Testing Avigilon's H.264 HD Camera
Avigilon is one of the better known and respected recent entrants in the video surveillance market. While they have received significant praise for their quality and usability, historically they had two important constraints:
- The company's cameras were only supported by Avigilon's own VMS. If you wanted to use their cameras, you had to migrate to their VMS.
- Their cameras supported only JPEG2000. While JPEG2000 offers some important benefits, the bandwidth consumption is much higher than H.264, the de facto standard in today's IP camera market.
In the last year, Avigilon released a line of H.264 camera with support for ONVIF, addressing both issues. Now that you can use Avigilon cameras with 3rd party VMS systems, it does raise the question of how well Avigilon's quality matches up against other IP cameras used more commonly in 'open' VMS deployments.
To that end, we bought an Avigilon 1.0MP-HD-H264-B1 [link no longer available] box camera and conducted a series of shootouts between Avigilon, Axis and Sony cameras in a variety of real world surveillance scenarios. In this report, we share our test findings.
Here are a few of the test scenes used including indoor, outdoor, day, night and WDR:
We analyzed each of the scenes to see how the Avigilon camera performed relative to the other cameras. We exported the simultaneous recordings and then digitally zoomed each camera to the same level to show differences. The sample comparison below from this test shows an example of the analysis (camera names/models revealed inside the Pro section):
We conducted our analysis across 10 scenes. They are:
- Indoor Even Artificial Lighting (Moderately Wide FoV)
- Indoor Even Artificial Lighting (Narrow FoV)
- Indoor Low-Light (Default)
- Indoor Low-Light (Normalized)
- Outdoor Daytime Even Lighting (Wide FoV)
- Outdoor Daytime Even Lighting (Moderately Wide FoV)
- Outdoor Nighttime (Wide FoV)
- Outdoor Nighttime (Moderately Wide FoV)
- WDR Scene (Bright)
- WDR Scene (Dark)
Additionally, we examined bandwidth differences across the cameras and scenes to see any tradeoffs in bandwidth vs quality.
The following are key findings from the Avigilon test:
- Under favorable lighting conditions the Avigilon 1.0MP-HD-H264-B1 provided image qualities on par or slightly better than the field.
- Low Light: A limitation of 1/30s maximum exposure time for the Avigilon 1.0MP results in Avigilon's images being darker than competitors supporting slower maximum exposure. [Update: new firmware allows max exposure to go to 1/4s. However, low light performance is still inferior compared to other MP cameras.]
- Bandwidth consumption was generally on par with the field under favorable lighting, but spiked considerably under low-light/nighttime conditions.
- Auto-focus occasionally had problems focusing especially when focal length is changed and in poor lighting.
- Pricing: The Avigilon 1.0MP-HD-H264-B1 at MSRP $645 is highly price competitive relative to other premium tier 720p cameras, e.g. Axis P1344 (online $745) and Sony CH140 (online $800). Also, its integrated lens is a premium feature that often incurs a significant price premium in competitive offerings.
- Hardware Limitations: The Avigilon 1.0MP-HD-H264-B1 does not offer on-board storage or analytics unlike many competitor's premium offerings. While these are statistically used infrequently, applications that require this should keep this point in mind.
- VMS Integration: Integration with a 3rd party VMS (namely, Exacqvision) via ONVIF provided reliable basic funtionality (zoom lens control, live view, and recording/playback). Integration with Avigilon's Control Center provided access to advanced camera configurations
In light of these findings we recommend the following:
- Given its relatively low price, generally solid competitive quality performance and the convenience of an integrated lens, the Avigilon H.264 box camera is worth seriously considering against the bigger IP camera brands for integration with 3rd party VMS systems. However, Avigilon tightly controls distribution and essentially blocks online sales, much more so than other IP camera manufacturers. This can be a big positive or negative depending on your position in the marketplace.
- For low-light applications, consider adding artificial lighting or a competitor's camera allowing a slower shutter speed.
