cctvexpert

It all comes back to a few simple statements but the people who are hell bent on IP will never remove their hands from their eyes or ears. FACTS: 1. IP is merely a transmission method. 2. No one is going to run video on a data network 99.999% of the time it is run on a separate network. 3. Whether it is IP or analog both can be run via UTP. 4. IP vs. Analog is just a matter of am I doing the compression at the edge or further back in the architecture (more importanly - why are you doing that) 5. IP and megapixel are not synonymous. There are traditional IP cameras that are non megapixel and in fact are analog cameras with encoders built -in ostensibly which is what an IP camera is. 6. IP is clearly good for large wireless applications due to channel limitations on analog wireless - and until something better comes along it works well for High Definition imaging (but it has its flaws - low light problems, high storage requirements, high bandwidth requirements, rolling shutter problems, motion artifacts 7. and last but not least and here is where the arguments start - analog is cheaper and easier to maintain. But an IT manager loves the job security that comes along with managing an IT network.

You missed the point. IP is merely a transmission method. The question is if you have to run a separate network anyway, then why would you spend the increased cost for IP cameras when you can run analogue cameras with the same basic components and have no network issue and save money. And yes, the connections are passive not active and do not require user intervention.

1. Call up the Nevada Gaming Control Board or any other gaming commission for that matter and ask for a set of their regulations. Network cameras CANNOT be used for gaming operations. They can be used in some cases as NON gaming cameras. The reason is lack of reliability. 2. You can start to build fault tolerance into a network with self healing networks and layer 3 switches and all the bells and whistles - so i am not sure who is going to spend $1 million on a solution that can be solved with proven technology in the form of analog or hybrid for a fraction of the price. All you accomplished is providing job security for some IT manager. 3. IP is merely a transmission method. Both analog and IP can transmit video across a Cat5 cable. The only difference is where the video is compressed. One does it out at the edge and the other does it at the head end. For the privilege of doing that you at least double the cost and reduce reliability by a factor of xxx 4. IP has a place with high megapixel subject to low light and fast motion problems or in wireless - period. That is where they make sense. High megapixel has little value in a "closed" space with fixed ingress and egress points. In a casino they have a tight shot at each door entrance so they have your face as they would in any other building. No need for a megapixel burning it up at 10GB per hour. As far as identifying the card or the suit if you have a decent wide dynamic camera with a good quality low f-stop lens you won't have a problem. You can even see the numbers on a dice. 5. analog Video over UTP using baluns has been around for almost 2 decades now - there is no black magic. 6. The CCTV world still has not learned to play nice and other than one particular manufacturer that I know of even though there are video compression standards set by ITU and MPEG - the video manufacturers choose to slightly modify the native codec to make it proprietary. The problem is also older codecs cannot handle compression on live video very well and only a few companies that have been encoding and decoding H.264 for several years now do it properly. 7. MJPEG is from the stone age. So when we talk IP using MJPEG is like saying I have a state of the art futuristic car with parts from the 1960s. Like I said IP not ready for prime time and all it is - is a transmission method as all keep forgetting. 8. As stated all IP cameras start out with an analog sensor then get converted to Digital to perform its magic then get converted back to analog if being transmitted analog or it can remain digital if transmitting IP. Many, many, many of the IP cameras go analog sensor to analog to digital converter, then to digital to analog conversion back to analog to digital. So if you think they are ready for prime time good luck. . Both IP and analog are just methods of transmission. . Both can be transmitted across a UTP cable. . Both can follow 568B guidelines . One goes to a network switch and the other to a balun hub. (difference - IP uses bandwidth - analog doesn't . Analog is live video capable back to the DVR with pure image - IP is a latent image already compressed . With respect to PoE you still have the 90 meter rule and still cannot power IR and PTZ since they can use in excess of 70Watts and PoE is less than 13 and the new PoE+ is only 25 - 35 max. Last but not least the reason they have so many problems with the Megapixel cameras is the CMOS chips they use are a toy. Do your homework and you will find that the CMOS chips in the IP cameras are the same ones they use in your cell phone camera. The reason is there are NO CMOS sensors designed for the CCTV industry. What everyone has to remember is CCTV only gets its technology from what the consumer market eventually passes down. That is why in CCTV cameras we use 1/3 CCD and not 1/2 CCD which would be better because that is what the consumer market dictates. The only reason higher compression technologies such as H.264 came around was because of trying to get HD and other video to fit on a DVD. For your own testing take any IP megapixel camera and lower the lights and record fast moving objects - you tell me what happens.

