woodyads

Sheriffa, just to help you understand why some installers are recommending single mode, and how much you should install Multimode preceded Single mode from a commercial perspective. Single mode is faster and goes a greater distance. There are still some proprietary interfaces that only support multimode because they don't need the speed, they are low volume or legacy products not worth the investment in upgrading. Government and enterprise and larger medium size businesses started migrating to single mode away from multimode between 2000 and 2005. At this stage multimode fibre was only good for 100mb-s only since 2005 have we seen gigabit multimode. Working in a geographically distributed LAN environment there would be times where we would put in a low end multimode switch on to a small remote office with less than 10 computers, however we would always install both multimode and single-mode to that office. Larger buildings that already had multimode we would pull single-mode only. The view being that when budget allowed all the switches would be upgraded and the entire network would be single-mode. The biggest drivers in the larger network space are Consolidation, Redundancy and Management. Consolidation Externally the cost of leased lines is formidable. So enterprise environments move to consolidate their networks to minimise their leased lines. Internally an enterprise site may have services like VOIP, access to process control systems from off-site vendors, onsite contractors networks CCTV the list goes on. These networks require fibre, very quickly using up spare fibres and or the room you have spare in your conduit. The answer is to moved to managed switches with VLANS, clustering or stacking the switches to increase the number of ports without requiring more fibre cable. Fibre itself is then trunked to optimise the bandwidth, so instead of having 2 x 48 port switches on their own separate fibre, the two switches are clustered and the they share the aggregate of the two fibre pairs which in turn provide redundancy through STP. This way if one switch is under more demand than the other, the load is shared across the fibre, reducing the amount of fibre you need. Redundancy Redundancy in an enterprise environment becomes a capacity issue, again consolidation is used to manage the fibre capacity. For redundancy you must have two fibre pairs running to each switch. Redundancy on separate switches is n x 2. On a group of stacked or clustered switches it is n + 1. So in the case of 4 switches separated they need 8 pair but consolidated they need 2 pair. Management Very similar to the notes for consolidation only to add the less points of management the better. So consolidation or clustering your switches to a single managed switch decreases the room for human error and makes a large network with many switches easier for the network administrator to get his head around. This all sounds well and good but when you add routers, wireless LANs, firewalls into the picture management becomes quite complex. A single switch with the wrong configuration can cause problems across the whole network, I have seen a laptop plugged into a port only configured for phones bring down the entire phone system. At the moment I am sitting at a site with a several clusters of 10 or more switches. All trunked back the head office into a firewall, everything is virtual, LANs, firewall ports, routers, it all sits in clouds underneath the physical layer. It has taken 3 months to find and solve an asymmetrical routing issue between the virtual firewall and the wireless LAN. So there is some pain involved with the enterprise level of network management not to mention cost. The other option, which is feasible in smaller networks, is to keep the networks physically separate, which means separate fibre, as services become available to smaller businesses you will need to provide managed converged networks or physically separated networks. This is plausible because of the cost and difficulty of network management which only becomes efficient at a large size. So as an CCTV integrator you must be sensitive to the client’s needs. If you put in multimode and a network engineer comes through in 6 months time and tell your customer they need to re-pull single-mode fibre then you’re not going to be too popular. Sherrifa, I hope this helps you understand why some people have recommended single mode while others have recommended multi-mode The day is coming when 1gb will not be enough for your 64 x 100megapix cameras, the wireless LAN you have installed in your warehouse for the touch-books for your stock pickers, the IP based cordless phone systems or possible the mobile substation that allows your employees to use the local PABX when in range so there are no charges for their mobile phones when making internal calls and only land line charges when making external calls, the stock tag swiping system than is now entirely IP based, along with your wireless bridge to the warehouse you bought across the road as your business expanded. So before you get greased up you may want to consider your fibre count. Your decision is one of risk vs return, multimode will get faster as Gbic technology improves, network management will become more out of the box and consolidating networks will become more feasible to smaller businesses and on lower cost hardware and single mode technology will come down in price. On the other hand the services you are going to deploy tomorrow will take you by surprise. So the question is not so much what you need today but what you need tomorrow and the net present value of your investment and the risk of under investing. My advice is, if you need to put fibre in the ground, investigate the cost of composite cable (single + multi) and plenty of it, include the cost of the pulling and be mindful of the capacity of the conduit. If you are going direct bury just put in lots. If the cost of pulling fibre is negligible, then you can go smaller fibre count multimode. If you’re installing in a larger environment then consider single-mode hardware that is management compatible with the existing hardware. If the environment you’re installing in is larger but has a ailing network then make sure you’re aware of the customers network plans for the future. Hope that helps. (Sorry about the essay but I am using parts of it in a project submission)