- Reduce maximum gain in low-light environments, especially when apparent illumination in the image appears adequate.
- When focus is not optimal, use the manual focus to improve image sharpness/clarity
- Configure VBR cap and maximum gain settings to help manage network/storage utilization
- If planning to integrate to 3rd party VMS via ONVIF, do a test before specifying or ask manufacturers for confirmation of real world integration success.
Physical Overview
In this video we provide a physical overview of the Avigilon 1.0MP camera. It features an 18x remote zoom lens. A PoE capable Ethernet port is provided on the rear along with some standard hardware features, such as an A/V jack, I/O terminal block, and auxiliary power ports.
Administration and Configuration
In this video we cover adminstration and configuration via the Avigilon camera's web interface. The setup menu is broken down into several categories (e.g. network, image and display, motion detection, system etc.). One nice feature is the option to revert to default settings either on a sub-menu basis or global basis. We cover key configuration tasks such as exposure/gain settings and compression/bandwidth settings. A key limitation is a maximum exposure length of 1/30s that makes low-light/nighttime deployments problematic.
Indoor Even Lit (Far)
In this scenario, cameras are digitally zoomed into a human subject ~33ft out, corresponding to a HFoV of ~31ft. The scene is illuminated via artificial lighting and is evenly distributed (~200 lux).
Below is the composite snapshot (download full video clip [link no longer available]):
With the exception of the M1114, all cameras provide details for identification. The CH120 provides the highest level of details, clarity and sharpness. Among the field the Avigilon 1.0MP appears second best delivering slightly less detail and overall clarity than the CH120.
Indoor Even Lit (Near)
In this scenario, cameras are digitally zoomed into a human subject ~17ft out, corresponding to a HFoV of ~16ft. The scene is illuminated via artificial lighting and is evenly distributed (~200 lux).
Below is the composite snapshot (download full video clip [link no longer available]):
Relative performance does not appear to change at this narrower FoV. Again the CH120 provides the best overall level of detail and image quality. The Avigilon 1.0MP appears a bit over-saturated, but nonetheless provides good detail and overall utility. Except for the M1114, absolute levels of identification are delivered across the board.
Below is the bandwidth utilization graph:
No surprises as results are within expected ranges. The Avigilon operates well below its 12mbps bandwidth cap, in this relatively simple scene. The Axis cameras exhibit similarly efficient bandwidth levels, as well.
Indoor Low-Light (Default)
In this scenario, cameras are digitally zoomed into a human subject ~33ft out, corresponding to a HFoV of ~31ft. The scene is 'dark' with essentially no lighting (~0.1 lux). All cameras are set to default settings. Note that the M1114 is color only.
Below is the composite snapshot (download full video clip [link no longer available]):
In this severe low-light (~0.1 lux) environment, the Avigilon is essentially 'blacked-out'. Like the Sony CH120, the Avigilon features a default max shutter of 1/30s; but the CH120 manages to deliver images with some practical utility. The relatively better image qualities produced by the two Axis cameras is no surprise given their default 1/6s max shutters.
Below is the bandwidth utilization graph:
The Avigilon bandwidth spikes to ~8.5mbps. Having a max shutter length of only 1/30s, the low-light induces considerable gain/noise. Despite this measure to increase apparent illumination the picture quality provides little to no utility and a particularly high bandwidth. In contrast the P1344 delivers a fair level of details at a low 'cost' in bandwidth.
Indoor Low-Light (Normalized)
In this scenario, cameras are digitally zoomed into a human subject ~33ft out, corresponding to a HFoV of ~31ft. The scene is 'dark' with essentially no lighting (~0.1 lux). Cameras have been normalized to a 1/30s shutter speed. Note that the M1114 is color only.
Below is the composite snapshot (download full video clip [link no longer available]):
With all cameras now at a 1/30s max exposure, Axis cameras have both essentially 'blacked-out', and little utility is delivered as a result. Both the Avigilon and Sony CH120 settings have not changed post-normalization, and no material image quality differences exist between the previous 'default' and this 'normalized' scenario.
