2009 Megapixel Camera Comparison ReleasedBy: John Honovich, Published on Jan 31, 2009
This 20 page report helps you decide amongst the leading megapixel camera products in the market. If you are evaluating megapixel cameras or moving to IP, this report will help you understand the best product for your needs.
[Update: The 2010 megapixel comparison report is now available. The 2009 report is still useful but the 2010 report has more in-depth analysis.
The report covers the following companies:
- Arecont Vision
The goal of this guide is to help industry professionals understand key factors in selecting megapixel cameras. Megapixel cameras are the fastest growing product segment in all of video surveillance and viewed as one of the key factors in the shift from analog to IP video. To appreciate the competitive positioning of various products, it is important to recognize the following 3 trends:
- The Shift from Standard Definition to Megapixel
- Key Differentiators
- Contrast between HD and Megapixel
Readers of this guide are expected to have moderate knowledge of video surveillance and at least have seen demonstrations or sample images from megapixel cameras. If you have not or want a refresher, see the following options:
- Images from an actual theft solved by a 2MP camera.
- Recorded 16MP Camera in an airport that shows recording of small detail very far away.
- Recorded 5MP Camera in a food court that shows monitoring throughout a large area.
Shift from Standard Definition to Megapixel
Two main drivers exist to move to Megapixel cameras:
- Reduce the number of cameras needed to obtain the image detail needed
- Increase the resolution for a given camera
This is basically the same thing. However, it emphasizes two different aspect of why megapixel cameras are deployed. Appreciating this distinction will help you better understand why megapixel cameras are selected.
- Camera Reduction = Cost Savings
- Increasing Resolution = Loss Prevention
For most applications, I recommend emphasizing camera reduction / cost savings. In security sales, where fixed or limited budgets are the norm, cutting costs are an easy way to approve projects. This is even more important in the current economy where budgetary pressures will grow. By taking this approach, you are not risking or speculating on the ability of the cameras to do something radically novel or questionable.
However, in some applications, an emphasis on loss prevention is key and appropriate. This is most likely where losses are common and high. For instance if you are evaluating deployment of a megapixel camera at a money changing station where you have unresolved disputes once per month, calculating the ROI/payback period becomes very simple. With a high number of incidences and even a modest amount of additional cases solved, the business case can be made. However, any given surveillance camera may only capture an incident once a year or every few years. For those cameras, it is hard to speculate how much money may be saved.
In general, megapixel camera adoption is being driven by camera reduction in areas/scenarios where cameras are already being deployed. The most typical types of deployments are large areas where traditionally multiple cameras have been deployed.
Three factors impact cost savings:
- Cost Difference between the cameras themselves: This is a factor of the relative cost of the cameras and the number of standard cameras a megapixel camera can replace. While megapixel marketing material suggests that a single megapixel is equivalent to dozens of standard definition cameras, this is misleading. In most deployments, the average number of cameras replaced is between 2 and 6 (depending on the coverage area and the amount of megapixels the camera provides). Since megapixel cameras are between 50% and 200% more expensive, this frequently results in a cost savings by itself.
- Cost Differences in Infrastructure for cameras: Since megapixel cameras reduce the number of cameras, the need for electrical power, network cabling, mounting and even potentially trenching may decrease. This is generally a significant advantage for megapixel cameras, ranging from a few hundred to thousands (if trenching is required).
- Cost Difference in storage/bandwidth costs: Since megapixel cameras frequently use less efficient CODECs, storage efficiency may decrease. This can result in storage costs increasing by hundreds of dollars relative to standard definition cameras. This is usually a negative for megapixel.
Key Differentiators for Megapixel Cameras
When evaluating megapixel cameras, a number of features/capabilities differentiates manufacturers and models. These are key elements to evaluate:
Number of Megapixels
Users must determine how many megapixels they need while vendors offer a wide variety of options. The most common offerings are 1.0, 1.3, 2.0, 3.0 and 5.0 megapixels. A few vendors offer 8MP or 11MP while Avigilon is the only vendor offering a 16MP camera. As the number of megapixels increase, the maximum number of frames per second tend to decrease. For instance, Avigilon's 16MP camera can only stream 16MP at 3fps. Additionally, over 5MP, the price for cameras jump dramatically. A 5MP camera is often in the range of $1000 - $1500 but an 11MP camera can be $5,000 - $8,000. As such, in most scenarios, it is less expensive to deploy multiple 5MP cameras than a single 11MP camera.
