New Surveillance Products Review June 2011

JH
John Honovich
•Published Jun 15, 2011 00:00 AM
PUBLIC - This article does not require an IPVM subscription. Feel free to share.

In this report, we provide in depth coverage of new surveillance product releases announced in May and June 2011. Below are the products we are reviewing including a link to each company's announcement:

This is a sequel to our Spring 2011 new products review in which we reviewed dozens of new products.

Usually, by the end of April, all the new significant product releases are done. This year, many significant new products were announced in the winter and new ones keep on coming. Here is a short overview of the significance of the products reviewed:

  • Axis is claiming ground breaking advances in low light performance, an always challenging area that could have a major competitive impact against leaeding D/N cameras, I/R and white light illuminators
  • Axis has released new encoders with massive price reductions compared to their existing offerings, changing the competitive landscape of the encoder market
  • Dallmeier has gained significant attention for announcing the highest resolution camera in the market by far (51MP)
  • With Hawk-i, major VMS supplier Milestone finally has an iPhone / iPad application that is cost competitive with the market
  • Panasonic is the first surveillance vendor claiming to provide commercially available age and gender detection
  • Vivotek is claiming to be the first to deliver the 'next generation' in video codecs (H.264 SVC)

    • In this note, we examine the features, competitive positioning and pricing of the Axis Q1602. This camera's announcement has generated a significant amount of interest because of Axis's new Lightfinder technology. With this, Axis claims that this camera can maintain details and colors even in very dark conditions.

    First, let's look at the key features and then contrast it to Day/Night cameras, integrated IR cameras, EMCCD and IR illuminators.

    The most notable specifications include:

    • Standard Definition / D1 camera - not megapixel
    • Progressive scan CMOS sensor - not CCD nor EMCCD
    • Day / Night camera supports black and white 
    • Minimum illumination specifications are not finalized; However, Axis projects dramatic advantages in color mode and significant advances in b&w mode relative to traditional surveillance cameras
    • Axis is basing its low light projections based on 1/30s exposure - not dependent on digital slow shutter
    • Similar general feature sets as a P-line camera: audio, on board storage, analytics, vari-focal lens, etc.
    • Q1602 MSRP is $999, estimated shipping date: August 2011
    • Most similar Axis camera (without Lightfinder): the P1343, an SVGA Day/Night camera with a $749 MSRP

    For Axis's pitch / overview, review their Lightfinder whitepaper and the embedded video below:

    As with any newly announced camera claiming low light performance benefits, it is impossible to appropriately guess how much better the performance will be until we test. Below are samples from Axis's own testing:

    Potential Positioning

Industry insights delivered to your inbox weekly.

Sign up to get notified of new reports, investigations, research and more.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

The value of the Q1602 depends primarily on a few factors:

  • Need for low light color imaging: Almost all other low light optimized products focus on providing high quality images in black and white mode (D/N cameras, integrated IR cameras, added IR illumination, thermal, etc.). The only two exceptions are white light LEDs and EMCCD cameras. EMCCD cameras are extremely expensive ($5,000 - $10,000 minimum). While they claim to provide outstanding color images in very dark conditions, given the vast price differences and likely superior performance of EMCCD, we see the products being in different categories. However, white light LEDs are an interesting option. For a few hundred dollars, spot lighting can be added. We suspect this would likely provide a better image in short range scene but would cost more overall (adding in installation and energy costs) and may create undesirable light pollution in some scenes. That noted, we see the Q1602 and cameras with added white light LED illumination as the two closest options.
  • Value of low light color: A lot of users are willing to forgo color at night. It is not clear to us what night time applications are a 'must have' for color. Potentially, this could help identify an unknown suspect by capturing the color of their hair, clothing, car, etc. To the extent that users do not require color, many other more existing options exist (D/N cameras, cameras with integrated IR illuminators, etc.).
  • Digital Slow Shutter as Substitute: Many cameras can provide high quality color images in very low light scenes by slowing the shutter. The downside is that this can create motion blur. However, in some scenes, it is not a practical risk. For instance, the Bosch Dinion defaults to 1/7s max exposure and is likely to provide similar detail/quality as the Q1602 in scenes with slow motion (like the above sample closet / storage room scene). Indeed, even the Axis P1343 with its default slow shutter might provide a suitable alternative.
  • The Q1602 Absolute Low Light Performance: To the extent that the Q1602 can provide materially better low light performance in its black and white in more extreme low light scenarios than competitors, the camera could find broader use. This would be especially true if the Q1602 could do this without the motion blur problematic in using digital slow shutters.