Agreed they all have issues at times and it is hit and miss. What is still the best policy is if possible stress test the drives before sending them out. Many companies will have a burn in process for anywhere from a day to a week or until they fill the drive and have a couple of passes through. Again, this does not speak to the longevity but you can assume that if you got it to perform acceptably through the burn in phase your failure rates should be reduced significantly. Also important to note is alot of people make a mistake and put a "DVR" drive inside a RAID machine. There are drives that are meant for a DVR but yet not for a RAID but for the most part RAID drives will perform well with DVRs.

Aventura Technologies and I believe GE but maybe not in USA makes an H.264 embedded DVR for many years now. With respect to H.264 there are many different profiles and some perform better than others. Alot of progress has been made since it was introduced as a standard back in 2003. In fact, the latest version is know as scalable video coding or SVC, which is actually much better than the baseline AVC profile. It has the ability to properly manage substreams which is key when we talk about network transmission. H.264 while called MPEG4 Part 10, is in no way, shape or form closely related to the older MPEG4 Part 2, which utilized Huffman not Arithmetic algorithm. It is a horse of a different color.

rather then getting into a long winded discussion wait a few months and see. Analog is analog all cameras start out as analog. You say you can't transmit analog over twisted pair then what can I say - it is being done. It doesn't need a protocol - it is analog. The only thing right now waiting is the encoder at the DVR end to be able to compress the HD signal and there are a number of DSP's already on the roadmap. Rather than talk to me ask Micron and TI about the chips, Hitachi and Sony Broadcast about the transmission maybe you will believe them.

it is MPEG4 and if you believe in super compression then i won't burst your bubble also about the tooth fairy

i hate to differ with you but dont confuse MPEG4 with H264 they are not related. The motion picture experts group calls it MPEG4 part 10 and the internation telecommunications union calls it H264 they should not be confused with mpeg4 part 2. Completely different profiles and architecture. 264 is a complex codec. MPEG4 part 2 does not have slice based based prediction arithmetic algorithm deblocking filters use of multiple reference frames interlace management weighted prediction and it goes on: 264 goes down to 4 x 4 blocks mpeg4 16 x 16 mpeg4 uses huffman algorithm vs arithmetic for 264 they are not even similar in their profile characteristics. transmission with simple codecs cannot be managed. Its the DSP which has substream capabilities which controls the data flow. In the end seeing is believing. If it does what you say then fine.

so many things to answer First, all cameras start out as analog - again CCD or CMOS but analog image sensors its all in the A/D and D/A conversion. Baluns well balun stands for balance unbalanced and in fact want noise. Baluns and UTP have interference rejection and crosstalk immunity. I can take 16 cameras and run then across cat5 using baluns and a balun hub and not compress the video and have a pure signal. This is all done passive. I can punch down to a patch panel or anything passive with up to 10 cuts and follow 568B standards. Perfect image uncompressed live video no latency. If i take 16 IP cameras and encode there is automatically latency and no live image. It is all active, loss of the network, viruses, bandwidth congestion, and on and on. Oh yea and i am going to pay more. IP is a transmission method thats it. All the megapixel cameras start out as analog is what you forget. You are converting it to digital. So in order to make the megapixel camera HD you just put in line a D/A converter, there is no magic. As far as transmitting the HD there is no difference in the architecture and methodololgy. If you are running megapixel and using PoE that is Cat5 cable. Can go into a novel about transmitting video both analog and digital over UTP. My question to anyone is why are you going to pay more and get less. FYI, the IP cameras in the market do not use open standards. While the codec may be H264 or MPEG4 or whatever they are proprietary if they werent then you could simply run them on any windows media player which you cannot. If you have 16 different IP cameras from a half dozen manufacturers you have to hope your NVR solution supports them. If the NVR gets upgraded you have to hope the upgrade supports them. Last but not least you cannot software decode megapixel cameras without maxing out the CPU, its a fact of life. You can fight about IP cameras all day long but they have limited use and the real pros know this. Axis and others can spend zillions on advertising but they cannot reinvent math and science. Latency will always be a problem of math and physics and there is no fault tolerance. There is also the issue of what each of the IP cameras support in their SDK. One camera may have some basic functionality and not have others and similarly some IP cameras well the point is there is no consistency. Even within Axis or any of the others one company and a zillion different SDK's. Not only do you not have a single SDK for the entire codec family but it varies from manufacturer to manufacturer and from model to model. From the programmatic aspects there is no way to build in proper software testing. If everytime you add a feature to an NVR you have to test 100 models, it is impossible to regression test or have any proper quality control in the testing aspects. I can show you 100 casinos running hundreds and thousands of cameras running on analog platforms with virtually no downtime. Show me how many enterprise IP camera systems are out there and show me one with any fault tolerance. Much to the chagrin of the IP manufacturers it is the opposite of their BS. the further you scale the more complicated it becomes and the less value it has because of having to try to transmit and manage.

