As promised in the last installment of this column, we're taking this column to discuss resolution. Both analog and digital need to be discussed so that you can understand the differences and the consequences of those differences. You need to understand the language of resolution in order to competently make decisions about how much is enough, too little or more than enough. So let's start with a simple definition of what resolution really is. In both analog and IP applications, resolution is a number that we assign to describe the detail of the projected, transmitted or viewed video image. The higher the number, the better or finer the detail of the image will be. It is important to understand what numbers we are speaking about and what exactly they represent.
With analog, our resolution is created when a horizontal scan line crosses a vertical scan line. At each point of crossing, a dot or pixel of energy is created (See image 1). More scan lines means more pixels. The more pixels an image has, the higher the resolution rating the image has, and with higher resolution, greater detail can be seen in the picture. Remembering previous lessons, we know two things. First, we are restricted by the National Television Standard Committee (NTSC) to 525 vertical lines of resolution. This was to conserve bandwidth and prevent crowding of our airwaves back in the early days of television. Second, we create an analog image in two steps. We first create the odd horizontal scan lines and then the even horizontal scan lines.
This creates two fields or half pictures. The two fields combined make for one frame. We have 30 fields or 30 frames per second. This is called 2:1 interface and was created to allow recording of the video signal onto a one half inch wide magnetic tape. The whole frame takes up too much bandwidth to be recorded on such a small tape. Therefore, resolution equals bandwidth or a lot of frequencies. It also means that we only see one field (the odd one) during the playback of a video tape. This is why the playback never seems to look as good as the live image. We have 30 fields per second being played back as opposed to watching 30 frames.
When we discuss analog resolution, we always refer to the "horizontal" resolution. This is because we can monkey around with the amount of it where as the vertical is fixed. We are still limited, however, to how much detail we can produce because of the capability of our cameras to produce, the cable to carry, and the various pieces of equipment in between to emulate the bandwidth of the signal. The net result of analog resolution however is always four-fold:
1. The analog image can be enlarged several hundred times before becoming greatly distorted. This is because the pixels and the spacing between them enlarge exponentially.
2. When we discuss the resolution of an analog image, we are speaking about the "frame" resolution. That is, if a camera is rated at 400 horizontal lines, each field produced is only 200 lines. Most equipment that converts the analog to digital and then back (for the purpose of enhancements or control) cannot handle the bandwidth of an analog signal. Therefore, the output resolution is usually much less than the input.
3. Although both fields are recorded on a tape, we only see one field played back. The even field is ignored unless you are working with a frame recorder in which case every other odd field and every other even field are ignored. Remember, the video recorder in our industry can only play back 30 fields, regardless of how many are presented. A forensic person, using the proper tools can play back either or both fields for study and exposition.
4. Our analog image will always be four parts wide to three parts high. That is to say, if we have an image that is 12 feet wide, it will be 9 feet tall. Therefore, increasing the resolution is a matter of cramming more dots into the same size mold. The detail improves, the bandwidth increases, but the image stays the same size.
Alright, let's move on at the same rate as the industry. Let's discuss digital or IP resolution. The first step to IP is to realize that we no longer create two fields and combine them to make a single frame of video image. We create a single image at a time; it's a full scan, so to speak. The second part to be aware of is that we do not work with horizontal and vertical scan lines, creating dots on the screen. Instead, we create digital pixels or squares of colors. Imagine one digital pixel or square of color for each four analog pixels dots. The analog pixels could be thought of as the corner points of the digital square. This is why it takes so much more bandwidth to create a single image of high resolution, digital video as opposed to a single field of analog video. A square carries a lot more space; space equals bandwidth or frequency; frequency equals resolution or detail.
The field of view for IP or digital cameras works in much the same manner except that the image is no longer fixed at 4x3 and we don't count horizontal or vertical sweep lines. With digital, we reference the Common Intermediate Format (CIF); which is measured by squares of color, a.k.a. digital pixels. The CIF resolution standard and definitions were driven by the digital video recorder (DVR) industry and align with the analog (NTSC and PAL) standards. The resolution ratings of all digital or IP cameras are multiples or divisions of CIF. Consequently, we are speaking about true grid resolution -- the number of squares of color that we have as based upon rows and columns. Our most common digital resolutions (to date) are as follows:
- CIF (CIF) 352 x 240 pixels - 84,480 pixels Quarter of CIF (QCIF) 176 x 120 pixels, 21,120 pixels
- 4 times CIF (4CIF) 704 x 480 pixels, 337,920 pixels
- 16 times CIF (16CIF) 1280 x 1024 pixels, 1.31 megapixels
- 25 times CIF (25CIF) 1700 x 1200 pixels, 2.04 megapixels
- 36 times CIF (36CIF) 2112 x 1440 pixels, 3.04 megapixels
- [Information is presented as Ratio, Label, Horizontal/Vertical Pixel Array, and Grid]
To keep perspective, some of you have been listening to pixel resolution versus grid or CIF, so let's do the math in a different format. Let's talk about pixel resolution. CIF = 352 x 240 = 84,480 pixels = 84.5 kpr (kilo-pixel resolution). 4CIF = 337.92 kpr. 16CIF (also referred to as HDTV) = 1280 x 1024 = 1,310,720 pixels or 1.3 mpr (mega-pixel resolution). To understand what this means to image size, see Image 2 above.