Axis Low Cost Encoders Examined

In June 2011, Axis announced a major expansion of their network encoder offerings. The release consists of a total of 5 models - (2) 4 channel units, (2) 16 channel units and (1) blade encoder. Compared to their current generation H.264 offerings, the new release provides significantly lower pricing that is quite competitive even with Asian competitors. Inside the note, we examine the features, pricing and positioning of the offerings.

Existing Axis Offerings

Before this release, the primary options for Axis encoders were in the Q series - Axis's most feature rich and expensive series. For instance, the single channel Q7401 appliance, the 4 channel Q7404 appliance and the 6 channel Q7406 blade. Online pricing for the single channel unit was ~$500 and for the 4 channel, ~$1000.

Differences in Series Features and Pricing

The new releases are M and P series encoders. Like their camera counterparts, M and P series are less expensive and have less features than Q series products (with M being the entry level and P being the premium or 'mid level' offering).

Specifically to encoders, the follow major characteristics differentiate the 3 series of encoders:

  • Q series: supports 30fps, multiple H.264 streams, maximum video analytics processing power, I/O and audio
  • P series: supports 30fps, a single H.264 stream, I/O and audio
  • M series: supports 15fps max, a single H.264 stream but not I/O nor audio, only 1 year warranty, other series have 3 years)

All series support PTZ control, ONVIF and AVHS (hosted video).

Here are the key differences in pricing for the 4 channel models:

  • Q7404 MSRP - $1099 (existing)
  • P7214 MSRP - $699 (new)
  • M7014 MSRP - $499 (new)

Clearly, the newer models represent a steep decrease in pricing. While the feature sets have certainly been constrained, we believe that many users will find the P and M series feature sets to be good enough. Audio, I/O, analytics, etc. are niche applications. PTZ support is very commonly needed and is supported even with the least expensive unit.

Other New Axis Encoders

In addition to the 4 channel units, Axis released 3 other new encoders:

  • (2) 16 channel encoder appliances. Both are 1RU high, one is an M series (M7010, MSRP $1,999) and the other is a P series (P7210, MSRP $2,499). These appliances have their encoders built in.
  • A 4 channel blade (P7224, MSRP $699) that can be inserted into the existing Q7900 rack. This is a lower cost option to the 6 channel Q7406 blade (MSRP $1499). With the new 4 channel P series blade, a total of 56 channels can be used with the Q7900 whereas the 6 channel Q series blade supports a maximum of 84 channels per rack.

Competitive Comparison

Given the significantly lower prices of Axis's new encoders, Axis will be quite price competitive even with the low cost encoders on the market. Vivotek is a good example of this with 4 channel and 8 channel H.264 encoder offerings. Vivotek's 4 channel unit's online price is about $630 and the 8 channel's is about $1000. Both Axis's M and P series offerings have quite close pricing with Vivotek's (on a feature set basis, the Vivotek units are a closer match to the P series, with 30fps, audio and I/O).

In general, we think Axis new encoders will put price pressure on traditionally lower cost Taiwanese and Korean encoder suppliers. Given Axis's software maturity and broad support, if Axis price is similar, we think most end users would face less risk and better overall value with Axis. It will be interesting to see how and if competitors respond with lower prices.

In terms of shifting users from analog to IP, we think this will have a modest shift on migration. On the one hand, Axis's new encoders do not break ground in terms of cost reduction. On the other hand, Axis's marketing power means that 'mainstream' users who prefer to go with a brand will be more likely to use Axis's lower cost. Prior to this, the cost of Q series Axis encoders was generally more than an entire mid level Hybrid DVR including embedded VMS software and storage. Now, the total solution pricing is much more comparable.

