Buyer beware when comparing video systems
- Published: June 01, 1995, By Thome, John D.
This is the second in a three-part series on video web inspection. The series will present the technological basics, define concepts and principals essential to successful adaptation of the technology in the pressroom and will explore the future of web inspection. This month's installment presents a comparison and evaluation of the types of web inspection systems.
Just as they would with any other electronic product, video inspection manufacturers promote features, technical features. Almost every manufacturer boasts about technical specifications and claims product superiority. You hear terms like refresh rate, three-chip CCD, diopter, EPROM and the list goes on.
Unfortunately, if you don't have a basic understanding of the technology and how it works, just like computers, it's easy to let a fast-talking salesperson lead you astray. For example, you could be told that a PCI bus makes a big difference to image quality and color reproduction, when in fact it has nothing to do with color reproduction or image quality. How would you know?
In this installment, we will attempt to demystify the technical side of video web inspection. We'll examine each of the major components of a typical system, explain their form and function and address major questions to which you'll need answers.
Aren't All Electronics the Same?
Nothing could be farther from the truth. There are different quality levels of electronics just as there are for any product or product component. There is industrial-strength quality and there is consumer-grade quality, and a cost associated with the difference. In electronic goods, these differences can mean large variations in the quality and reliability of the electronic components. When are the boards made, and what is the skill level of the work force that made them? Are the boards hand soldered or machine stuffed? Are the electronic components surface mounted? The answers to all of these questions have an impact on reliability and durability and should be high on your list of requirements. Obviously, the construction and quality level of the system will affect maintenance and service, and with electronic devices it is easy to drop some serious dollars on service or repairs.
When comparing video systems, start at the heart of the unit with the frame grabber. It is the printed circuit board that digitizes and processes the image. Video inspection systems today use either 16-bit or the more expensive 24-bit boards. A 16-bit board is capable of generating 256 colors, while a 24-bit board can generate 16.8 million colors. The 24-bit frame grabber will deliver more gradients of color, but frankly, it's difficult for the average person to discern the difference.
Built-in static and surge protection devices are another feature to demand. Some manufacturers today put internal components in their equipment to help control and minimize electronic static discharge and voltage surge and to minimize the cumulative impact static has on electronics. These components add another level of reliability to the system, but they also add cost, so be sure to ask each supplier precisely what kind of protection their equipment has built in, how their systems are tested and what level of electronic static discharge their equipment will withstand without failure.
Field testing has documented the value to the converting industry of custom engineered electronics, heavy-duty protective housings, high-capacity fans, and static and surge protection in a video inspection system.
Such systems have repeatedly been found to be significantly more reliable in the press room - requiring fewer repairs less often. Video inspection systems that are essentially consumer-grade personal computers have been found to be far less reliable in the press room environment. Over time, the cumulative effects of static, moisture, heat and dust, which is the worst enemy of a personal computer, are just too much for the equipment to handle.
Single Versus Parallel Processing
As video systems become more sophisticated process-control tools, the need to process huge amounts of information, and to do it in milliseconds, becomes a tremendous design challenge. Some manufacturers have elected to utilize PCI bus in their equipment, and along the way, they have made significant claims about the affect of the bus on system performance.
A PCI bus is a new style bus and will most likely become the standard for personal computers in a year or so. Its primary advantage over other bus designs is that it is 64-bit and operates at processor speed rather than bus speed (8 mHz versus 33 mHz). Its impact is that it makes transfer of images between the frame grabber and memory faster.
The PCI bus will have no impact whatsoever on image quality, color reproduction or image-processing speed for high-end features like color monitoring, bar-code monitoring and other forms of defect detection.
The most successful manufacturers have elected to build systems utilizing a number of microprocessors and to employ a design scheme known and parallel processing.
The most advanced and sophisticated computer systems all use parallel processing because the advantage is maximized throughput.
Distributing the processing power across the entire system provides much higher throughput, and to the converting industry, that means faster response time and a more flexible interface for the user.
Think about your own experience with Windows.
Even with a Pentium processor, if you open too many documents or windows at the same time, you can get bogged down and the system becomes unresponsive.
There is plenty of power, but there is not an effective distribution of that power.
Critical Elements in a Video Inspection System
There are a number of design elements that are extremely important to the function, reliability and quality of any video inspection system.
They are simplicity, camera quality, press synchronization, the lens and strobe system, traverse design, the monitor, the electronics just discussed and the controls or interface used to run the system.
Ask questions such as: Is the system easy to learn? How much space is required to mount the camera? What is the image size at maximum and minimum zoom?
Why is Simplicity First on the List?
No single characteristic has a greater impact on the overall productivity delivered by a video inspection system than simplicity of operation.
Years of experience and research designing and building video inspection systems have proven over and over that the simpler it is for someone to learn and use a video inspection system, the more it is used.
