Testing Pelco's Sarix Megapixel Cameras (IXE20DN)

JH
John Honovich
•Published Dec 06, 2009 00:00 AM

While Pelco is a leader in analog surveillance cameras, they are not even in the top 10 for 2008 global network camera sales. Pelco aims to rectify this with their Sarix family of IP cameras [link no longer available] supporting both megapixel resolution and H.264 processing.

We tested a Pelco IXE20DN [link no longer available] (firmware version 1.3.7.9081-A1.3199) using Milestone Enterprise's 6.5f software [link no longer available] with Device Pack 4.5. The test was conducted using H.264 resolution at 30fps. The test included filming indoors and outdoors during the daytime, night time and under measured low lux scenarios. We also tested at multiple bit rates to examine the impact of adjusting bandwidth consumption.

Our key findings include:

  • Day time video quality had sharp focus and fairly true to life colors
  • Low light (under 1 lux) video tended to be dark with minimal motion blur and low visible noise (using default settings)
  • Integration of Sarix and Milestone was unstable, suffering from various problems
  • Limitation of supporting 3rd party VMS systems can block use
  • Relatively high H.264 bandwidth consumption with 3rd party systems should be considered
Compare to our other megapixel camera tests including: Axis Q1755, Arecont Vision 3105, Basler's 1.3/2MP cameras, Sanyo VCC-HD4000, Stardot SD500BN, and Vivotek IP7161.

Product Overview

The Sarix IX line of 'box cameras' support the following key features:

  • H.264 encoding, 1080p resolution at 30fps (prior to the current firmware version, the maximum frame rate was 15fps for 1080p)
  • Auto-focus capability triggered by physical button on side of camera and through web interface
  • Support for baseline, main and high profile H.264 encoding (with baseline being the least bit rate efficient and high being the most - see overview [link no longer available]). Use of these profiles depends on VMS support.
  • Defaults: Maximum exposure setting of 1/30s during the day and 1/8.3s during the night.
  • CBR only streaming (no VBR) with default bit rate of 6.5Mb/s for H.264 baseline, 30fps, 1080p
  • Note: The lens is sold separately from the camera. For this test, we used the Pelco 13M  2.8-8mm Auto-iris lens [link no longer available].
  • The IXE20 series offers multiple models supporting both ObjectVideo's analytics and Pelco's own analytics.
For full details, review the product datasheet [link no longer available] and the installation/operation manual [link no longer available].

Pricing

The IXE20DN, the 2 Megapixel Day/night version with no analytics is available on-line for about $1200 USD. The Pelco 13M 2.8-8mm lens is available on-line for approximately $130 USD.

Image Quality

Below is our screencast commenting and showing Sarix's image quality.  More importantly, is a ZIP package of video clips for you to review yourself.

Download the package of Sarix test video clips and image snapshots [link no longer available] (266 MB total).

In general, we though the video quality was solid both day and night. What stood out was that night time performance was achieved using the defaults. By contrast, some cameras were quite poor in low light (like Sanyo) or defaulted to an unrealistic 1 second exposure time (like Basler) which transformed moving objects into 'ghosts'.

Watch the screencast examining video quality:

Physical Setup

From a physical perspective, the camera offers a modest number of features including auto-focus, service jack for analog video/servicing and an SD card slot. Note: the auto-focus on the camera only performs a 'quick focus'. The 'full focus' functionality, important for cameras that are significantly out of focus must be performed from the web interface.

The screencast below overviews the physical elements of the Sarix IX camera:

Software Configuration

The screencast below describes the general software configuration process:

The key software challenge we had was integrating the Sarix camera with Milestone's VMS software [Note: We did not test this with Pelco's NVR/VMS systems. 3rd party VMS support is limited currently with Genetec being the only other option recommended by Pelco].

We spent over 8 hours troubleshooting this with multiple phone calls to both Pelco and Milestone technical support. They were unable to resolve.

We were able to establish a fully functional connection for multiple times during the two week testing period. However, periodically, video stopped displaying. [This happened with the VMS running on multiple machines and with the Sarix camera reset to factory defaults].

One issue we suspect is impacting integration and video display is Sarix's current requirement to use an old version of Quicktime (7.5.5). According to Pelco, Quicktime's most recent release caused problems with video decoding. For instance, when we first started our test, we could not see any H.264 video until we 'rolled back' our Quicktime version. Quicktime, by default, automatically attempts to upgrade to the most recent version. This may have been a factor.

Bandwidth Consumption

Bandwidth consumption is a key consideration for the use of Sarix's cameras. While Sarix supports multiple H.264 CODEC profiles (baseline, main, high), Milestone only supports baseline for Sarix cameras (which is similar to the profile used in other H.264 cameras). Significant bandwidth savings may be possible by using main or high profiles. However, these may only be available through Pelco's own VMS systems.

A key practical matter is selecting the target bit rate for the CBR stream. When configuring using the web browser, Pelco offers a range of acceptable bit rates. For H.264, 30fps, baseline encoding, the lowest rate in the range is 2.488 Mb/s and 6.5 Mb/s was the default.

Pelco does not support a quality scale that allows you to adjust encoding quality. In our tests, we found that setting lower bit rates acted as modest quality scale but nowhere near as extreme as we thought would occur.

We tested at both 2.5 Mb/s and 6.5 Mb/s to examine variance in image quality. At 2.5Mb/s , most of the videos were as clear and fluid as the 6.5 Mb/s videos. There were a few exceptions: (1) the daytime indoor video at 6.5Mb/s showed slightly more details than the 2.5Mb/s. (2) A few of the low light/night time videos, though set at 2.5Mb/s, streamed at closer to 5Mb/s.

I recommend that integrators carefully consider the bit rate to use.  It is likely that the bit rate can be set significantly lower than the defaults, resulting in much lower bandwidth consumption.

Camera Comparison

Contrasting the Sarix to Axis' Q1755 and the Arecont Vision AV2105 can illuminate tradeoffs. While there are many H.264 megapixel cameras, these 2 represent relative highs and lows for on--line pricing with the the Q1755 at about $1400 (lens included) and the AV2105DN at about $650 (without lens). With lens, the AV2105 will be about $750-$800 and the Sarix IXE20DN will be about $1350.

Contrasting Axis Q1755 and Sarix IXE20DN

Prices are similar. However, Axis has much broader 3rd party support and a built-in 10x optical zoom that not only provides auto-focusing but simple adjustment of the FoV. Axis supports VBR with much more bandwidth consumption during the day and similar consumption during the night.

Contrasting AV2105DN and Sarix IXE20DN

Pelco is $500-$600 more expensive than Arecont. Pelco does provide a setup jack and auto-focusing providing easier setup and potentially sharper video over the long term than the Arecont Vision. On the other hand, Arecont's H.264 cameras have much broader 3rd party support than Pelco's. Arecont supports VBR which provides much lower daytime/low motion bandwidth consumption but much higher consumption for low light and night time.

Image Quality Comparisons

While we have tested each camera individually, we have not tested them at the same time in the same scenes. As such, while image quality is key, we cannot offer an educated opinion at this time.