The key metric to decide on the number of megapixels is usually the horizontal pixels of the camera. This is because most surveillance applications care significantly more for width than they do height.
While megapixel manufacturers emphasis the increase in pixels (which can be up to 100x more than a traditional camera), I recommend using the horizontal increase column above as a more appropriate rule of thumb. For instance, a 3MP camera can generally replace about 3 traditional fixed cameras (in real world deployments).
Low Light Performance
For outdoor deployment, low light performance is a critical factor. While this is important for any surveillance camera, it is especially important for megapixel cameras. All cameras depend on absorbing light to create images. Without enough light, an image can be produced. While a number of factors impact producing an image, one important one is the size of the pixel. With a megapixel camera, there may be 10 to 50 times more pixels, however the size of the chip clearly cannot be increased by the same amount. As such, megapixel is inherently subject to greater concerns over low light performance.
This guide does not make recommendations on low light performance. However, it is strongly recommended that applications requiring night-time surveillance with limited lighting, test megapixel camera performance.
Megapixel camera CODECs can be a secret tax on the use of megapixel. In general, megapixel cameras use less efficient CODECs than standard definition cameras. For instance, where MPEG-4 is common for standard definition, MJPEG is common for megapixel. MJPEG is roughly 50% less efficient than MPEG-4. This difference in efficiency can result in storage costs that are hundreds of dollars more per camera. The cost penalty can even be greater when compared to standard definition cameras using H.264.
Nonetheless, this is a complex issue. A number of 1.3MP cameras have supported MJPEG for a while. All of the new 1080p HDTV cameras claim to support H.264 (many are still being beta tested). Arecont Vision supports H.264 in its entire line of megapixel cameras up to 5MP.
One of the concerns, even for megapixels cameras that do support H.264 is the resulting processing load on recording servers or client PCs. This can depend on the implementation of the CODEC and how the recording server of viewing application is designed. As such it is impossible to universally declare or make a recommendation. In my testing of ArecontVision's 2MP H.264 camera with Exacq's VMS software, I did not see any problem. However, that is one camera with one VMS vendor. Broader conclusions will require further specific testing.
3rd Party Systems Supported
Since megapixel cameras are IP cameras they face the same issue of 3rdrd party support as IP cameras in general. With no standards in place today, each megapixel manufacturer needs to work with each video management system provider to establish support. Statistically, megapixel cameras are less likely to be supported than standard definition cameras (simply because of lower historical demand for megapixel and more recent introductions of many megapixel manufacturers.
Additionally, simply because a video management vendor supports a manufacturer, it does not guarantee that it supports the specific model nor does it ensure that it supports all of a model's features. IQinVision demonstrates this point in their Video Management Support Matrix Chart. IQinVision lists 24 software features. Of the 20 VMS systems that provide support, their level of support ranges from 2 features up to 15 features. You will generally find the same type of support variance for most vendors.
A number of megapixel providers support video analytics with IQinVision's premium line providing the longest and widest support. Both Cisco and Pelco have announced future support for TI's DaVinci chip and associated analytics that run on it.
Nonetheless, there is not a significant need nor benefit for running analytics on megapixel camera. Specifically, there is not a major increase in performance or range to be achieved. As such, most users will select less expensive standard definition cameras for video analytics use.
Traditional or standard definition cameras come in a wide variety of housing from box to dome to pinhole to cube, etc. Megapixel cameras do not have the same variety. Most megapixel vendors only support box cameras. Only a few have dome cameras. Only a few are ready-made for outdoor use.
The logistical/operational issues of deploying cameras can become a significant issue in deciding on cameras to use. For instance, one of the most common reasons IQinVision and Mobotix are selected is because of the camera housing/body options they provide.