Seriously, I outsourced the first trailer I had built it had an 8m mast with two Pelco Esprit PTZ's on a crossbar. The mast was not telescopic with a hinge on the back of the trailer 2m above the ground. The whole thing looked really good, however the cameras needed to be removed to tow the trailer so the PTZ’s wouldn't get damaged. 30kg of cameras and PTZ, 8m mast 1.5m to counterweight, do the math. 30kg / 1.5 x 6.5 = 130kg counterweight around 3.5m behind the axle. So if someone took the trailer off the tow vehicle before fitting the cameras they would have to weigh in at 100kg to keep the draw bar down. Back one up to the crest of an open pit, disconnect from the Landcruiser and watch the thing throw itself down a 400m deep pit. We had a procedure to stop this from occurring and it never did happen, but it was always on my mind that someone wouldn’t read the procedure. So I eliminated the problem altogether and built new trailers taking control of the design. No big issues as it had many other issues that I fixed in the next 3 versions of trailers. The other problems were mainly power capacity and safety related.

I have already done this but from a different angle. Don't know if it would work for you, but will give you a few things to think about. Possible Major differences. 1 Heat. We work between 0 and 55c 2 Remote power only (this really sticks up the price) 3 We always had connectivity back to servers (wireless 5km or less) 4 Always looking at mining equipment 400m~2km away 5 Security not an issue Possible similarities 1 Large masts are problematic. Mast movement when looking distances. (note your IR triggers may have issues) 2 Counter weighting masts can make them dangerous, don't build a catapult. 3 Electrical regulations with earthing etc. Issues A DVR distributed NVR wouldn't work for us due to the heat, so we used Centralised NVR only. However you probably don't have the bandwidth back to your servers. Power. If you need remote power get back to me. Gensets suck Solar is way better if you have sun but its expensive and weighs a lot, (batteries) Long distance cameras. If you use any long distance cameras then your PTZ housings are really inefficient from a power consumption perspective. So if you’re concerned about any of these things drop me a line.

Multimode is going no where! We put it in every day. Government projects where money is no object are migrating to 50 micron and 50 micron laser optimized but short runs will be multimedia for a long time to come. As a general rule Multimode is used from MDF to IDF closets and short run building to building. Good points, I did put in a group switch for a PABX back in 2004 that was legacy multimode only at the time and I guess there is little reason for that to change, and I guess as individual devices may be multi mode, like fibre channel disk arrays and I have come across some radio units that run fibre back to their indoor units. My comments did say run of the mill IP devices, like switches etc. I guess there are some routers that will come with multimode only fibre cards. I did deal with several low end switches that supported multimode only, but I slowly replaced them with Cisco IP3000's as a standard and moved totaly single mode. I always had legacy runs of multimode around but all my new runs were single only. The Cisco IP3000 are industrial switches that are managable with the rest of the corportate network so I deployed Vlan, Spanning tree, Trunking and IGMP Protocols over them. I wouldn't get that level of (centralised) management from a cheap switch with inbuilt multimode. Also most of my runs were between 500 and 10km so multimode would have a capacity cost. Just another point is I ran an NVR site only so no analogue over fibre. So I guess if you have legacy non IP gear around, PABX, Serial over fibre, or DCS systems then make sure you put in multimode + singlemode. If you aren't consolidating your networks like I did, then cheap switches with multimode may be the best solution. ssmith10pn, I am an end customer who does all my own intergration and am about to move jobs and consult on another installation. CCTV is only a very small part of the system, the rest is monitoring and distributed control systems in a mine. Can you give me any specifics where you have had to use multimode for IP video or any managable low end switches that have worked for you, I am always up for new ideas or changes in direction.