Maximum productivity is obtained when system functions are executable in one or two simple steps, and the steps should be intuitive so they are easy to remember. The fewer the number of controls, the better.
Color-coded controls are learned faster and remembered longer. Complicated control panels and remote controls with tiny touch pads greatly impede use of the equipment.
What You Need to Know About Cameras
There are three criteria to use when evaluating the camera in a video inspection system: total resolution, linearity and accuracy of color reproduction. But before you can begin the somewhat complicated process of evaluation, you need to know more about video camera technology.
The printing process uses cyan, magenta, yellow and black, known as CMYK, to express all of the colors of the rainbow. Present electronic and video technology, on the other hand, digitizes all colors into red, green and blue, known as RGB. The translation process is not perfect, so color variations may occur, especially on less expensive equipment. Using the highest quality components minimizes the effect of this translation process and produces the most accurate color reproduction.
All charge-coupled-device cameras use at least one integrated circuit chip with thousands of light-sensing elements. A typical single-chip camera in a video inspection system may contain as many as 480 lines in horizontal resolution of these light-sensing elements, with 786 elements in each line in vertical resolution.
In a single-chip RGB-filter camera, an alternating pattern of red, green and blue filter are placed in front of each element. The camera requires one red, one green and one blue element or pixel to make up a single dot on the monitor. The true resolution of the camera is really 330 lines horizontally times 484 lines vertically or 159,720 pixels.
A charge-coupled-device camera using a green-stripe filter will require two elements horizontally and two elements vertically to make up a single dot or pixel. This reduces the true resolution of the camera to 470 lines horizontally multiplied by 242 lines vertically, which amounts to 113,740 pixels, or 41% less resolution.
Three-chip cameras, on the other hand, use three chips, each with its own separate single-color filters. Using a prism, the light entering the camera is split so that all three chips see the same subject matter. Because of this splitting of the incoming light, three-chip cameras require considerably greater light on the images to be seen. Higher levels of signal amplification are also required, which can add unwanted noise.
The real value of a three-chip camera is in the added spatial resolution and quality of color tone it provides. Generally speaking, the larger the size of the image area seen, the greater the need for a three-chip camera. As a general rule, if you are viewing an image area larger than 7 or 8 in. in either direction, you will benefit greatly from the use of a three-chip camera. Also, if you are printing 6-point type or smaller, a three-chip camera is a must. Converters monitoring bar-code quality or doing a great deal of process printing will also benefit greatly by investing the additional money on a three-chip camera system.
Are Three-Chip Cameras Better for Monitoring Color?
Yes and no. There is no question that three-chip cameras provide better spatial resolution, more pixels represented in a given area or space. They don't, however, have greater total resolution than a single-chip camera. Remember the discussion of cameras and filters: using a single-chip camera with an RGB filter, three pixels are homogenized into single color on the monitor. With a three-chip camera, each pixel is individually represented on the monitor. That's greater spatial resolution.
The impact on color is most noticeable for process work. Color tonal quality is significantly better with a three-chip camera than a single-chip camera. For applications requiring line and screen work such as laying down areas of uniform solid color, the three-chip camera won't make a significant difference versus the significant additional cost. Bear in mind that a three-chip camera will add as much as $5,000 to the $10,000 cost of a system.
Another issue to consider is image area. If improperly designed, a three-chip camera will cut the field of view of the system approximately in half.
Format refers to the actual size of the charge-coupled-device chip. One size is not inherently better than another. The primary reason for different sizes of chips is due to the production needs of the chip manufacturers. Four times as many 1/3-in chips can be produced on the same piece of silicone compared to 2/3-in. chips. Size is irrelevant. It's the number of sensing elements or resolution that's important.
Lenses and Diopters
A critical component of the camera is the lens, whose job it is to concentrate light and aim it at the charge-coupled-device chip. The quality of the lens will affect image sharpness, image clarity, color accuracy, optical contrast and consistency across the entire image area.
Diopters are close-up lenses used in video inspection systems to govern and enlarge the image area. Increasing the image area will have a cost in reduced magnification.
For example, a 6:1 lens with a standard + 7 diopter sees a maximum image area of about 2 1/2 in. x 3 1/4 in. but will reduce the maximum magnification about 25% to 15 times. Similarly, a 10:1 lens with a +4.6 diopter will see a maximum image area of 3.8 in. x 5 in. and provide 25 times magnification. Changing the diopter to +2.6 results in an image area of 7.5 in. x 10 in. but reduces the maximum magnification to 12 times. If you're printing process work, you must have a magnification of at least 15x.
As you increase the size of the image area seen, it will also be necessary to mount the camera father from the web, which means greater strobe power will be required to properly illuminate the web and produce a quality picture. Be sure to consult with your vendor to determine the right combination of lens and diopter for your application.