Panoramic / Wide Angle Lenses
With significantly more pixels available, a number of specialized lenses/cameras are becoming available to better utilize those pixels. These options attempt to solve 2 fundamental issues in traditional video surveillance:
1.Capturing an entire room: It has been traditionally hard to capture the activities across a room. You could use multiple cameras or sacrifice less important areas. Now, there are a number of options for 360 degree imaging. Immervision offers lenses that can be connected to megapixel cameras (tested up to 1.3 MP). IP Video Market Info Premium Subscribers Only p. 7
Mobotix and Grandeye offer cameras with built in fisheye lenses that provide similar capabilities. Two key challenges for these cameras is (a) how far these cameras can see clearly and (b) what VMS systems support them.
2.Capturing a wider field of view: Most surveillance camera capture image of no more than 70 to 90 degrees. With standard definition cameras, this is generally fine because there are only so many pixels to cover a wide area. With significantly more pixels provided in megapixel cameras, increasing the width of the field of view becomes more viable. The main provider of specializes lenses for megapixel cameras is Theia. Their lenses cover up to 135 degrees and MSRP for $450. The lens is compatible with many but not all megapixel cameras. Learn more about their technology and application in this article.
Contrast Between HD and Megapixel
Historically, everyone who sold cameras with 1 million pixels or more called them megapixel cameras. However, in the last 6 months, a number of manufacturers have begun touting their new megapixel cameras as HD.
Indeed, Axis markets its new high-end Q1755 camera specifically as HDTV, not megapixel. This caught me by surprise and confused me. A few colleagues e-mailed me with the same questions.
HD vs. Megapixel
HD is a type of megapixel camera. All HD cameras are megapixel but not all megapixel cameras are HD. While standard definition cameras (e.g., analog cameras and 4CIF IP cameras) have no more than 400,000 pixels, all megapixel cameras (including HD) have 1,000,000 or more pixels.
HD's Key Features Compared to Typical Megapixel
The key features of HD for video surveillance is:
- Maximum HD resolution is 2.1MP, maximum megapixel resolution is 16MP and up to 5MP cameras are common from numerous vendors
- HD video format is 1280 x 720 or 1920 x 1080 (megapixel cameras can offer many more formats)
- HD aspect ratio is 16:9 (compared to 5:4 or 4:3 in other surveillance cameras)
- HD frame rate is 30/25 (where megapixel cameras are often 3 - 15 frames)
- HDTV has quality compliance standards (where megapixel simply specifies the number of pixels)
Which is Better?
This is not a clear cut case - it will depend on applications. Some applications really need more than 2.1MP (the maximum for HDTV today) and do not need high frame rates (those should choose non-HD megapixel cameras). Other applications will care more about high frame rate and maximum image quality (where HD cameras will be best).
However, since Axis has put its marketing efforts behind HDTV, expect HD's importance to increase significantly. Indeed, I see this as a cunning competitive move on Axis' part to generate perceived differentiation over incumbent megapixel providers such asArecont Vision, IQinVision and Mobotix. It also leverages the consumer interest on HDTVs. This aspect may not improve security but it sure does increased perceived quality.
For most surveillance applications, the 16:9 aspect ratio of HD cameras will be more efficient to the 5:4 aspect of traditional megapixel cameras.
In surveillance, you generally need to view far wider than you need to see high. The reason is that people and vehicles are only so tall but they can be anywhere across a wide area.
However, this does not mean you have to use a HDTV camera. For instance, if you use a 3MP camera (2048 x 1536), crop the bottom and top and have the same effective aspect ratio and resolution of a 1080p HDTV camera. Additionally, this camera may be also be less expensive than a 1080p HDTV camera.
Most applications are fine with 12 frames or less, which is quite typical for 3MP cameras. While a lot of attention will be drawn to 'full frame rate', the reality is surveillance applications have traditionally used far fewer frames per second and been successful at doing.
On the other hand, there will be certain applications (like casinos or money counting) where milliseconds count. A 12 frame per second camera provides an image every 83 milliseconds; A 30 frame per second camera every 33 milliseconds. An HDTV camera cuts out 50 ms. Only in circumstances where very fast tiny movements are key will HDTV make a difference.
Download the complete 2009 Megapixel Camera Comparison Report.