Guys, Multimode went by the wayside years ago. I haven't installed multimode since 2007 and that's because I had a RAD serial to fibre multiplexor to support. As far as normal run of the mill IP gear goes no-one looks seriously at multimode. Single mode and multi mode cost around the same, Cost of burying fibre is usually the main expence. The ends supports can be more expensive but not as expensive as having to change from multi to single when no one supports it any more. There was a world wide shortage of single mode around 2005 that was because everyone was replacing their multi mode.

Not a dumb question at all. The reason why we look at high end cameras and lenses is because of the cost of infrastructure. Because we work in a dynamic environment where we build mountains and valley in a day, we either have to place our cameras miles away from the action or go mobile. These are around $145k and will support in excess of 12 cameras. But we run many other comms systems off them as well. When we started we were paying 70k for a trailer that could only support 2 cameras. So we sized up and threw in as many synergies as possible and came up with these beasts. 5.5kw solar. 1.7 tonnes of battery. 15m mast with 45kg headweight. We have three of these some with cameras on them and others we mount cameras on frames right on the side of the pit. After taking out the cost of the other services powering a camera is down to under $8k each. So in many cases its worth throwing the money at the better camera and lens unless your willing to fork out some big bucks for trailers of other infrastructure.

Has anyone looked into or used the Cisco video equipment?

That's a nice peice of equipment, the EH5729 might give it a go. but its only 150mm high and about 750mm long. But you should be able to get a special order from Pelco for a longer one. Just need to secure it so it doesn't get ripped off. What camera do you recommend for that type of equipment? I am not particularly impressed with most CS mounts. They are fairly flimsy when you consider you are hanging a couple of $k off the end of them and are sitting them in a several $k of PTZ and power system. At the end of the day the camera is the cheepest component of the excersize and starting to be the weak link in the system.

If anyone is interested in serious solution you can talk to http://www.orionintegration.com.au. We are only an end use but are our own systems engineering department controlling around $10m of bleeding edge control systems on $300m worth of heavy machinery, so we make up anything we want or need. CCTV is not our leading system its just a bit of icing on the cake and doesn't have the big returns of our other systems. However being in an extreme environment we have had to manufacture our own stuff to come with heat, power and dust solutions for any of our systems. Recently we started talking to Orion Intergration and they had done many things we are already doing or want to do. So now we are outsourcing most of our CCTV mods to them. They are dealing with some big lenses, long distances and extreme temperature stuff. If you start heading down the modifications area talk to them and they could save you alot of heartache. Just got off the phone to them about this topic and here is some extra info. 1: Stick a 240mm lens in a Pelco Esprit with an IP camera in a hot environment and the camera will crap itself. You need to add a cooling system, this will cost about 5watt. This is because the 240mm lens reduces the air volume in the camera. Computer memory is about the first thing to fail electrically. Analogue cameras can take it IP's can't, but why do you want to stick a $1500 lens on a 640 x 480 camera. 2: If you are going to big lenses look at Cannon SLR lenses. They are much better than standard CCTV grade stuff. 3: If you go for a bigger housing on an esprit the stepper motor will stop working if it senses an overload. I mentioned that I was worried about overheating in the previous post They said its not an issue just the thing will stop working. They make modifications to overcome this 4: Pelco will sell an Esprit with a longer housing as a special order. But the other part numbers I mentioned earlier may be the go they are from the old legacy systems. They are substancially wider making for more air space. However more windage as well. 5: Orion do make extensive modifications like turning Esprit to IP and power mods to reduce power consumption (a must for solar installations) You can spend up to $20k AUD of one Esprit PTZ from them. Before you all turn away and think that is rediculous my first attempt at Solar power mobile trailer based cameras cost $36k per camera for power and remote connectivity. After investing $600k on 4 massive 5 ton trailers I now have that cost down to $8k per camera though the synergies of many other systems on the trailer and the sheer size of the trailer. We are about to build a further 6 smaller trailers for $400k including 12 cameras. So turning a 2 camera trailer into a 4 camera trailer for the cost of $20k per PTZ can be worth it. Or if you put a single camera in a remote location the cost of power can be enourmous. Reducing power consumption it the best way to reduce the infrustucture cost. These guys have done all the hard work, just pass the bill on to your customer and get it right the first time. 6: Anyone interested in the Esprit IP conversion. We were doing this ourselves, Pelco don't do it because they have to change the slip rings to meet RF compliency. If your willing to take the RF issues on yourself you can make the mods your self. (like there are lots of devices going to be affected by the RF from an unshield ethernet cable in a PTZ sitting 10m from any other device.) Anyway its more that just running a few extra cables. The RS485 cables have to bypass the baseboard which has filters in it. So you need to modify the filtering in the mainboard in the top of the PTZ. Again you can get Orion to do this for you.

