wattwire

silicuda, Man, your situation really sucks. I can't imagine the stress you must be going through. I have a high resolution (700tvl) analog license plate camera that I've been meaning to test (for resale purposes) for quite awhile and this seems like the perfect excuse to get going on this project. If you conduct the following information gathering exercise, I'll do my best to simulate the conditions at my local and post test images: 1. Determine the best location at your place to mount the LPR (lic plate recognition) camera. This spot will be the best for seeing cars head on or from the rear--with an angle that is most direct. Of course, the closer you can get to the road, the better....but should also be able to mount the camera about 3m high (to make it out of reach of the general public). 2. We will call the above spot "Pt A". From Pt A, measure the distance to the middle of the road where a captured license plate might travel. Note, this isn't where the camera will be focused; but rather the shortest distance at a perpendicular angle (90 deg) from the camera to the middle of the road. We'll call this "Pt B". I'm assuming that there's a road in front of your place that runs in a straight line. 3. Now take a second measurement, the distance from Pt B to where the camera will be focused for capturing plates. My camera has a 6-60mm varifocal lens, so the overall length from this point to the camera shouldn't be more than 35m. We'll call this "Pt C". 4. Take / post the following digital pictures: a) From the camera location (Pt A) to Pt C during the day, preferably of a car passing by. b) Same as above, but at night c) Daytime: from A to B, B to A, C to A, just to see if there are any other factors such as branches, mounted lights, or anything else that might interfere with capture. I also have a Dahua 1080p 8-22mm with IP camera with 36IRs I can test. I should also explain that the above Points A,B, and C form a right triangle which will allow me to calculate the camera angle (to the straight line of the moving vehicle) and the distance from the camera to the capture area. I mentioned 35m above as the maximum distance because with a 60mm lens fully zoomed, the horizontal view area is about 3M wide. This can be figured out with a lens calculator, like the one on Pelco.com. Also, this might be a silly question, but do you currently have a DVR installed? If not, you have to think about how all this is going to be recorded an retrieved later. -d

A search on Newegg of "powerline networking" shows these devices have come way down in price. You can now get a PAIR (Tenda P200) of adapters for $30.00 w/ free shipping. I've used other brands with great success. They're fast and easy to install.

Nice shots Shockwave! What Dahua model are those from?

Even with a good LPR camera, it's going to have to be aligned almost directly in line with vehicle travel direction. This is going to be hard to achieve on a public street unless your property is at the end of a "T" intersection.

I sell and install Dahua IP cameras and am happy to share my actual power consumption measurements for a few models (day / night): HDB3200 (your 1080p mini dome w/ no IR): 3.2 / 2.9w HDW2100 (720p eyeball dome with IR): 2.1 / 3.1w HFW2100 (720p small cyclinder with IR): 2.4 / 4.8w HDBW3202 (large VPD with IR): 5.0 / 6.1w Notes: o These are short runs for testing of < 30'. For long runs there surely will be significant voltage drop, so I did a test and connected one of the cameras up to a 200' cable but saw no power output drop from the switch (it was previously on a 10' cable). So the camera was able to handle a voltage drop okay. o I used a managed POE switch to take these measurements which made it pretty easy. Now back to your original question--if you're only planning on having 5 cameras, the most economical solution may be to buy a switch with 4 POE ports and then use an individual POE injector for the 5th camera. The Trendnet TPE-S44 and ZyXEL ES1100-8P are a couple of models to look at for good price / performance. The Zyxel has a higher rated power budget but you probably don't need that. -d

