The Many Factors that Determine Video Quality

What Determines Video Quality?

Video quality is one of the least understood and most complex aspects of digital video; it’s no wonder why this is our most popular question from customers. The actual quality of your video is determined by many factors that work together in a complex synergy to create a digital video file. There are many myths about video quality which this article hopes to dispel.

Video quality is dependent on a balance two primary devices; the camera or “input” and recorder. Each of these two devices has several sub-factors that work together to determine the overall output quality of that specific component.

Camera – Factors That Determine Video Quality

The output quality of camera is derived from the performance of five main components; lens, image sensor type, image sensor size, effective pixels, and the signal processor.

Lens                                                                            Impact: 2/5

            The lens is used to direct light onto a camera’s imaging sensor. Lens clarity is vital to the performance of the lens. If the lens is not clear it will not project the maximum amount of available light onto the image sensor. Lens uniformity and consistency is also important; any deformation or change in refraction properties on the lens will redirect the light and project a deformed image onto the image sensor. The size of the lens is also important as it will spread out the incoming light over a larger area reducing the impact of small imperfections.

Image Sensor Type                                                  Impact: 5/5

The image sensor’s task is to convert light into a digital value that can be read by the processor. It does this through an array of hundreds of thousands of cells called pixels. Image sensors can be broken into two categories; CCD and CMOS. CCD sensors are generally more sensitive, have less noise, and better low light performance then the CMOS technologies. This is due to CCD using a cleaner transfer method of the actual pixel value. CMOS technologies have a lower cost and much better power efficiency compared to the CCD type.

Image Sensor Size                                                    Impact: 4/5

The size of the image sensor is also critical to image quality. Image sensors range in size from less then a few millimeters to over an inch. The larger the image sensor the more area the chip will have to read the light and thus provide the most accurate value. Some cameras have 3 image sensors which combine to create a more accurate rendering by dividing the task or reading red, blue, and green color; 3CCD cameras generally require more light.

Effective Pixels                                                          Impact: 3/5

Images sensors are comprised of arrays of pixels ranging from a several thousand to millions. These pixels read an “analog” light input which is represented as a digital value. The more pixels you have the “sharper” the image will be which creates a better rendering of the actual picture.

Signal Processor                                                       Impact: 3/5

The signal processor converts the data sent by the image sensor and corresponding circuitry to a format which can be transmitted and read by the recorder. Not only will the quality of this chip have an effect on the video but the format in which is converts the data to will also play a large role. Signals can be broken into two distinct categories; digital and analog. Since the inception of video technology analog has had a tight grasp on the video market and almost all signals were transmitted using an analog format. Recently advances in digital technology have brought digital transmission to the main stream.

Analog signals can be broken down into two components, luminescence (luma) from which we derive brightness, and chrominance (chroma) from which we derive color. Chroma can be further broken down into a red, green, and blue component. These components can be transmitted in different ways which affect the video quality.

Analog video is commonly transmitted in the form of Composite, S-Video, or Component. Composite transmits chroma and luma through one line. This is the most popular transmission type but limits the quality of the video. S-Video separates the luma and chroma signals providing a slightly clearer image; and component separates chroma into a two components and has a third line for luma which also provides a clearer signal. The drawback to this is the physical number of wires that are used, which has kept most systems on the standard composite format.

Analog video is transmitted in horizontal lines of resolution. In general the more lines of resolution the higher quality picture you are going to have, given the other factors. Think of this as the data rate of analog video. The higher the rate, the more information there is for the DVR to record.

Digital signals can be transmitted through DVI, HDMI, or HD-SDI. Unlike analog, digital formats are much more evenly matched. Digital signals are less prone to show noise in the video or have a distorted signal when compared to analog. Since these digital formats are newer to the market, they can be less reliable and more expensive.

Recorder – Factors That Determine Video Quality

The quality of your video is also dependent on the quality of your recorder. The task of the digital video recorder (DVR) is to convert an input signal from the camera and create a video file which can then be stored, edited and viewed. Digital video recorders are available in many forms and configurations from mass market to application specific DVR’s. The quality of video that you will obtain from your recorder is largely dependent on encoding and resolution. Encoding can further be broken into compression type, bit rate and encoding efficiency.