The problem with most PTZ's housings is they don't support IP camera's. (housings that will fit a decent lens). We use a Seikou HZCG12240 240mm lens with a DVTel IP camera in an Esprit but we have to modify the PTZ to carry the IP signal up to the camera. And this combination just fits with some modifications to the curcuit board to make room in the housing. There is a Vendor in Australia who will do this for you, or Pelco are planning an IP version soon. For bigger lenses Pentax look the go but I have never used them. You can get bigger Pelco housings and put them on Legacy PTZ pedastools or even put a bigger housing on the Esprit but I stopped short of the latter option fo concerns of burning out the motors with the extra weight and windage. (The Esprit housings continuously use the motors to keep the housing still against the wind). Housings that may fit the Esprit are EH4718 (457mm) or the EH4722 (558mm) the Standard Esprit housing is only about 350mm long. I have not made this modification myself it is possible but the holes don't quite match up so its no walk in the park. And as I mentioned before stress on the motors may be an issue. I can't help you with the number plate issue as we look at trucks the size of the 3 story house up to 2.5 km away. And we have massive image distortion due to heat and dust. Typically we deal with 50 ~ 60 degree thermals comming out of a 500m deep pit. I am yet to deploy the megapixil camera and expect better results out of that but we have bench tested it and the modifications we make to the PTZ. This projects with 6 mobile camera towers has been approved and will be in opperation in the next few months. Then I might be able to help you wiht some number plate information. This picture is taken with dirty lenses down a very hot pit. The amount of hot exhaust being expelled by the vehicles alone warps the image. It is taken with a PAL cameras with a 240mm lens. Squashed into a Pelco Esprit. In the preview the image looks worse than real life but as you see you wouldn't read a car number plate unless it was cristal clear with no distortion. To put things in perspective a car doesn't come up to the track height of this digger.

Its a not too much of a problem. Every now and then you can face them up during a storm but we go 6 months without rain. So you just wear the dirt and clean them once ot twice a year. Now we tend to mount them on the hand rails of trailers with working platforms or on ground based frames so they are easy to service. We are yet to stick one on the telescopic masts and they were bought with that in mind as an option. We have three 15m air powered telescopic masts that can handle a head weight of 45kg. However we worked out that it is important to hold the camera as still as possible when looking long distances. Masts cost about $17k each. I guess the vehicles we are looking at are dirty in the first place so it hides it pretty well. All the stills were taken from the same camera and only on one is it obvious the PTZ cover has lots of dirt on it. Still it would be nice to clean them a bit more often

Hey AK357 did you have a camera in mind that would do the job better? Its got to support Multicast, be the right size, take a c/cs lense or a 240mm lense, efficient codec, fit in a PTZ or have a well supported PTZ protocol. I would have liked to have found a higher Pixel camera.

Good fishing up here too

hey woody, you arent worried about the radiation? The third arm comes in quite handy really.

I don't care too much about frames, I am not watching a movie. Most of the time when we play things back its at fast motion anyway. We use it for replaying safety incidences. Understandable that you would care if you were in a casio and you where watching for slight of hand. What we want is to be able to see the number on the truck. We currently run with a throttle of 1000KB/s which is about 5 frames per second. This give us 12 days of recording and keeps bandwith low. DVTel had a very good NVR security model that let me share out to many uses in different groups and group my cameras and control PTZ access to different groups. Its security model now isn't that crash hot, but since the system I put in 4 years ago has done well I will stick to it until I change the whole setup.