Yes, different metrics. This is the key to my confusion. The arithmetic of reducing the horizontal resolution to accommodate a 4:3 aspect ratio is pretty straight forward. But isn't the CIF standard for NTSC already set at a 4:3 ratio? (352/264). So the theoretical maximum number of horizontal pixels that would benefit the quality of a digitized image is 440? (330 * 4 / 3). The only reason I referred to CIF in the first place was because it was a known number of pixels. Hmmm, maybe we need to investigate what pixel is and what TVL is. Actually, I'm more confused than ever now. This is a good thing though, because I'm probably on the verge of learning something. Amir, thanks for having the patience in explaining this. Understanding how TVLs relate to pixel dimensions of digital images recorded by a DVR has got to be one of the most important concepts there is in CCTV. It plays a hand in how we select cameras, transmission methods, DVRs, and display devices, and more. So at the risk sounding ignorant, I'm going to take another shot at trying to get at the bottom of what a DVR can record (with no compression) from a baseband signal coming from a camera over coaxial cable. I think my biggest stumbling block is understanding what a TVL is. Here's what I've gathered from what I've read and been told: The maximum number of vertical lines going across the screen whereby each can be individual resolved. Said another way, if you could make a uniform bar code and place it in front of a camera, how many vertical lines there would be before the black ones merged together. I guess this also relates to the modulated electrical pulses in the signal (but we don't need to get this deep). I feel much more confident about what a pixel is as it relates to a digital image. It is the smallest unique picture element that can be modified. What I really want to know is: What is the maximum number of unique (not merged) pixels of information than can be recorded/digitized on a DVR coming from the baseband signal of an analog camera? If we took the uniform barcode example, how many white spaces and black bars could we individually distinguish before they started to merge together in the digitized image? Thanks!

Actually, that assumption is not correct. NTSC does exceed the resolution of CIF. As I said: "And in the case of a DVR doing a poor job of compressing the content, then the extra resolution NTSC has over CIF is probably lost," it is the combination of poor video compression and NTSC that might eliminate the extra resolution benefits of 4CIF. Bear with me a bit as I suffer from mild retardation at times. I'm probably missing something here about the basics. For a moment, let's put aside discussion about compression. My understanding is that TVL represents resolution going horizontally across the video scan line. I understand that the scan lines are interlaced and alternately refreshed. The resolution of CIF for NTSC is 352x240. So my original question was focusing on the fact that 352 > 330. What am I missing here?

Amir, that was a nicely written tutorial in the drama filled thread about camera resolution. I feel very lucky to have the attention of an experienced video engineer. More people need to point out that we're working with a signal standard here (NTSC) that doesn't dynamically change to accommodate higher resolution cameras and recording equipment. Having said this, I'm still struggling a bit trying to understand the maximum resolution that video transmitted over coaxial cable can be recorded at. In particular, I have to ask, "How is it that recordings at 4CIF look better than 1CIF given the fact that 1CIF exceeds the 330tvl frequency limitation of NTSC?" Is it possible that the baseband signal from cameras is exceeding the NTSC standard of 4.2 Mhz and approaching the 6.75Mhz "studio frequency" you mention in the other thread?

CIF images are going to be soft no matter how little compression artifacts there are. Sure, if you are bandwidth starved, then an H.264 at lower resolution may be superior but if motion is small, even that won't be the case. How do you get near DVD when you have one quarter the resolution? I'm working with a Blackfin DSP based H.264 DVR now and I can vouch for the concept that the 1CIF looks very good and exceeds the quality of "D1" recordings from other DVRs. Unfortunately, for my DVR, the D1 quality only increases marginally--even when testing with a very high quality 540tvl camera. There's definitely a difference between the two, but not nearly what you'd expect for having 4X the frame size. My theory on why 1CIF can look nearly as good as D1 is based on the belief that the image constructed in the DVR's frame buffer doesn't come anywhere near representing the resolution we think it does. As a matter of fact, I bet the real resolution is closer to 1 CIF than it is to 4 CIF. The frame buffer is 4 CIF, but the image has been expanded to fit. An analogy can be made from working with an image editing program like Photoshop. You can take a not so sharp looking picture and make it look good by shrinking it some (1CIF)...and if you expand the shrunk image back up, it doesn't look that much different from the original. However, there's very little you can do to improve the original picture because the image information just isn't there, even if you try to expand/increase the resolution (4CIF).