Below is the bandwidth utilization graph:
No change in the high bandwidth utilization of the Avigilon camera, as no settings were changed. However, normalizing the Axis cameras to a 1/30s shutter results in a near complete loss of utility, as the Axis cameras are now mostly 'blacked-out'.
Outdoor Daytime (Far)
In this scenario cameras are digitally zoomed into a human subject ~75ft out, corresponding to a HFoV of ~70ft. The scene is illuminated via natural sun light and is evenly distributed (~10,000 to 20,000 lux).
Below is the composite snapshot (download full video clip [link no longer available]):
The Avigilon edges out the field at this particularly challenging FoV, managing to deliver a fair degree of facial details. In comparison the CH120, P1344, and M1114 images are less crisp than that of the Avigilon. Odds of identification are slightly better in the Avigilon versus the CH120, P1344, and M1114.
Outdoor Daytime (Near)
In this scenario cameras are digitally zoomed into a human subject ~33ft out, corresponding to a HFoV of ~31ft. The scene is illuminated via natural sun light and is evenly distributed (~10,000 to 20,000 lux).
Below is the composite snapshot (download full video clip [link no longer available]):
The Avigilon provides the best overall image quality among the field. Indeed, sharpness and clarity are particularly strong in comparison to the CH120 and M1114. The CH120 which edged out the Avigilon in the indoor scene with favorable lighting, tends to get washed-out in this outdoor scene.
Below is the bandwidth utilization graph:
Across the board, bandwidths are at reasonable levels, as expected in this fairly non-complex environment. Indeed, on the whole, utilizations are just slightly higher than the 'Indoor Even-Lit' scenario, suggesting a slight increase in complexity (e.g. more motion, lighting effects, vegetation etc.) in this outdoor scene.
Outdoor Nighttime (Far)
In this scenario cameras are digitally zoomed into a human subject ~75ft out, corresponding to a HFoV of ~70ft. The scene is illuminated via artificial lights (~0.5 to 10 lux).
Below is the composite snapshot (download full video clip [link no longer available]):
At the default max shutters, the Axis cameras provide an advantage in low-light as evidenced here. The Axis M1114 arguably provides the greatest overall utility, due to added richness of color (namely, indications of clothing color and complexion). The Avigilon with a default of 1/30s max shutter is particularly noisy, despite a not too severe low-light environment (~5 to 10 lux).
Outdoor Nighttime (Near)
In this scenario cameras are digitally zoomed into a human subject ~33ft out, corresponding to a HFoV of ~31ft. The scene is illuminated via artificial lights (~0.5 to 10 lux).
Below is the composite snapshot (download full video clip [link no longer available]):
Images from the CH120, P1344, and M1114 all provide additional utility over the Avigilon even at this narrower FoV. Indeed, while most images clearly depict a subject holding a license plate, in the Avigilon it's not clear what the subject is holding.
Below is the bandwidth utilization graph:
A fairly high bandwidth utilization is observed in the Avigilon camera (6.77mbps). Moreover, the image quality is the poorest of the lot, as it contains considerable noise. The other D/Ns (P1344 and CH120) provide not only a more efficient bandwidth profile but also images of distinctly higher utility, as well.
WDR Scene (Bright)
In this scenario cameras are digitally zoomed into a human subject ~40ft out, corresponding to a HFoV of ~37ft. The scene's WDR character is produced by a ~2000 lux bright area and a ~200 lux interior 'dark' area. The subject is positioned in the bright area.
Below is the composite snapshot (download full video clip [link no longer available]):
Overall none of the cameras are able to provide details within the back-lit region of the scene, as it is competely washed-out. However, the Avigilon manages to deliver the best level of subject detail among the field in this very challenging WDR scene. In total however it is questionable how much more valuable the marginal gain is given that identification is still fairly unlikely.