Evaluating Cameras
When comparing different video inspection systems, it's the total resolution of the camera that determines the system resolution.
The camera also determines the system's ability to reproduce colors accurately and its ability to reproduce areas of high-intensity detail. One way to evaluate this is to test the camera's linearity or its ability to transfer color accurately under changing lighting conditions.
Some cameras are designed to boost the colors they deliver, which means the picture may look more vibrant, but those colors won't be true reproductions of the original color.
Accuracy of color reproduction is also influenced by signal to noise ratio and the light sensitivity of the camera. Both of these, in conjunction with strobe power, determine camera noise. The more noise in a camera, the less consistent will be the colors it produces. Noise is also evident in the form of the ghosting of images on the monitor, especially in detail areas, and by color variations on the screen.
Xenon strobes are used in video inspection systems because they provide what is referred to as the flattest spectral output, which simply means they deliver uniform light over a wide range of colors.
Xenon strobes are also known to produce the best color balance, or consistency from flash to flash, and generate short-duration, high-intensity light. This is crucial at high-magnification levels and with the short focal-length lenses being used in video inspection systems today.
Circumferential strobes should be avoided as they produce hot spots on varnished, ultraviolet-coated or reflective substrates.
Monitors
Like speakers in a stereo system, the video monitor is the platform on which images are produced, and just like speakers, there are many levels of quality and price. There is also a direct correlation between quality, price and longevity.
Some high-quality monitors contain built-in voltage devices to help them tolerate voltage spikes without causing damage to the monitor. A hood on the monitor can help reduce disturbing glare.
There are two basic kinds of monitors on the market, interlaced and non-interlaced. Both paint or refresh images on the screen one line at a time. Remember, there will be upwards of 800 lines to refresh.
The interlaced monitor first paints all of the even-numbered lines, and then it goes back and paints all of the odd-numbered lines. The result is image flicker.
A noninterlaced monitor paints all of the lines every time the screen is refreshed. As a result, picture quality is superior. There will be no flicker as long as the scan speed is fast enough, usually 60 mHz to 72 mHz.
Monitor resolution is unrelated to use when evaluating a video inspection system. The camera, not monitor, resolution will determine the number of horizontal and vertical lines in the image.
The refresh rate refers to the number of times per second the monitor paints the image on the screen. A standard television paints 15,750 lines/sec., which means images are repainted on the screen 30 times/sec.
Most noninterlaced video inspection systems paint 31,500 lines/sec., or 60 times/sec. The most commonly used technology today in video inspection systems paint 38,000 lines/sec., or 72 times/sec.
A higher refresh rate reduces eye fatigue and allows closeup work on critical color areas of the web. It also contributed to greater edge definition, stability and sharpness in the images seen on the monitor.
Dot pitch refers to how small a dot the monitor can reproduce, with one dot equal to one pixel. The lower the dot pitch of the monitor, the finer the image detail the monitor can produce. A super-fine-pitch SVGA monitor, for example, can have a dot pitch of .026, which is the lowest presently available.
Monitors must also be evaluated for distortion or picture squareness. Less expensive monitors use magnets to correct the geometric distortion of electron flight paths, which are the scanning beams. Over time, these magnets tend to fail, resulting in picture distortion.
The single most important way to evaluate a monitor is in how accurately it reproduces color. You should also evaluate it for convergence, which is the alignment of red, green and blue dots on the screen.
Following are three quick tests designed to help you evaluate components and performance of different video inspection systems and to make more knowledgeable decisions.
Monitor Convergence Quick Test
Put a sheet of clean, stark-white paper in the viewing area of the camera. Set the monitor brightness level at 75% of full. Zoom in all the way on the white paper, and then intentionally defocus the system so that you don't see the graining of the paper. Now stand back and take a close look at what you see.
Lower quality monitors will have hot spots, color variations in the corners or dog bone images on the screen.
Camera Noise Quick Test
Look at an image with small text and lots of detail. Magnify the image to full power and look closely at what is on the screen. The less ghosting of the images you see, the lower the noise level of the camera in the system.
A second way to test noise in a camera will require a sample with lots of color, both screen and process.
Because you will need to look at the same image each time for this test, the press cannot be running during the test. Manually secure the image under the camera, and then make sure that the system is in manual or snapshot mode so the camera is taking continuous pictures of the same image.
Carefully watch the images as they appear, paying particular attention to color. The variations in color which you see from one image to the next are a visual representation of the noise in the camera. The less variation you see, the lower the noise level of the camera. This translates into more accurate and consistent colors.
Camera Linearity Test
If your supplier does an in-plant demo, test the linearity or accuracy of color reproduction of the system's camera. Examine a variety of colored images under both very high and very low light conditions.
Do this by manually adjusting the iris on the system. Look at colors like yellow, orange, red, brown and blue. You are trying to see how accurately the colors are reproduced under extreme light conditions.