One interesting future element is combining these low cost encoders with iomega NAS devices for AVHS deployments. A 1RU 16 channel encoder for $1999, a few TB NAS for under $500 and AVHS service could be an attractive package for retail chains. We suspect Axis is looking at this given their announcement of these new encoders on the first day of the NRF LP show.

DVTel's Quasar HD Line Examined

In this note, we examine DVTel’s announcement of their Quasar line of HD cameras [link no longer available]. This release includes multiple models of fixed and minidome cameras: The CF fixed camera models, and the CM minidomes. Each of these is available in either 1.3 megapixel/720p or 1080p versions, with true day/night and wide dynamic range functionality in all models.

Compared to Existing Line

Compared to the current generation of DVTel's HD cameras, the HD Classic [link no longer available] and Elite [link no longer available] models, the Quasar lineup features multiple improvements, such as improved nighttime sensitivity and digital noise reduction, and increased frame rates while dual-streaming.  Continuing in the tradition of the previous generation of cameras, Quasar CM minidomes will be part of their Snap-In Camera System. This allows dome housings to be reused simply be replacing the camera module with a new electronics package. This reduces overall cost by eliminating some hardware cost, and reducing installation labor. The CM series domes are also available with motorized lenses which enables remote zoom functionality.  This functionality is found in some competitor's cameas, such as Pelco's Spectra [link no longer available].

New Features

DVTel also claims the Quasar line improves performance thanks to two proprietary features: Scene Adaptive and Smart Picture Quality algorithms. These algorithms adjust picture settings such as gain control, sharpness, and contrast as well as auto-iris and exposure controls based on user thresholds to maximize picture quality and reduce bandwidth needs. DVTel claims these features provide a competitive advantage. Since they are not yet released and we have not tested them, we cannot provide an opinion.

All Quasar models also feature an on-board SD slot for local recording. Video clips stored on the SD card may be emailed or uploaded via FTP upon event, or accessed via the camera’s web browser interface. Once downloaded, users of DVTel’s Latitude VMS may also access these video clips for use in CaseBuilder, their investigative and incident reporting tool.  While not as full-featured as edge storage functions found in other VMS's (see Genetec, March Networks, Axis, or Mobotix), this support for edge storage is an improvement over many manufacturers, which require users to at best, access the camera via web browser, and at worst, remove the flash card from the camera, in order to view stored video.  

Introductory Quasar models will support current ONVIF specifications (version 1.02.2). When ONVIF 2.0 is adopted, all Quasar models will be updated to the 2.0 specification through firmware. Moving from an end-to-end software and camera supplier to a more open supplier of devices which may be used on third-party ONVIF-compliant systems should serve to strengthen their position in the industry.  However, being a DVTel-only dealer line limits how many integrators will choose to use these cameras with third-party NVR's.  We imagine most DVTel integrators with access to these cameras would choose to use them with DVTel's own software.

Pricing and Availability

The Quasar line will be available in Q3 2011. The CM-3211 indoor 720p minidome will have a $719 MSRP. Quasar CF fixed cameras range from $849 to $995 (720p and 1080p, respectively), not including a lens. 

Pricing Comparison

The pricing will likely be fairly competitive to cameras with similar feature sets (megapixel, WDR, D/N, SD card support, etc.). DVTel uses a traditional security dealer discount structure so street pricing should be notably lower than MSRP. We suspect the overall pricing, after adding in cost for lens for the fixed cameras and factoring in discount structure will be similar to Axis P series pricing.

Dallmeier Panomera Camera Systems Examined

Dallmeier's announcement of their Panomera Megapixel camera series [link no longer available] likely drew the most interest of any new product introduction at the IFSEC 2011 show. The two most novel claims Dallmeier made was that (1) the Panomera provided up to 51MP resolution and (2) that the Panomera camera could deliver the equivalent of up to 215 MP of resolution. These would far outstrip any offering in the marketing. In this note, based on an detailed discussion with Dallmeier, we examine the features, pricing and positioning of the Panomera camera series.