Continuing forward with perspective, those of you that have photographic digital cameras will be happy to know (or perhaps not so happy to know) that a 4 megapixel image (approximately 2,000 x 2,000) is equal to 400 ASA film resolution and a 6 megapixel digital image is equal to 100 ASA film resolution. Our original digital standards were developed for the VGA monitors (Video Graphics Association) at 640 x 480. The field of view for each camera will need to be calculated in advance of system installation or design. This allows you to choose the proper focal length (FL) lens with very accurate and expected results, prior to spending a dime.
As with analog, the quality of the image or the detail that we are able to obtain, transmit or record is limited to our ability to first the image, carry the image, and store the image. Because, just as analog resolution equals bandwidth, so does IP resolution equal bandwidth or frequency. However, with IP we have a lot more latitude as to what we can do to preserve our detail. We can limit the number of images that are produced and transmitted per second. Do we really need 30 images per second? Not in most cases. The second thing that we can do is to compress the image. This is done via several different methods and will be the topic of our next column.
The biggest downfall of IP resolution is that enlarging the image is a matter of making small squares of color (or shades of gray) larger. Imagine a round ball made of square blocks. Imagine next that this ball is the size of a house and each block is one foot wide by one foot tall. When we view the ball from two blocks away, it appears round and smooth. But as we walk closer and closer to the ball, we see that it is not smooth. It is jagged, as only squares being placed into a curve can be. The key is to use smaller squares in the same space. However, that takes more bandwidth and the catch-22 grows.
See you next time as we discuss how to choose camera lenses.
A PERSONAL REQUEST FOR HELP TO THE INDUSTRY
For the past 31 years, I have been bragging to friends and family about the security industry and it gifts to the general public. So I am now addressing that same industry, personally with a request for help.
In about six weeks, a small Catholic School on the lower west end of Davenport Iowa will be permanently closing its doors. Sadly, this is happening after 100 years of dedicated service to the education of children of all faiths. This is a school where my wife and her brothers, my three sons and a countless number of other folks that are close to me, have attended. This is a school where the teachers and staff have worked for below average wages and given from their hearts for as many as 25 years. Because we are a small parish and an equally small community, our staff was never afforded compensation or retirement benefits. They will be saying goodbye with nothing more in their pockets other than the fond memories and broken hearts that years educating others will create.
Several weeks ago, I helped to found a committee of parents, parishioners and friends of St. Alphonsus School. The committee's sole purpose is to raise money for an employee appreciation fund to help these departing employees. All costs involved with our project, including legal, financial, postage and marketing have been donated by a wide array of professionals in the area. This allows us to use almost 99 percent of all money raised toward our employees.
Since the funds we are trying to raise are beyond the capacity of our local area, each member of the committee has reached out to friends, neighbors, and fellow workers. I am reaching out to the security industry as a whole. I realize that the world is full of good causes and sad cases. I also realize that this cause does not warrant national attention, and yet here I am. I am not trying to save the world or prevent any major disasters. I am only reaching out to an industry that has given me so much. I am reaching out to you, as an individual or company, to ask you to reach into your heart and help us to say thank you to a small group of people, a group of folks, that just like you, have given so much from their hearts for so long. Please don't think that any donation is too small or to large. All donations are fully tax deductible for those that itemize their taxes. Unfortunately our time is very short as we must close our efforts as of June 1, 2005. I personally have pledged to match 1 percent of all funds raised up to the first $200,000.
Please help me in this worthy cause by sending a check or money order today. These should be made out to: St. Alphonsus Employee Appreciation Fund and sent to the same name at P.O. Box 163, Davenport, 52805.
Thank you, from my heart for whatever you are able to do to help us.
- signed, Charlie R. Pierce
About the Author: Richard R. "Charlie" Pierce has been an active member of the security industry since 1974. He is the founder and past president of LRC Electronics Company, a full service warranty/non-warranty repair center for CCTV equipment. In 1985, Charlie founded LeapFrog Training & Consulting (Formally LTC Training Center), a full service training center specializing in live seminars, video-format certification training programs, plain language technical manuals and educational support on CCTV. He is an active member of ASIS, ALAS, CANASA, NBFAA, NAAA and SIA. He is the recipient of numerous security industry awards, and is a regular contributor to Security Technology & Design magazine. Look for his columns to also appear regularly via SecurityInfoWatch.com and this website's e-newsletters. He can be contacted via email at [email protected].