Hello, Wow...Thanks again wattwire. Ok.. Actually i would like to set up CCTV system for my house ( inside the house and house compound ). The distance between cameras and the recording equipments is about 250 meter , 450meters far ,maybe some of them further....And can you roughly explain what is Cat5e with passive baluns and Cat5e with active baluns ? Thank You In the simplest terms, video baluns are small devices that allow the use of UTP cable to transmit baseband video which is the output from security cameras. Baluns work in pairs--one is connected at each end of the cable and usually has a bnc connector to connect to the camera and DVR/monitor. The terms passive and active could be substituted for the "non-amplified and amplified". I do not necessarily endorse this company, but Active-Vision has a selection of baluns for sale and also provides more information: http://www.active-vision.com/Video_Balum_s/35.htm?gclid=COe_oqzJ6ZYCFQv7agodkheIPg Here is an article on what UTP is: http://en.wikipedia.org/wiki/Unshielded_twisted_pair#Unshielded_twisted_pair_.28UTP.29 There is lots and lots of discussion about using coax vs. UTP. In general, the advantage of UTP is that the cable itself is cheaper and smaller in diameter making it easier to work with. Note, the savings in the cable itself may be offset by the cost of the baluns! Coax is heavier and tougher. My opinion that it is more reliable based on the fact that there are fewer connections and the strength/durability of the cable. Fewer connections for coax because a balun requires two at each end--the wires going into the balun and the balun connecting to the video device. Having said all that, I use both UTP and coax depending on the circumstances. -dw ps, you never mentioned how far your run lengths are for supplying power to the cameras. Don't overlook the requirement to power the cameras. Improperly powered cameras is one of the main sources of CCTV trouble.

Survtech makes a good point and I'm beginning to get the picture. The associate has a large stock of non-Pelco conforming PTZs and wants to use them with cheaper DVRs that only recognize Pelco D (or vice-versa?). This is a little off subject, but I have to share the fact I've recently installed a Wonwoo WSJ-300 PTZ and was delighted to discover that it had a multi-protocol "auto sense" setting. All that I needed to do was set the baud rate/parity and it just works given that it receives one of a long list of protocols it recognizes. This could solve the problem of having a DVR and PTZ joystick controller that don't share a common protocol. For instance, I have an older Pelco KBD300a controller that only speaks Pelco P. The Wonwoo PTZ could work with this controller and a Pelco D (yes, "D" also works at 4800 baud) only DVR at the same time. Slick. -dw

I think you got some bad info about the range of Cat5e (or Cat6) video transmission using passive baluns. Or perhaps you left off some 0s The range is greater than that of coax. If you use active baluns, the range jumps well over 1km. Yes there amplifiers for coax but GENERALLY (others can pipe in) the strategy for long runs in order of distance is: 1) Coax 2) Cat5e with passive baluns, 3) Cat5e with active baluns or fiber. 4) Long range microwave (wireless) The best thing is for getting help in this forum is to state your scenario: 0) What is the general environment (new construction, commercial building, private property, in the jungle, on a farm, etc.) Also the climate can be a factor if the cable to be run outdoors. 1) How long is the distance between cameras & recording equipment? 2) What is the distance between cameras and power source? good luck, -dw

http://www.ikegami.com/cb/products/pdf/tech/simplesystem.pdf Make sure your coax has a solid copper center conductor and copper braided shielding. There should be no aluminum nor steel used by the cable. If you're using siamese cable, the ikegami document doesn't apply--the run lengths are much shorter for the power conductors due to voltage drop. I can't make heads or tails out of this question, "Is there any hardware like switch for coaxial ?" -dw

Oh man, that really sucks. Those Bastards! Even though it makes for good video, I feel really bad for you. Speaking from experience, being the victim of a crime like this is no fun. If you were nearby, I'd buy you a drink or two. Be on the lookout for one or more of these chumps in the future. There's some truth in the saying, "The criminal always returns to the scene of the crime." In any event, the odds are that one or more of these guys lives close to you. BTW, I assume you gave a clip or two to the authorities. Tell us the details--what form was the video in (USB Thumb drive? CD? .AVI? Tape?), what comments if any did the police make about the video? Cheers, -dw

Whoops, had to edit my message to get the polarity of the power correct. That's what I get for making late night posts while tired. The common ground is true, I was just thinking in terms of the output connections normally seen with a dome camera. I was assuming that the camera didn't have a microphone. If it does, then Soundy is right on with the wire assignment.