How to Design H.265 Systems

May 17, 2018
A closer look at how the new technology will impact legacy systems, VMS, file sizes, display, analytics and more

As with many new technologies, it is often difficult to immediately ascertain its best fit in the security landscape. Such is the case with new H.265 cameras. If you are a systems integrator trying to figure out how to counsel customers on the best approach for taking advantage of this new technology, you are certainly not alone.

These new-to-market models can reduce file size by half without compromising video quality, saving customers money on storage hardware and network infrastructure. At the same time, displaying and processing H.265 video requires far greater horsepower, which can add to system costs elsewhere.

Balancing a range of considerations is essential to providing meaningful guidance to customers. Here’s where to start:

New vs. Legacy

Perhaps the biggest factor that will determine the net benefit of H.265 cameras, for any given project, is whether they will be part of a new system or if they are being incorporated into an existing installation.

Greenfield: For new installations, all hardware should be specified based on H.265 processing requirements. VMS manufacturers who support H.265 by providing separate hardware specs for the new standard are beginning to emerge. Customers purchasing such workstations and servers will initially pay more for them but, over time, this will be countered by savings in bandwidth and storage.

While it is not yet, H.265 will become the default standard within the foreseeable future, and it would be short-sighted for customers to try saving money by purchasing new hardware incapable of supporting it. Either they pay for it once, now, or they will find themselves paying twice when they need to upgrade down the road.

Plus, with appropriate hardware in place, the use of H.265 cameras throughout the installation can only enhance system performance and value.

Legacy: For customers who already have a fully functional VMS network and are looking to add new H.265 cameras, or swap out older cameras for H.265 models, the equation is more muddled. In these cases, a project that is budgeted for just new cameras can open a can of worms when it is determined that the new cameras compromise the performance of existing display stations and other functions, like search and video analytics.

In some cases, it may make sense to stick with H.264 for the time being, especially for systems using older PCs for monitoring purposes.

Almost all H.265 cameras can also support the H.264 standard, so forward-thinking customers may still find it wise to invest in H.265 models as they add to their network, even if the cameras’ superior encoding capabilities won’t be put to use until some later date.

The Impact of File Size

In terms of overall efficiency, H.265 is just like H.264 in that it performs the very best in situations where there is not a lot of motion, such as hallways or parking lots. These scenarios enable the encoding engine to take maximum advantage of similarities between individual video frames, which is the underlying principle for both H.264 and H.265.

Using H.265 as compared to H.264 delivers a net benefit with video featuring a lot of activity moving in predictable patterns, such as the view of a crowded escalator or expressway. In these cases, improvements in compression efficiency can approach 60 percent or more.

Reduced file sizes translate to lower system bandwidth requirements, and the degree to which this can impact total system cost depends on the size of an installation, the strength of its existing infrastructure and system topology. Larger installations with more cameras stand to benefit more, as do facilities that lack a modern, powerful network infrastructure, like a fiber optic backbone to support the type of heavy, high-speed traffic generated by VMS systems.

System design also makes a difference. Especially with enterprise or multi-site installations, decisions about where video is recorded relative to camera location, as well as how viewing stations are supported, affect how much video data is traversing the network at any given time. Customers who have previously had to make compromises on video quality settings in order to keep network traffic in check will find that with H.265, they now have the freedom to either increase the quality of video or to add more cameras to the system.

Advantages related to storage are straight-forward. Smaller files mean recording systems already in place can now either store video from up to twice as many cameras for the same period of time, or video can be stored for twice as long before running out of space; thus, when purchasing new hardware, less money needs to be allocated for storage.

The Display Challenge

Compared to H.264, H.265 can require up to eight times the processing power to encode, and twice the power to decode. Fortunately, the encoding process is handled almost exclusively by cameras, and the cost of new H.265 cameras is not much different than for H.264 cameras.

Decoding, on the other hand, needs to happen anywhere that video is displayed within a VMS system. Since this requires twice as much horsepower as decoding H.264, hardware currently used for viewing H.264 streams will only be able to display about half the number of concurrent video streams in H.265. This is where the challenges for legacy systems come into play. For example, if a viewing station is configured with layouts that display eight 2MP cameras, and that CPU is already performing to its maximum capacity using H.264, it will only be able to display four cameras in H.265.

One way that VMS system designers can minimize the need for upgraded hardware to support H.265 is by taking advantage of camera features that can reduce some of the load on VMS workstations and display devices.

Cameras that offer triple codecs (encoding to M-JPEG, H.264 and H.265) along with multi-streaming (transmission of two or three simultaneous, independent streams) make it possible to transmit H.265, high-resolution video to a recording device, while standard resolution (VGA) video in M-JPEG or H.264 is sent to the display station for inclusion in multi-camera display layouts. As many as 16 VGA M-JPEG streams can be supported with the same processing power as only two H.265 streams. Then, when a specific camera is clicked on and expanded to fill the majority of the screen, the VMS can switch over to display the high-resolution H.265 stream for that one camera.

This display strategy offers very little downside, assuming that the cameras in use support multi-streaming and the VMS client software can be configured as described. The visual superiority of the higher resolution video stream is far more evident when expanded, and it is only within this expanded view that megapixel resolutions really pay off.

With this approach, it is only necessary to support one H.265 video stream at a time at the viewing station, keeping the burden of video decoding within reason. At the same time, full, high-resolution video from all cameras is stored to disk in the highly compressed H.265 format, making that video available for forensic use as needed.

Search and Analytics

It is difficult to talk universally about the impact that H.265 will have on the processing power needed to perform search and analytics, as different VMS engines handle these functions differently. For example, many IP cameras equipped with motion detection and other analytics capabilities have the ability to pass this information along to a VMS server as metadata – descriptive data that gets recorded alongside the video stream. If a VMS is able to search through the metadata to find incidents without first decoding the corresponding video stream, the fact that the video is H.265 is somewhat irrelevant.

Another variable is whether analytics engines are integrated into the VMS platform or are handled by a separate, dedicated server. If you are designing or upgrading a system that uses analytics, make sure to initiate a conversation with your VMS manufacturers’ tech support team to make sure the analytics will continue to work as expected with video from the new cameras.

Your VMS manufacturer’s support team is actually the best place to start in making sure realistic expectations are established for all aspects of your project, and that best practices are employed to deliver the maximum benefits that these cameras can deliver. Customers can rightfully expect dramatically superior compression and the advantages that come with that, but the cameras’ value must be examined within the context of their entire VMS system.

Guy Arazi is Director of Product Management for Vicon ( To learn more about H.265 technology and its impact on VMS system design, download Vicon’s free whitepaper at