Let's start with the basics:

  • Panomera is a multi-imager camera systems that uses multiple cameras inside of a single housing. This is not simply one, super 'big' camera like the Avigilon 16MP. 
  • Panomera uses rows of MP cameras inside the housing. The lowest resolution version provides 3 cameras in a row. Other versions offer 3 rows of 4 cameras (total 12). The maximum resolution version incorporates 17 3MP cameras providing the maximum 51MP.
  • Dallmeier claims, "If you tried to achieve this quality with a conventional megapixel camera, you would have to use a camera with 215 megapixels"
  • Supports H.264 resolution, 1080p/30fps and 3MP/14fps.
  • The first generation of the series, currently shipping, does not offer a mechanical cut filter. The second generation, scheduled for Q1 2012 will add this.
  • Currently, no 3rd party VMS sytems support the camera. However, Dallmeier says that Panomera uses an open API and that several manufacturers plan to integrate with it in the future.
  • The pricing ranges from 3,500 Euros for the entry level to 30,000 Euros for the high end model. That's roughly $5,000 to $50,000 in USD.
  • Panomera states that their decisive advantage is: "completely novel lens and sensor concept. For conventional HD or megapixel cameras, this means that the indicated resolution, let’s say, 12 megapixels, is evenly distributed on the entire viewing angle. The farther you now 'move to the back'of the scene, meaning, zoom into the picture, the higher the loss of detail, causing the picture to become blurry. By contrast, with innovative geometric construction principles, Panomera uses its megapixels such that even objects that are farther away can be displayed with the same resolution as objects in the foreground of the picture." We believe that Dallmeier is using different focal length lenses for the individual cameras within the 'box' to achieve this and the 'equivalent' 215 MP resolution.

Below is Dallmeier's marketing video. The still demonstrates an example of the camera system itself (notice the multiple rows of embedded cameras):

Additionally, Dallmeier provided us a comparison of installing Panomera vs traditional megapixel [link no longer available] and the video quality achieved by Panomera vs traditional megapixel [link no longer available].

Our Analysis

We believe that this series may find a niche for very large scale areas such as stadium. However, we believe this offering is overhyped and can generally be substituted with simpler, cheaper commonly available alternatives.

Here are the specific weaknesses that constrain use:

  • Without a mechanical cut filter, we would anticipate very poor low light performance. This is a real problem because the outdoor environments where this camera is most likely to be used almost always suffer from low light challenges. When the second generation version is released (planned for Q1 2012), this should be resolved.
  • Without 3rd party integration, it will be hard for most anyone to use this camera. It could take a year or more for other systems to support.
  • The cost is fairly high - even for the entry level version. There are many multi-megapixel camera offerings available for $1,000 or less.
  • Most importantly, we think most users could achieve the same basic results by mounting a few megapixel cameras on a pole (or using an existing multi-imager camera such as the Arecont or Avigilon units). At some level, it appears that this is basically want Dallmeier has done - mounted a variety of megapixel cameras with different focal length lenses into a single housing.

On the positive side, if you really need to cover a very large area or cover it 'densely', we think Panomera may be a more elegant solution than trying to hack one's own multi-camera setup.

Last, and perhaps most strongly in favor of Dallmeier, people love big numbers and higher resolution (e.g., the IFSEC award [link no longer available] and buzz). Many, many people will equate the higher resolution with being the best, which is certainly powerful from a marketing perspective.

We think this is overhyped and most users should look to building their own that has a mechanical cut filter, supports their existing VMS and is a cheaper (perhaps substantially) than Dallmeier's.

Hawk-i Milestone iPhone App

A new iPhone application supporting Milestone's VMSes, named Hawk-i [link no longer available], has been released. While many affordable options exist for surveillance apps generally, until now the only apps supporting Milestone's VMS were extremely expensive. This had placed Milestone, one of the most frequently cited VMS providers, at a competitive disadvantage. In this note, we examine the positioning of Hawk-i and contrast it to LexTechs Labs [link no longer available], the most commonly used Milestone mobile solution.

