Surveillance Recorder Standards Examined (PSIA / ONVIF)
Security managers suffer from no standards for surveillance recorders. Even large organizations with thousands of cameras and hundreds of sites rarely use more than 1 or 2 manufacturer's recorders (even if they use more than a dozen manufacturer's cameras).
Emerging interoperability specifications from PSIA and ONVIF aim to eliminate this problem.
Operational Constraints from Poor Interoperability
The main practical issue is that each manufacturer has their own proprietary viewing software that almost never supports viewing video from other manufacturer's recorders. Because of this, if security managers used multiple recorders, operators would be forced to use multiple applications.
The limitations are serious. Operators would have to switch back and forth to view different cameras, could get confused about what cameras are associated with which manufacturer's recorders and would not be able to display all cameras on a single monitor.
Financial Penalties from Having to Use a Single Vendor
Because of these drawbacks, the pressures to use a single recording vendor is high.
However, the costs and penalties of 'locking in' to a single vendor is also high:
- If better quality (or less expensive) alternatives arise over time, it's hard to switch.
- If some sites are better served by other manufacturers (specific features needed or lower price points for different camera configurations), it's hard to choose them.
Recorder standards depend on the voluntary adoption by manufacturers. Key benefits include:
- Simplifying 3rd party DVR/NVR integration: PSIM and access control manufacturers will clearly find these standards advantageous. Their solutions require integration with a large number of 3rd party DVR/NVRs. These interoperability standards will reduce the cost and complexity of tasks they currently undertake.
- Reducing Cost of Integrating with PSIM and Access Systems: Video recorder manufacturers who need to integrate with many access and PSIM systems may be attracted to supporting the specification, simply to reduce the cost and complexity of performing integrations with a variety of other vendors. On the other hand, such integrations usually do not win projects - the key driving favor for most manufacturers. To the extent they do, video recorder manufacturers may find it easier to perform integrations on a 'one-off' basis using their existing APIs.
- Supporting rival manufacturer's systems: Upstart recorder manufacturers may adopt these standards in an attempt to displace incumbent manufacturers. This would overcome a historic constraint that encouraged end users to stay with their incumbent recorder manufacturer. On the other hand, incumbents who stand to lose because of increased interoperability may find this to be a disincentive.
- End User Requirements: If end users require the support of these specifications, video surveillance recorder vendors will be motivated to add support. However, this will be unlikely to occur until a sufficient number of recorder vendors add support. As such, this could take some time to occur. Also, depending on what specific functionalities an end user require, they may be reluctant to disqualify manufacturers who do not yet support the specifications.
Obstacles to Recorder Standards
The biggest obstacle to widespread adoption of recorder standards is the disruption to existing competitive advantages for incumbent manufacturers.
The lack of standards means that switching costs to rivals are high. High switching costs increase the power of vendors, generating higher profits and revenues in the existing market.
Vendors may sacrifice this advantage if they believe that standards will help grow the overall market allowing them to compensate smaller profits from their existing customer base for greater revenue from a larger market. However, this is likely not the case for surveillance recorders where the overall market is not growing quickly. This is especially true for DVR manufacturers that have most of the existing market but flat to declining sales.
In a shrinking market under attack from newer competitors, incumbent DVR manufacturers may / should find it more advantageous to delay or refuse recorder standards. On the other hand, PSIA notes that many large incumbent manufacturers are members and contributors to PSIA specifications. What will be interesting to see is if and how quickly these incumbents provide production support for such specifications.
Examining PSIA's RaCM Specification
On September 11, 2009, Version 1.0 of the "Recording and Content Management" specification was released. The 97 page document may be obtained by registering at the PSIA's forum [link no longer available]. Version 1.1 is expected shortly.
Prerequisites
Supporting the recording specification depends on implementing parts of two other PSIA specifications: (1) the IP Media Device (IPMD) Specification [link no longer available] and (2) the soon to be released CMeM specification. The IPMD specification is necessary for supporting live video streams. The CMeM is necessary for supporting metadata based search and video analytics.
