Testing Canon IP CamerasBy Ethan Ace, Published on Jun 30, 2014
Canon now owns one of the top global VMS offerings but how good are their IP cameras?
Will the combination of Canon IP cameras and Milestone VMS create the next Avigilon? Or will Canon's cameras be the anchor that drags down Milestone to mediocrity?
Here are our key findings from this test, both for Canon cameras in general, and specific to each camera:
- Both cameras were discovered and added to Milestone XProtect without issue, with streaming and configuration working as expected.
- Adding cameras via ONVIF resulted in MJPEG streaming only in some VMSes, requiring specific URLs to use H.264.
- Both cameras lack a bitrate cap when using VBR encoding. This may be problematic given their higher than average bitrates in low light scenes especially.
- Low light (~2 lux) performance better than other non-IR low cost 1080p cameras tested with better visibility of subject and chart legibility.
- Performance in dark scenes (<1 lux) better than other 1080p minidomes, with the ability to detect test subject, but easily outperformed by low cost integrated IR models. Canon does not offer any integrated IR cameras.
- Moderately better color fidelity in full light than competitive cameras, with subject skin tone and wall color more accurately represented.
- Ethernet and I/O connectors located flat on bottom of dome can cause difficulties mounting to some materials and more prone to dust and dirt collection compared to cameras with recessed connectors or cable pigtails.
- Higher than average bandwidth consumption in low light, especially compared to low cost integrated IR competitors. Bandwidth consumption in full light was average compared to competitive cameras.
VB-H610D Full Size Dome
- Very low light (<1 lux) performance lagged behind 1080p competitors such as the Samsung SNB-6004 and Bosch NBN-932V, though similar to the Pelco IME319 and Sony SNC-VB630.
- Slightly better color fidelity in full light than other cameras tested, with subject skin tone and wall color more accurately displayed.
- Remote adjustment features allow adjustment of the camera along all axes, instead of simply focus and zoom found in many dome models, potentially offering moderate reductions in installation time.
- Higher bandwidth consumption at all light levels than competitive cameras, especially in very low light scenes where greatly increased noise causes bitrate spikes.
- Smart Shade Control was inferior to true WDR cameras, with subject details more difficult to discern against strong backlight, and the subject and chart very difficult to see.
- The VB-H610D is large even by full size dome standards, due to the pan tilt mechanisms used for remote adjustment. At 7" in diameter, it is larger than most other indoor vandal resistant domes as well as many outdoor domes.
The Canon VB-S800D is priced approximately similar to other 1080p fixed lens minidomes from manufacturers such as Bosch and Axis, but as much as $200 above lower cost competitors such as Dahua and Hikvision. Additionally, many low cost competitors sell models with integrated IR for prices well under the VB-S800D.
- Canon VB-S800D: ~$340 USD online
- Axis M3005-V:~$345 online
- Bosch Flexidome Micro 1080p: ~$350 online
- Dahua IPC-HD2200C: ~$160 online
- Hikvision DS-2CD2532F-I ~$215 online
The VB-H610D is priced comparably to most top tier indoor 1080p cameras, but still higher than lower cost models from Samsung, Hikvision, etc. Note that none of these competitive models include the same remote adjustment features as the VB-H610D, though some include motorized focus/zoom:
- Canon VB-H610D: ~$920 USD online
- Axis P3346: ~$930 online
- Hikvision DS-2CD755F-I: ~$500 online
- Samsung SND-6084: ~$480 online
- Sony SNCVM600: ~$800 online
The VB-S800D minidome is an above average performer compared to competitive minidome models. However, its lack of integrated IR is a disadvantage in dark scenes, with many low cost models now offering this feature. Additionally, higher than average bitrates and no VBR cap option may result in wasted storage space.
The key advantage of the VB-H610D is its remote adjustment feature which could speed installation, since installers no longer need to aim the camera on site, simply install it and move on. It could also reduce service costs if the desired field of view changes after installation. However, its average imaging performance, large size, and high bitrate are distinct disadvantages compared to top models today.