WDR Scene (Dark)
In this scenario cameras are digitally zoomed into a human subject ~40ft out, corresponding to a HFoV of ~37ft. The scene's WDR character is produced by a ~2000 lux bright area and a ~200 lux interior 'dark' area. The subject is positioned in the 'dark' area.
Below is the composite snapshot (download full video clip [link no longer available]):
There is no clear 'winner' that can be declared in this scenario. All cameras provide images with low overall clarity and sharpness. Identification is not highly probable, but indications of subject build, hair-style, gender, and complexion are revealed across the board.
Below is the bandwidth utilization graph:
Interestingly, bandwidths, in general, are even lower than the simple 'Indoor Even-Lit' scenario. Note that the Avigilon delivered relatively solid images at a reasonable cost in bandwidth under this typically difficult WDR scenario.
Motion Analysis
The Avigilon camera has a limitation which precludes setting exposure lengths greater than 1/30s. This constraint does not lend itself well to flexible low-light/nighttime deployments. As a result, however, risk of motion blur is essentially eliminated. Be aware the common side-effect seen in low-light environments of increased gain/noise, and bandwidth spikes.
Bandwidth Analysis
Average Bandwidth by Scene
In scenes with favorable lighting (Indoor Even-Lit, Outdoor Daytime, and WDR Scene) the Avigilon camera exhibited similar bandwidth consumptions as the other two D/N cameras tested (P1344 and CH120). However, in low-light/nighttime scenes the Avigilon spiked to bandwidths considerably higher than the D/Ns, skewing 'by scene' averages upwards.
Overall Average by Camera
At 4.74 Mbps the Avigilon's overall average bandwidth is markedly higher than the other two D/Ns. By percentage it was ~500% higher than P1344 and ~130% higher than CH120. Compared to color-only M1114 it was ~25% higher. The reason was Avigilon's bandwidth spikes in the low-light/nighttime scenarios.
Methodology
Here are the four (4) cameras used in comparisons with the Avigilon 1MP:
- Avigilon 1.0MP-HD-H264-B1 (MSRP $645) -720p D/N; 1/2.7" CMOS; F1.6; 0.4/0.04 Lux (Color/BW)
- Axis P1344 (online $759) - 720p D/N; 1/4" CMOS; F1.2 Computar ; 0.05 Lux (BW)
- Axis M1114 (online $475) - 720p D/N; 1/4" CMOS; F1.2 Fujinon; 0.6 Lux (Color)
- Sony CH120 (online $460) - 720p D/N; 1/3" CMOS; ; 0.3 Lux (BW)
Each of the four (4) cameras were simultaneously recorded to an Exacqvision VMS during five (5) key scenarios. All cameras were set to their defaults during testing, except during the low-light tests. During low-light tests D/N cameras were 'forced' into B/W to ensure consistency of comparison. Furthermore, cameras were normalized to 1/30s shutter speed for one of the tests. All camera lenses were adjusted to a uniform lens angle (~50 degrees) and recorded/analyzed at their - VMS supported - maximum resolutions.
Here are the key default settings and firmware versions for each camera:
- Avigilon 1.0MP-HD-H264-B1 - 1/30s maximum exposure; fw 1.0.4.4(3305)
- Axis M1114 - 1/6s max exposure w/ AGC; fw 5.09
- Axis P1344 - 1/6s max exposure w/ AGW; fw 5.06.1
- Sony CH120 - 1/30s shutter; AGC '42dB'; fw1.12.03
Here are the five (5) key scenarios:
- Indoor Even Artificial Lighting - ~200 lux
- Indoor Low-Light - ~0.1 lux
- Outdoor Even Midday Lighting - ~ 10,000 to 20,000 lux
- Outdoor Nighttime - ~0.5 to ~10 lux
- WDR Scene - ~2000 lux (bright area) and ~200 lux (dark area)