When most people talk about high-quality video, the conversation likely starts with — and quite possibly ends with — resolution. However, that’s only part of the story. There are many factors that go into a video camera’s ability to deliver the high-quality images necessary for today’s surveillance applications.
That’s not to say resolution isn’t important. In fact, says Tim Shen, director of marketing for Dahua Technology, based in Hangzhou, China, resolution should be high on the list of considerations. “I would recommend that dealers and integrators first consider camera resolution and determine how many megapixels are required for the system,” Shen says.
Before delving into the inner workings of cameras, it must be said that the primary goal for dealers and integrators who are working with high-quality video should be making sure that a camera’s capabilities are aligned with the specific needs of end users and applications.
“There’s lots of whiz-bang stuff out there and you’re going to pay a lot of money for those, but you may not really need to. So be conservative,” says Doug Gray, product manager for Hikvision USA and Hikvision Canada, City of Industry, Calif. “Most cameras that you select will overcompensate for what you need to do and may do a much better job than you’re expecting.”
While all applications may be unique, there are a few common camera elements that affect image quality, each of which dealers and integrators should look at closely when designing and installing video systems.
Perhaps more than any other factor, the lens has the greatest impact on image quality. It’s the most basic factor that’s likely to be taken for granted.
“The lens is an important thing. With high-resolution cameras, that’s become more obvious. It used to be that a lens was a lens pretty much, during the analog days. Now you really need to think about what lens you match up with the camera,” says Fredrik Nilsson, general manager of Chelmsford, Mass.-based Axis Communications. “The lens is interesting because everything else in the camera aligns with Moore’s Law in that it gets better and faster all the time for a similar cost because of developments in memory chipsets and processors; but the lens doesn’t follow Moore’s Law. It’s mechanical so, if anything, gets more expensive because of raw materials and manufacturing labor.”
However, it’s the lens that largely determines image resolution and clarity. The wrong lens can limit camera performance and inhibit its ability to deliver quality video. A main culprit is often a mismatch between the lens and the imaging sensor.
“Resolution, as defined by the image sensor, does not necessarily translate to the optics that may be included with a camera,” says Allen Chan, senior product manager for Sony Electronics’ Security Systems Division, Park Ridge, N.J. “A camera lens needs to employ higher optical performance and lower F-numbers so that it can take advantage of all the information generated by a high-quality image sensor.”
The materials used to make a lens play an important role in determining how effective it will be at delivering high-quality images. While plastic lenses can be much less expensive for a manufacturer, Jeff Whitney, vice president of marketing for Arecont Vision, Glendale, Calif., says for the most part, they should be avoided at all costs.
“The optical quality of the lens can be inferior to a good glass lens in both image quality and length of service life,” he says. “They also can have significantly lower image quality in high temperatures. Plastic lenses typically are found in lower cost surveillance cameras as a cost-cutting measure.”
To ensure the highest quality and best performance, glass lenses are preferred. However, a glass lens doesn’t necessarily guarantee high-quality video.
“Some manufacturers are less strict than others when it comes to the quality of the optical glass, Whitney explains. “This can lead to imperfections in the lens that may negatively impact image quality in undesirable ways and offer inconsistent imagery.”
Therefore, when evaluating the lens, look for one that is made with quality glass and is designed to handle the resolution and other capabilities of the camera.
Being charged with the task of regulating light sensitivity and accurate color reproduction places imaging sensors high on the list of factors in determining image quality. Of particular note is the shift that has taken place recently in the type of sensor that is preferred for video surveillance applications.
“It used to be that people wanted a CCD image sensor because they were better than CMOS. Now with the very rapid development of CMOS sensors because they’re in every cell phone, every car, and every digital camera and so forth, CMOS is actually dominating now,” Nilsson says. “The quality of the sensors is developing all the time with light sensitivity and resolution and other dynamics,
Because there are only a few manufacturers in the world that produce professional megapixel camera-grade sensors, Whitney says camera companies often have the same sensors as their competitors. One big difference that impacts image quality, he adds, is the lens underscoring the importance of matching the lens with the camera’s capabilities and the application.
Like any camera element, the image sensor has to be considered as part of the whole when evaluating options.
“From the camera perspective, most people believe that the image sensor determines image quality. While the sensor is certainly an important component, it’s not the only component that impacts image quality,” Chan says.
Network & Storage
Higher-quality, higher-resolution cameras provide more pixels and more details, which increases both bandwidth and storage requirements. With storage costs making up approximately 50 percent of the total cost of a video system, this could be an expensive proposition for end users, says Cheryl Bard, regional marketing manager for Fairport, N.Y.-based Bosch Security Systems.
“A camera with powerful encoding and intelligent digital noise reduction can reduce bitrate by up to 50 percent to significantly reduce storage costs and network strain,” she says.
Compression and encoding alone, however, are not always the answer. In some cases, they can even hamper image quality.