All my stuff is wireless. But that between infrustructure not directly to the camera. I use Cisco 1410. 5.8GHz 54mb/s bridges between trailers that hold my cameras. I can mount up to 20 cameras on a trailer (Power capacity limit). On the trailer I run Cisco IE 3000 (Industrial Ethernet switch) The link and switch support all the Vlan, STP, Multicating IGMP and QoS protoclols. As its all the same gear you have some guarentee that the protocols are compatible. I also run a Tropos mesh 2.4GHz + 5.4GHz which doesn't support Multicasting. this is for my high availability gear like slope radars. So design the network to pass data via either wireless system apart from my cameras which I VLan though the 1410's only. Tropos are close to supporting multicast over mesh so I expect to be testing that in the near future. I also run ELPro 2.4GHz for IP, Serial, Modbus and Digital IP as clients to Tropos. They are nice low powered devices and could support cameras once Tropos can multicast. I then run another separate 2.4 GHz mesh for my heavy vehicals. Every thing runns through the IE 3000's and connected to the 1410 backbone via vlans and trunking. All in all I have about 500Mb/s wireless data coming in and its pretty bullet proof. Ihave about 300 IP addresses on wireless.

Ikegami ICD-808P they are 4 years old. These use the HZCG 12240 lens. We pipe them into 7501EXT DVTel encoders at the source. We are moving to the DVTel DVT9580A camera. It uses the HZDC12240 direct drive version of the lense. It should be a lot better as its direct digital to IP. Removes the need for an encoder and is 1280 x 1024. We have bench tested the camera and lense, done a proof of concept on converting the Esprite to IP but are yet to field test the final package. I did ask my distributer to look for a better lense but he couldn't come back with anything that would fit in a housing. Since then I have come up with some bigger housings to fit the Esprite but the modification is pretty heavy. At the end of the day my image quality will always be hamstrung by distance. The further away the more atmospheric disturbance and the more dust. So moving the cameras closer to the action is the solution I have now solved with mobile comms trailers Link to photo's of comms trailers http://www.cctvforum.com/viewtopic.php?t=17803&highlight=

Thanks to zmxtech for the lens calc. According to this at 200ft a 4ft wide view would need a 246mm lens. The Axis 233d has a 119mm lens. I haven't found any all in one cameras that have a lens anything like 200mm. The HZDG 12240P lenses is about $2000 AUD. We just manage to fit one of these lenses into a Pelco Esprite but it is a tight fit. We make some physical mods to the lense and housing. And we are currently modifying an Esprite to support a megapixel IP camera. I will let you know how it goes. Mods to the Esprite include adding a Mil-Spec plug to the bottom to make them plug and play. Convert the power supply to make it solar friendly. Re-route the serial cables so the serial data comes from the camera. And replace the serial interface from the input to support IP to the camera including PoE. Here are some photo's unfortunately I couldn't find anything at 200m worth focusing on apart from piles of dirt so these photos are at least 500m away. Camera res is only 800x600, very dusty environment so all the trucks and equipment are covered in dust and are as dull as the photo's look. It is also very hot and deep so there are laminates of different air temperatures that bend the light coming to the cameras. So here are a bunch of near and far shots with a 240mm lens In the first two photo's the truck is 616m away, I have live feedback of GPS of all the equipment so I can accurately calculate the distance to the meter In these 3 photos the shove is 1364m away and 285m deep. The drills are 858m at 285m deep. The last shovel is only 510m away and 285m deep. The closeup is a bit blured as the capture snapped between two frames.

Hahahahahh, yep as I said synergies. We have 4 slope radars which are a wicked technology that scan the walls and set off alarms if they are going to come down. Typically we get about a months warning, so they set up critical monitoring, about 2 weeks to go they will bund off the area to catch rocks and divert traffic. 2 days to go they will close off that area. 2 hours they will stand on the edge and watch it go. 6 hours later they give the all clear, they cost about $1m each. Then we have Modular which is the Fleet Management system running off mesh at a cost of $4m. So its easy to justify $600k on comms trailers. Sticking cameras on them is just a bonus, but in saying that the site love the camera system. Unlike most security systems we get everyone using them, Trainers, Geo's, Geotechs, Dispatches, Managers and supervisors. We also use it to audit the Fleet management system and check the data comming from the vehicles matches the data collected by the system. Shovel dig rates, Clean up times, kick off times etc. Funnily enough we need those big trailers to supply enough power to make the cameras cost effective. My first attempt at a small trailer only had enough power for 2 cameras and was very inefficent. cost for the infrustucture was $25k per camera, not including the camera or NVR system, just comms and power. Now I have it down to around $5k per camera. plus got some wicked trailers out of them. As for why look at dirt, it cost around $300m per year to dig that hole. Save 1% and we pay for all my systems in 3 years. Our group save around 10% of mining costs.