Encoding                                                                   Impact: 5/5                 

Encoding is the most important factor of the DVR in determining your video quality. Encoding is the process of converting the camera signal into a readable file; which can then be viewed on a television or computer. Encoding has three important properties that should be explored; compression type, bit rate, and encoding efficiency.

Compression type                                                    Impact: 4/5

            Compression is the process of reducing the amount of data that is recorded. Uncompressed video is not feasible for most DVR’s as the file size and data rate would be too high for current hardware and media. One of the biggest components of video compression is motion compensation. This involves a full frame image that is divided into sections called macroblocks, and then the following frames are recorded as movements of those macroblocks rather then the pixel data itself. This greatly reduces the amount of data that needs to be recorded while still maintaining good video quality.

There are many different compression types on the market. The most popular types of compression are MPEG-2, MPEG-4 Part 2, and MPEG-4 Part 10 also known as H.264. These standards define the way in which the video will be decoded.

Decoding methods were designed with different applications in mind, thus have their advantages and disadvantages.

MPEG-2 was designed for professional and broadcast video and was later adapted as the DVD standard. This method allows for higher video quality with the drawback of having larger video files. MPEG-2 is still the best choice for pure video quality as it retains more accurate pixel information.

MPEG-4 Part 2 was designed for consumer grade and video communications where bandwidth was more important then quality. This allows for reduced file size with the drawback being reduced quality. This is the compression type found in most consumer grade recorders.

MPEG-4 Part 10 was designed to compress larger streams such as HD video and transmit using existing infrastructure. Part-10 offers more efficient compression then Part 2. This is the compression standard adopted by most consumer grade HD products.

Bit Rate                                                                      Impact: 4/5

The amount of data the recorder saves over time is called the bit rate. The greater the bit rate recorded, the better your media player is able to replicate the actual scene captured by your camera. To put this in perspective DVD quality is approximately 5Mbps or 5,000,000 bits per second and HDTV quality is about 15Mbps. By raising the bit rate you loosing less data in the compression process there by creating a higher quality video.

Bit rates can either be variable or constant. Variable bit rates can change the amount of data recorded based on complexity of the video in order to keep consistent quality. Since the amount of data recorded is correlated to the amount of movement in the video, variable bit rates can be used to minimize file size while maintaining high quality Constant bit rates will consistently record the same amount of data regardless of video conditions, in effect sacrificing quality. Constant bitrate are much better suited for streaming applications where video is not being recorded. Of the two, variable bit rates will provide significantly better quality then constant bit rates.

Since variable bit rates are more complex and require more processing power to encode, some DVR’s will not record using this method. When choosing a recorder be sure to ask the reseller or manufacturer if it records with a variable bit rate or constant bit rate.

Encoding efficiency                                                            Impact: 4/5

Encoding efficiency should not be underestimated as it can have a much bigger impact on video quality then other more discrete properties of the recorder. Since compression type only defines the way in which the video is decoded, different encoding methods can have a significant impact on video quality. That means that you can have two recorders with the same statistics; compression type, bitrate, and resolution; and one may record much better video then the other. Unfortunately there is no easy way of determining the actual quality and efficiency of the encoding without knowing the specific encoding chip used or comparing the video visually. A general rule of thumb is that you get what you pay for, but you can never go wrong by comparing raw video.

Resolution                                                                 Impact: 3/5

Resolution is another key ingredient to having a high quality video, as it determines the upper limit of the video quality. Resolution of digital video is the number of pixels that appear in the video and it is measured in horizontal pixels and vertical pixels. It is a common myth that higher resolution; high definition, means higher quality. Since resolution determines only the number of pixels in an image, it does not necessarily mean that every pixel is going to have useful and accurate information that contributes to the video quality. Instead, resolution is able to define an upper limit by requiring more quality information as the pixel count increases.  With the limitations of cameras currently on the market along with the write speed limits of media, which can determine the bit rate, it is typical to find standard definition systems recording higher quality then high definition systems.

It is important to remember that all of these factors work together to create a high quality video. All of these factors experience a diminishing marginal return, some more drastic then others. It is imperative to find a system that best utilizes all of these factors to create a well balanced system to best optimize your video quality per dollar. We hope this article was informative and helps you to find the system to best suit your needs. We encourage questions on comments to be sent to sales@datatoys.com.