There is no such thing as "self learning" unless you have invented a neural intelligent algorithm which i doubt is the case since the processing power required would take up 100 cameras. People should be honest about analytics and tell it like it is. I have been working with analytics for a decade now and there are simply things it will do and wont do and hardware analytics do suffer in a number of areas. As far as having it self calibrating and self learning we can get into a whole discussion of math and science why that is not possible. And to even use the words plug and play with video analytics well shame on you. We have tested virtually every product out there over the years and watched a number of shoot outs by public agencies and they all fail in the same areas. That being said, analytics is a great tool but not a panacea. It will never be plug and play as each and every scene is different, each and every shadow is different, lighting is different, background activity is different, etc. Even the best software algorithms which do true scene content analysis and are light years ahead of the objectvideos of the world still need extensive calibration and optimization. If it is so good can you give 3 references of people using it successfully in an enterprise environment. There is your answer. Talk to the customers using it system wide.

264 is the best because it uses variable bit rates and only sends the changing frames down when necessary depending upon the configuration of the I, B, P frames.

IP is a transmission method, be clear about that. This is first and foremost. Once you decide on an IP camera you are now locked into a proprietary platform. Want to add another camera you first have to make sure the NVR supports it. Next issue add a megapixel camera how are you going to decode it for viewing since they use up 100% of the CPU you throw at it. See how long a server lasts running at 100%. Encoding at the head end makes sense using non-proprietary cameras make sense because you bypass the network and if you have 100 cameras that are all analog you dont have to worry about changing things out. You can argue till the cows come home but the problem with a proprietary camera is clear. Also processing in the camera is "limited" - look at an axis or any other h264 camera and try to run it at 30fps high bit rate. Guess what happens it drops frame and has issues. You can run an analog camera using 264 in a dvr at high bit rates, high resolution full frame rate. So again you are paying more and getting less. We can have this discussion in 6 months when analog HD cameras hit the market and oh by the way remember that the market for HD cameras is 1% - so we come back to 99% of the world still wants a basic non proprietary camera and if you think that is IP when there is a legacy world out there I have nothing more to say on the subject. You lose your network or get it hacked and you are SOL and is the reason government and gaming do not allow IP cameras.

I guess you didnt read. Not sure why you think you can use a fraction of the cameras unless they are megapixel and if you have a bar and its dark should be real interesting how you are going to address low light issues. IP is only a transmission method. Good Luck!

its not that simple.