Since the recording specification depends on the IP camera specification, this will be simpler for manufacturers implementing PSIA for IP camera interoperability. On the other hand, manufacturers implementing ONVIF may delay and wait for ONVIF's recording specification to be released to minimize redundant effort.
The specification requires the implementation of ZeroConf for recorder discovery. Since most recorders do not support ZeroConf currently, it will need to be added.
The specification supports and identifies a number of video CODECs including H.264, MPEG-4, MJPEG, JPEG-2000, etc (see Appendix A). While new CODEC types can be registered, the current specification does not implement driver downloads for proprietary CODECs. This would require the manual installation of drivers when integrating with DVRs using proprietary CODECs - an issue with a number of incumbent/legacy recorders.
Features of PSIA Specification
The PSIA specification implements a broad number of recording/content management features including:
- Live streaming of video including PTZ control
- Configuration of recording parameters (when to record what channel, time frame, etc.)
- Playback of recorded video (both continuous streams and motion/event triggered video)
- Time Synchronization across recorders (using NTP)
- Searching by time, channel, motion and metadata (such as video analytic event types)
- Insert metadata into recorded video (such as annotating or adding notes to video)
- Check the status of recorded video sources
- Supports a number of playback requirements (section 12.4) including: frame by frame playback (forward and backward), I frame only playback (to reduce bandwidth load), lossless playback mode (to ensure every frame is displayed) and playback at the next point forward for event/motion based recording
Impact of Recorder Interoperability Specifications
If recorder specifications are broadly adopted, we anticipate 4 significant industry changes:
- Growth of PSIM: Providers of command and control software will find it easier to integrate with video recorders. Today, most PSIM systems only integrate with a handful of the leading recorder vendors. Given the fragmented state of the video surveillance recorder market, a large percentage of products are not supported. The growth in PSIM will still be constrained by limitations in integrating with other access control system such as intrusion detection and access control (though PSIA has a working group in area control dedicated to these segments).
- Mix and Match Recorders: End users will start considering how to mix and match recorders for different applications and over longer time frames. For instance, a large corporation with branch offices may use Genetec or Milestone for their corporate facilities but deploy low cost DVRs for the branch offices.
- Accelerate Migration to Up-Start Manufacturers: As end-users look for their next generation of video surveillance recorders, these specifications will make it easier for end users to 'free' them self from the incumbent. Smart up-start VMS manufacturers will add support for this specification so they can manage legacy DVRs, providing a simpler transition to the use of their software/user interface.
- Free/ Low-Cost Video Management Software: As these specifications mature, some manufacturers will find it advantageous to offer this software for free or low cost that integrates with any PSIA and/or ONVIF enabled camera or recorder. This could come in the form of selling complimentary products or 'monetizing' from services as is common with open source software providers.
- Enhanced Premium for 'Powerful'/'Usable' User Interfaces: Many VMS UIs are cumbersome to use, often generating comparisons to spreadsheet software. As it becomes easier to interface with recorders, basic support will lose its value. Manufacturers are likely to differentiate themselves with 'easier to use' UIs and end users will have greater flexibility to choose such offerings.
Projected Timetable for Recording Standards Adoption
Since the PSIA draft specification was just released, it is difficult to estimate with precision the timetable for widespread production adoption of these specifications. Nonetheless, it will certainly depend on 2 key factors:
- How much resources it will take to implement the recording specification? Because recorders are more complex than cameras, implementing the recording specification will likely take more resources than the camera specification. Also, it may take more manual testing and revisions to optimize compatibility.
- How motivated will manufacturers be to adopt the recording specification? As examined above, incumbents may be resistant to adopting such standards. Even for up and coming IP video companies, recording specifications likely represent less of a growth opportunity than IP camera specifications.