In this video we review the physical construction of the VB-S800D:
And here we review the features of the VB-H610D:
The 610D is radically larger than typical high-end domes:
This video reviews the web interface of the Canon cameras, essentially the same in both:
Remote Adjustment Tool
In this video, we demo Canon's remote adjustment features using their software tool and the VB-H610D.
By default, both cameras stream only in MJPEG using ONVIF. In order to connect and stream H.264, users must use a custom URL, adding "#profile0" after the IP address (172.20.128.33#profile0, for example). Even using this URL, we had some intermittent issues with the camera switching to MJPEG after a reboot, but disabling and reenabling it remedied these problems.
Both cameras are supported by direct drivers in Milestone XProtect. We had no issues connecting or streaming in XProtect Enterprise using Driver Pack 7.3, with all features (audio, I/O, etc.) supported.
VB-S800D Minidome Performance
We shot out the Canon VB-S800D minidome against Axis, Bosch, Dahua, Hikvision and IQEye in our conference room at full light, low light, and in the dark. In these comparisons we have included integrated IR cameras from Dahua and Hikvision, priced below the VB-S800D, in order to show the performance difference between it and low cost IR models available.
Starting with full light (~160 lux), the Canon VB-S800D true color representation of our scene and test subject, with performance otherwise similar to other models.
At ~2-3 lux, prior to the Dahua and Hikvision cameras switching into black and white mode with IR, the VB-S800D has the best low light performance of cameras tested. We can see our test subject's face easily, while it is obscured in other cameras.
Below 1 lux, however, the VB-S800D provides only detection of our subject and chart, still better than competitive minidomes, which provide no usable image, the integrated IR models provide details of the subject's face and the test chart.
VB-H610D Full Size Dome Performance
We shot the Canon VB-H610D against Axis, Bosch, Pelco, Samsung and Sony in our conference room scene at multiple light levels, and also in a warehouse for a WDR scene.
Beginning in full light, the VB-H610D provides true color representation, with the wall color and subject's skin tone and clothing closer to actual than in other cameras.
Lowering the lights to 2 lux, the VB-H610D performs similarly to other top 1080p models such as the NBN-932V and SNB-6004, with moderately less noise and artifacting than the Sony and Pelco cameras.
Below 1 lux, the VB-H610 is easily outperformed by the Samsung SNB-6004, as well as the Bosch NBN-932V, with usable image details roughly similar to the Sony SNC-VB630.
Canon's Smart Shade control performed worse in both light and dark WDR scenes than the true WDR cameras in this test. Against strong backlight, it marginally improved visibility of the subject. However, in the dark area next to the open overhead door, the subject and chart can barely be seen at all.
Both Canon cameras' bandwidth consumption averaged higher than competitive models. The charts below show bandwidth with all cameras normalized to 1080p resolution, 10 FPS, 1/30 shutter speed, using H.264 codec.
Canon VB-S800D Bandwidth
The VB-S800D's bitrates were average in full light, similar to Axis, Bosch, and IQInVision. However, at ~2 lux, its bitrate was the highest of all cameras tested, running more than double some cameras. Note that at this light level the M3005 provided very little usable image, resulting in a very low bitrate.
Canon VB-H610D Bandwidth
The VB-H610D has the highest bandwidth consumption in full light and <1 lux scenes. In the dark bitrates were especially high, more than double the next lowest cameras, and over six times the Samsung SNB-6004.
Cameras were tested using these firmware versions:
- Axis M3005: 22.214.171.124
- Axis Q1604: 5.50.3
- Bosch Flexidome Micro 5000: 5.90.0126
- Bosch NBN-932V: 5.90.0126
- Canon VB-H610D: 1.2.1
- Canon VB-S800D: 1.1.0
- Dahua IPC-HFW3200S: 2.420.0000.0.R
- Hikvision DS-2CD2032-I: V5.1.2 build 140116
- IQEye IQD52W: V4.0/092
- Pelco IME319: 126.96.36.199.8280-A0.0
- Samsung SNB-6004: 2.22_131218
- Sony SNC-VB630: 2.1.0