Hawk-i Overview

Here are key points about the Hawk-i mobile application:

LexTech Labs Contrast

Here are key points about LexTech Labs mobile offerings for VMS systems (such as Milestone):

  • Offers 2 versions: a single user optimized offering called iRa Pro [link no longer available] and the team oriented C3 [link no longer available]
  • Both versions are similarly expensive: iRa Pro is $899 per client license; C3 is $900 per user or $250 per camera (selectable by client)
  • Both offerings support all paid Milestone versions including Corporate
  • The C3 offering includes additional 'team' features - central management, no need to setup users or cameras on individual devices, web browser support, ease of access for outsiders/police, additional platforms such as BlackBerry

Positioning Review

A number of key advantages stand out for both offerings:

  • LexTech: Milestone Corporate support, BlackBerry support, centralized management and sharing access via web browser
  • Hawk-i: Same core live and recorded video access at ~90% less cost

As such, we would expect Hawk-i to be an attractive option for small to medium size Milestone deployments that only need a few mobile clients and want to save thousands vs LexTech. By contrast, larger operations (such as campuses) who need dozens of mobile operators and want to facilitate sharing with third parties, are more likely to be willing to spend the significant premium on LexTech.

For Milestone, while many VMS vendors are now providing mobile apps for free, we believe Hawk-i will significantly reduce the competitive disadvantage their mobile offering had previously.

Panasonic Business Intelligence Kit Examined

In May 2011, Panasonic introduced a Business Intelligence Kit [link no longer available]. The most noteworthy aspect of this kit is functionality that automatically detects faces and estimates the age and gender of people. This is the first mainstream commercial offering of this technology we have seen in the surveillance market. In this note, we examine the functionality, pricing and competitive positioning of the kit.

Let's start with the basics:

  • The kit is offered as an add-on for Panasonic's NVR (the WJ-NV200). This NVR includes facial detection / recognition (see our overview of this function). The kit only works with this NVR and is not available to be integrated / added on to 3rd party systems.
  • The kit adds face and gender identification, statistical analysis of people trends (breakdown of age and gender over time) and increases the watchlist size (from 16 to 64) and concurrent matching size (from 8 to 32) for facial recognition
  • A 3 month free trial for the kit is offered with purchase of the NVR. The MSRP of the Business Intelligence Kit is $1,273 USD. By contrast the MSRP of the NVR with 1TB of storage is $3,273 USD(online pricing is about $2,000). We would expect online pricing for the Business Intelligence Kit to be $800 - $900 USD.
  • The intelligence can only be applied on a single camera, automatically set to H.264, 1.3MP, 5fps (see operating manual [link no longer available] for details).
  • In the technical documentation, Panasonic notes a number of limitations: matching may not work outdoors, for partially concealed faces, with strong direct light or when a subject is moving fast (see page 6 of NVR manual [link no longer available]). Up to 500 faces/people can be analyzed per hour.
  • While the documentation does not cite a maximum FoV width, from their sample images, we suspect it is relatively narrow (12 feet or less wide).

Panasonic supports both a real time output for face/gender analysis and a statistical one. Below are examples:

Here is video from a Japanese trade show demonstrating the kit in action:

Our Analysis

We believe the gender and age estimation has potential with some important limitations.

On the positive side:

  • Gender and age estimation does not need to be extremely accurate nor does it matter if any particular individual is misidentified (unlike facial alerting which is highly sensitive to this problem).
  • For retailers, this could be a useful feature as it provides them a data point that is not easily determined by traditional means.
  • The pricing is fairly 'cheap' and one can try it out for a few months for free.

On the negative side:

  • Only supports Panasonic cameras and recorders.
  • Only supports 1 camera per recorder
  • May have problems with various lighting and environmental scenarios (though less likely to be a practical problem as most use cases will be indoors).