“If you don’t have enough bandwidth, you can’t get enough video throughput. Then you compress the video harder and if you compress the video harder, you don’t have the same image quality,” Nilsson says. “You can decide to not compress the video at all or have very basic compression but then it becomes unmanageable. Even though everyone is using standard compression today, mostly H.264 and soon H.265, there are variations to those that can compress the video quite a bit without losing the image detail that is very specific to surveillance scenes.”
Another way to combat bandwidth and storage challenges is by segregating video surveillance from other networks.
“When working with high-quality video, a private and isolated network is always recommended for IP cameras due to the high-bandwidth consumption of video,” says Alex Goussev, product manager, Speco Technologies, Amityville, N.Y.
As resolution increases, light sensitivity decreases, so higher-resolution cameras require more light to create quality images. This creates the need for dealers and integrators to modify their sales approach a bit, Nilsson says.
“What integrators typically sell to customers is resolution,” says Fredrik Nilsson of Axis Communications. “We have to get better at talking about low-light capability because people tend to do bad things in the dark.”
Higher-resolution cameras tend to switch from color mode to black-and-white mode earlier in the day than lower-resolution cameras, Chan says, making minimum illumination a key factor for evaluating cameras. However, that’s sometimes easier said than done. For example, one vendor may measure using a shutter speed of 1/30 second, while another uses 1/4 second.
“When the shutter speed is slowed down by 50 percent, it results in a minimum illumination number which is approximately two times better,” Chan says. (For more on the varying specifications between manufacturers and models, see “Spec Sheets vs. Reality” on page 85.)
In situations where lighting is really poor, infrared IR illuminators can be deployed. The challenge when using external IR sources is maintaining overall picture clarity as objects move closer to or farther from the camera. A built-in IR light source allows the camera to better control the illumination level or exposure.
Low-light isn’t the only challenge cameras face. A high contrast between light and dark areas in a scene can also cause less-than-ideal image quality. Wide dynamic range (WDR) can address these challenges, but there are two types: digital and software-based or “true” WDR.
“With digital WDR, you have a range that it will adjust to. With true WDR, you’re taking multiple exposures of a scene and restoring those images into the best possible view based on those exposures,” Gray says. “WDR is important, but you have to recognize that there are two types and that the true WDR is considered better.”
As evidenced by the continued adoption of 4K video in the industry, the demand for high-quality video continues to grow. However, all cameras are not created equal, and sometimes it can be difficult to tell the difference. These four main criteria will provide installers and integrators with a clearer picture of the differences between models and manufacturers, allowing them to make the best choices for each customer and application.
Spec Sheets vs. Reality
Spec sheets may seem like a useful tool for evaluating cameras, and in many cases true. However, dealers and integrators who lean too heavily on the data provided by a manufacturer do so at their own peril.
“Unfortunately, there is no standardization around how camera specifications are written, which has been problematic in the security industry,” says Sony’s Allen Chan. “Camera specification sheets appear to offer comparable features across the board, but they do not tell buyers the true capabilities of a product.”
That’s why in nearly every case, it’s important to field test cameras, preferably from multiple manufacturers, in a real-world situation to determine which will deliver the best performance for a particular application. This is particularly important when high-quality video is a primary goal.
“The dealer or integrator always should test the cameras they are considering in as true-to-life a situation as possible in order to really evaluate camera image quality,” says Jeff Whitney of Arecont Vision.
One of the most common discrepancies between spec sheets and reality is lighting performance, says Doug Gray of Hikvision USA and Hikvision Canada. While the numbers listed on the spec sheet may technically be correct, manufacturers may achieve them using any number of settings.
“Light levels are often fudged in various ways. The lux reading should be associated with a specific lens aperture, like F1.4, F1.8 or whatever it may be because you have to measure light level based on a specific reading through a lens,” he says. “A lens with an F-stop of 1.2 will show much better sensitivity than F1.4, where less light is coming in through the lens. You need to make sure if you’re comparing light levels, you’re also comparing the F-stop levels between them.”
Coloring in the Lines
Given the importance of video in the post-incident forensic investigation, a camera’s ability to accurately reproduce color is essential. However, not all cameras are pre-set with video surveillance in mind.
“Was the subject’s hair red or brown? Was their shirt dark blue or black? What color was the car? Often when a camera is viewed on screen, the colors seem very vivid and image quality looks similar from one to another,” says Arecont Vision’s Jeff Whitney. “But the dealer or installer should always be sure that those are the actual colors in the scene, as some cameras are set not necessarily for optimum color accuracy but to produce eye-popping images.”
Therefore, dealers and integrators should look for cameras that produce the most natural colors in a scene, and the only way to do that is with testing.
“To ensure maximum image quality, systems integrators will need to check the white balance to confirm that the camera is capturing accurate color in daylight, and then repeat the same check in night mode,” says Tim Shen of Dahua Technology.
Whitney offers some advice for the best way to approach testing.
“Take a sample like a leaf on a bush or an article of clothing and compare the actual sample to the screen. Is the image different from the actual article? Often, the most vivid images may not be the most accurate,” he says.
For more on the topic of video image quality, visit SDM’s website where you’ll find the following articles.