For me the big difference between DVR and NVR is working in an Enterprise environment, NVR makes much more sence. I can leverage off existing infrustructure, I have no Layer1 network (medium, fibre, coax, RF, etc) dependence as all security and addressing is done virtually. My servers don't have any video dependent hardware in them and I can put all or parts of my NVR in a virtual server. If hard disk capacity, speed or size becomes an issue I can move video streams to another archiver without having to recable, while the database can sit on any other server instance and the licence server can be on an entirely different instance. Network wise it means I can beef up my network as I have synergies to increase the over all value. I can justify high end network hardware and if I decide to install dedicated switches I can join the two VLANs and Trunk the fibre for backhaul. There is so much flexibility available in an IP network. Analogue video medium is very limited. In the server room I don't have any coax at all, this is a big plus. Just fibre and all on one protocol. Full IP means that security and addressing is in the IP packet where as Analogue Video has layer 1 addressing in other words physical. If you want to move your DVR to another room then you need to relay your cable. Furthur more my server room is 1km from my closest camera, I just want to put in one run of fibre anywhere (I use Fibreflow so I don't need to pull fibre see below) Coax I need conduit and pits. I can create a redundant loop with fibre and the IP protocol will look after it. Again with analogue routing addressing and security must be done at layer 1 in other words physical connection. NVR is superior when connecting areas with wireless because it reduces RF management. All my cameras are on remote solar powered trailers so the Cameras come back to a switch on the trailer then are backhauled through wireless bridges, 2 independent mesh systems, to a point where they come back on fibre with a microwave redundant backhaul. While you could multiplex analogue and pass it over a radio link why would you bother when passing IP over wireless has instant synergies with other IP systems. NVR has much better distributed points of failure than DVR as the encoders are all separated and the server services can be distributed virtually. If a DVR card fails you must take down the server. Where as if you have issues with a switch you can run up a switch along side and move the links acros.s If you have to work on an analogue trunk link you must take down that link. With IP if your running redundant backbone you can service fibre medium seamlessly. The downside of NVR is you must have a solid holistic understanding of all IT functions to make the synergies work for you. And if you don't have high networking standards the synergies might not be there **Fibreflow looks like very thick fibre but it has 3 to 50 hollow tubes in it each about the size of drinking straws. You direct bury it then you blow fibre through it. The advantage is you can go 1km without pits. Much better than laying conduit with pits ever 50m. It comes with upto about 50+ tubes. The stuff I use is 7 tubes and will take 12 core per tube. So you can blow more fibre through if you need it later. Small comms trailer and externa camera stand Large comms trailer. 5.5kw solar power over 4 redundant circuits, 4.6 tonne with 1.6tonne batteries, 15m air powered mast. Network 3 x fibre links, 2 x 5.8 Ghz, 3 x 2.4 Ghz, 2 mesh, 3 backhaul, Supports Cameras, 2 -Way radios, FM repeats, 3 meshes, 2 wireless bridges, 4 slope radars, Serial, Modbus, I Phones, Laptops, Digital IO, 120 heavy duty vehicles all over IP. Got 3 of these to move around when necessary. Cost about $200k each

Although some IP cameras will have poorer compression than others at the end of the day the encoder in a camera and the encoder in a DVR card both create a video file what ever that may be (AVI MPG or a proprietry file). This is where the compression is done to provide a digital file format for the hard drive not for the internet or network. An NVR encoder with a better codec will result in a better file than DVR with a poorer codec and vise versa.