At the very least, we think this will be a valuable marketing differentiator for Panasonic as this functionality is commercially so uncommon yet engenders interest and potential value. This is something we plan to test, minimally to better understand how well face and gender detection work.

Vivotek SVC Codec Support Examined

Vivotek has announced upcoming support for SVC (Scalable Video Codec) for its Supreme line of IP Cameras. SVC is important because it has the potential to be the next 'big' CODEC, after H.264 AVC, adopted by video surveillance solutions. In this note, we examine Vivotek's plan for supports, potential issues and benefits of migrating to SVC.

What is SVC?

SVC is an extension to H.264, providing new features to H.264 compression. The key new features are the ability to dynamically 'scale' or change the resolution, frame rate or quality of the video. By contrast, in 'regular' or 'traditional' H.264, an individual video stream can only be set and viewed at a single resolution, frame rate and quality level (e.g., 1.3MP, 30fps). With SVC, the video management system can change the settings of the stream on the fly (e.g., scale down the 1.3MP, 30fps stream to only send or store 480p, 5fps).

What are Vivotek's Plans?

According to Vivotek, SVC support will be available as a firmware upgrade for their new Supreme line of IP cameras (running on recent TI chipsets) but not their legacy series. SVC H.264 support will be provided up to 1080p /  30fps. Firmware upgrades are planned to start in Summer 2011. Additionally, Vivotek says SVC support will be added to their own VMS software and believes 3rd party VMS providers will add support as well (but did not cite any definitive plans for others).

As a side note, Vivotek claimed that their SVC camera support is the 'first'. However, there is so debate about that as GE Security lists support for SVC in an SD IP camera line from 2010 [link no longer available]. However, given GE Security's integration into UTC, we have not seen any practical impact of those cameras. Regardless, anyone releasing SVC H.264 IP cameras in 2011 is at the leading edge of video surveillance releases.

What is the Upside of SVC?

If you are familiar with JPEG2000 and Avigilon's HDSM, H.264 SVC will likely provide similar benefits. Indeed both JPEG2000 and H.264 SVC are scalable video codecs. The big difference is that JPEG2000 is based on JPEG compression while H.264 SVC is based on the much more efficient compression of H.264. 

There are two main uses of a scalable video codec:

  • Streaming to clients: With SVC, you can dynamically change the size of the stream depending on the remote connection and viewing display of a client. This can improve client responsiveness and more efficiently use bandwidth. For instance, instead of trying to send 9 MP streams to display in a 3 x 3 matrix, the server can dynamically scale down the video to 9 CIF streams (reducing bandwidth consumption and CPU/GPU consumption on the client PC).
  • Storage management: With SVC, it is is fairly easy to periodically reduce the resolution or frame rate of video. This provides a simple yet powerful way to reduce long term storage utilization. You can do this with 'regular' H.264 but this is generally much more limited as pruning is limited to I frames (see a Genetec discussion [link no longer available] on such limits).

The main alternative to achieving these benefits with 'regular' H.264 is to use multiple streams at different quality / frame rate settings. With this approach, the VMS can choose between the streams for the right 'size' given the application or time frame. The downside of this is that this generally needs to be manually set up, is time consuming and is generally not done.

What Limitations for SVC H.264 Exist?

There are a number of limitations and issues that will block SVC adoption:

  • The biggest one is that it needs to be supported by both the camera and VMS. Just like when systems migrated to H.264, it takes investment and motivation for both sides of the system to add support in. Currently, VMS support for SVC is basically non-existent.
  • Most cameras will not allow for firmware upgrades as SVC encoding is considerably intense than regular H.264 encoding. As such, this will require a move to a next generation of cameras (like the move from MPEG-4 to H.264 a few years ago).

Finally, while we expect to see more vendors use TI chips to add in SVC support, we think wide spread VMS support for SVC will be slow unless and until a player like Axis with major marketing and VMS clout makes a significant push around SVC. That said, we do believe SVC is the most likely next big codec but it could take a few years to see wide spread use.