicon being used, Golston said. “The gap closes over time with more advanced silicon being introduced that can take advantage of the capabilities of H.264.”
Broadcasting generally uses more advanced H.264 encoding techniques than surveillance uses today, Christensson said. “Broadcast encoder inputs can be more complex than the typical surveillance scenes, so there are less ‘shortcuts' to take. The cable and satellite broadcasting network also pays a huge premium for bandwidth, so they choose the best encoding technology to get more compression.”
“An example of this is that a broadcast-grade H.264 encoder uses more powerful noise filtering, mode decision and motion estimation search algorithms that can result in three to five times better compression ratio than the typical low-cost H.264 encoders you may find in surveillance DVRs today,” Christensson cautioned.
Some users ask for remote video quality that exceeds what their network bandwidth can afford, Chen said. “The advent of HD multimedia in the consumer space has raised the bar for what customers expect from video products, and some people have a hard time understanding why they sometimes can't get beautiful 720p video at 25 fps.”
While these expectations are not unrealistic, in many cases network cameras and DVRs are not yet 'up to spec' to match consumer HD technology, Christensson said. “There are also many moving parts involved which make it difficult for customers to understand. If one parameter is optimized, trade-offs are required on other functions or features. We see big improvements now and the next couple of years as surveillance technology catches up with broadcast technology. We'll see lower latencies, better compression and more stable and fluid video surveillance takes its time to catch up with the consumer market. delivery.”