ok guys here we go All video starts out as analog because the light level hitting the image sensor (CCD or CMOS) is analog. So the initial sensing is performed in an analog electrical form which is then processed to either a discrete digital number for each pixel for each color and then processed digitally using math on the digital signal processor (DSP) or it passes through directly as pure analog. This is where it is converted using an analog-to-digital converter. From the A/D converter it is either passed straight thru as a pure digital signal or converted back to analog with a digital to analog converter. So now that we have established that the image sensor (CCD or CMOS) are analog - then the question is what follows. The IP camera is confused in many ways because the only way it differs is in the transmission method as described above. With an IP camera you have latency in the camera and you can never view real time. Sorry to jump around but whether it is HD or non-HD it follows the same basic principles. So all that megapixel is that it uses the image sensor then converts it to digital and then stays digital. If you convert it back to analog the way you do an analog camera it would work just the same. There are HD sensors being designed. There are plenty of HD encoders in the broadcast world so it is logical that it is a matter of migrating the same technology to the CCTV world. In fact, HD encoders are comonplace in the IPTV world. What you have to remember with an IP camera is that while it is technically a PC device it is limited in its processing power and you can obviously do alot more at the head end where you can have more resources in the form of an enterprise dvr. Today there are DVR/NVR hybrids that can record H264 on up to 64 channels in real time and use combinations of hardware encoders and decoders in a single server chassis that makes alot of sense compared to trying to manage a network of an abundance of IP cameras. IP cameras do not scale well despite what the peddlers of IP say. If you are going to scale a solution with 100s or 1000s of cameras you are then talking about layer 3 switches, significant network management, bottlenecking issues, packet management, and maintenance concerns. Ghosting well you don't get ghosting on good quality dvrs in fact i have worked with some broadcast quality ones where you cannot tell the difference between the original and playback. Also understand the nature of IP and you can never view a real time live image as you have built in latency at the head end since there is latency from the process of encoding to transmission to the monitor. Last but not least there is a reason that no gaming commission in the united states allows IP camera for casinos. Why, because there is no fault tolerance, if you lose the network you are done. You can try to build clusters of network switches with all kinds of backups but there are issues and the costs go into the stratosphere. Well one more thing, a network device can not only be hacked but pick up viruses. If someone hacks into the network and they take it down, then what. In the DVR methodology you are more secure. While i am beating IP down a bit the fact is it does have a place but it is not a panacea. If you are in one physical building there is not a single reason for using IP cameras. It costs more and you get less. The use of video over cat5 and fiber when it makes sense will always be a more stable, secure and reliable solution. If you have video on 2 sides of the street you can still run one cable in a number of methods. Try to control by the way a ptz across networks it gives latency a new meaning. This can get into a looooooooooooooooong winded topic but i think the long and the short of it is as least from my perspective analog is not going away if you understand the technology and what can be accomplished. HD is an important step but is in its infancy and needs work. Customer expectations need to be managed. FYI, for 2008 the market for IP cameras was still less than 10% despite everyone professing its death. Look at the most recent Axis press release where they say the market is not what they anticipated and are cutting back. They blame it on the economy yet the market for analog cameras did not suffer and maintained a 90% share. The reason is talk to any enterprise installer and see how many successful installations they have completed in the 100s or 1000s of cameras on an IP platform.

how do you differentiate between animals and humans how do you account for changes in weather how do you account for lighting issues how do you calibrate (no analytics is plug and play) its an algorithm

H264 is by far the best solution especially now with the latest profile of scalable video coding. Nuvico is a good machine but an older codec.

you should be able to reach them they are located in the UK check www.digitalvideos.co.uk there should be all the contact info there

since the magazine publishers will never suggest anything that would hurt their advertisers maybe we should set something up where we can compare all these manufacturers recorded as well as remote viewing quality.

Aventura Technologies I believe works well on iPhone and other PDA devices. I am not sure but I think also Avermedia does also but not positive about that one.

What you should be using is mobile DVRs which are designed for operating outdoors and in extreme environments. While an embedded DVR may be able to work in more extreme conditions than a traditional PC Based DVR it still is not designed to be working outdoors for a myriad of reasons. When we have worked with Departments of Transportations we have either used encoders and backhauled to a controlled area and then recorded at that location or we just used mobile dvrs in NEMA enclosures.

Wireless is obviously a category unto itself which much has been written about. With respect to solar there are plenty of companies that make the products. The questions they will ask you are what is the power specifications of the camera obviously to determine how much draw and then the location of the wireless so they can estimate the amount of sunlight. You can count on about 4,000 for the solar for a fixed camera and double that for a PTZ camera. The wireless portion depends on the distance. There are point-to-point IP devices that can transmit over long distance non line of site or you can go the traditional way.

simple examples of hardware based systems with fixed parameters would be: Lets say you are interested in left objects around an airport for obvious reasons. There are luggage trolleys everywhere on an airport. People strand them everywhere. If everytime one was stranded an alarm went off then the system would become useless. So you would need a method first to identify what a luggage trolley is and then the difference between a full trolley and an empty trolley. This can only be done with software systems on the fly as opposed to hardware systems which would require major headaches.

hardware based systems such as the videoIQ are a problem. Since the core parameters are "fixed" you have little ability to teach the system to learn patterns and be able to exclude known items that would create false alarms.

what determines how close lines are together is the resolution that has been selected CIF is 352 x 240 2CIF is 704 x 240 4CIF is 704 x 480