Just because they are able to set up an IP camera, doesnt mean the video surveillance application is being done right. As to learning "IP", a 5 year old can do that these days. So what I am trying to work out at the moment is can you setup two Vlans on one switch that are connected together without a router then use IGMP policies to filter out specific multicast groups. This is an option to the easier task of filtering ports via a firewall or bridge. While I can make this work is it in the design specs of IGMP V3 and VLAN 802.1p/q Or is there a chance that if I deploy this it will fail if we introduce new switches at some stage in the future. Also how difficult will this become to manage and will Dynamic vlan port allocation become an issue. Has anyone got a list of switches that don't default to IGMP and treat multicasts as broadcasts by default? Does anyone know the answer to these?

On the comment of fear and control. In my experience dealing with the bureaucracy in large IT departments is much harder and takes more technical expertise than setting up the system yourself. Not having access to the interfaces and blindly trusting IT to get it right is not really affordable. You have to be capable of knowing what the underlying problem will be from the symptom with out the help of the access to the network interfaces. This presents a very high risk to any project you are likely to undertake. So while you can overcome the fear of technology there is another level of competence you must conquer to deliver these projects. I don't blame anyone from shying away from these larger contracts. Better clarify the larger contracts. Intergrating IP/NVR over the data network in an enterprise environment. As this is were there is a lot of scope for work if you have the skill and the imagination as discussed earlier. This is not suitable in casio's etc where separation or the data and video network are required.

[quote name="ak357 Thanks for info ! Can you tell us How Do You test Ethernet switches For Throughput ' date=' Bandwidth, etc Which tools do you use ? Thanks[/quote] Capacity planning for NVR / DVR is a mixed bag. Bandwidth from IP cameras is easy, it is throttled and can be controlled by and administrator. Computers are much harder because individuals with no capacity planning can demand a 20mb file from a file server or email. It is when you mix Video over a data network things become hazy. Capacity planning for switches is a bit more difficult. The backplane on a switch won't necessarily support the aggregate bandwidth of all the ports, particularly from cheap switches. DVR's are harder to capacity plan for than NVR's as NVR distributes the multicasting between the cameras instead of aggregating it at the server. The big questions is how to divide up the traffic among the switches and if your using trunks how to prioritize traffic. My major concern is safety systems and high cost heavy vehicle monitoring monitoring systems not the NVR. So my Vlans are based around calculable network traffic and random traffic that is client demand orientated is not allowed on those Vlans (which is protecting with wireless side capacity issues). The testing I am currently doing is based around spanning tree, failover, trunking and IGMP to see how the switches react when plugged into the wrong port or the effects of the multicasts and broadcasts between Vlans. I am not really testing for bandwidth issues as that is more a bottleneck at the wireless side. For monitoring bandwidth you can use a SNMP system like Cacti. I don't let random data generating system on my critical Vlans so I don't really need live monitoring. My wireless system can aggregates around 240mb-s from moving vehicles so my switches have no issues keeping up.

You both got it right, Soundy said that the cameras are easy to set up which is right, a HTML setup page is easier that dipswitches once you have found the camera. However as Survtech implies, managing switches is quite a difficult skill to acquire. The skill is not that difficult its just hard to get the skill. I currently have a couple of Cisco 2955 and 3000 industrial Ethernet switches I am bench testing for there behavioural anomalies in regards to VLAN, trunks and IGMP. You can't just read this stuff in a book or get certified. You must take the devices, set them up then break them to see what symptoms they will give you in the field when set-up wrong. I do a lot of reading and bench testing. My policy is just because it works doesn't mean you have it right. Then back to Soundy, there is a lot of new business out there if you are willing to put the time in and learn IP and NVR. Our company has just released IT GLD (Group level document) that every site must comply by. While these documents don't mention anything about Video surveillance they do not approve putting hardware into computers. It would be very difficult to integrate a DVR system into the network. Where as NVR is no problem, I saw this coming four years ago and that's why I didn't look past NVR. I guess a warning for those in integration land is the big accounts will go to IP/NVR in the future. It is easier for a IT team to learn how to set-up an IP camera than a CCTV integrator to learn IP, switching, SQL deployment, server and capacity management in an enterprise environment. The question is will the big global IT outsources like CSC and Unisys give the camera installation side to existing integrators or open up the market to non traditional CCTV competitors.