Simple question about IP cameras

[kao 0]

A couple of things that make me chuckle about IP systems...

 

People say you can use your existing network infrastructure so you don't have to run coax cable...

 

... ok - that's great, but how many places have network workstation drops at the far north-east corner on the roof? Looks like you're gonna be running cable ... even if it's not coax.

 

The "CC" part of "CCTV" is CLOSED CIRCUIT. IP cameras basically blow that whole concept away. Welcome to OPEN CIRCUIT ...where if someone is savvy enough, they can hack at your cameras all day.

 

I'm a software engineer, and in my day - I've seen PLENTY of network outages.... switches blow, routers croak, etc... Ever seen a coax cable "crash"?

 

We have a client that has about 60 analog cams and has been sold on the marketing hype of IP. He's slowly starting to move over - new cameras installed are all IP. .. so far, he's got about a dozen IP cams up and working. A couple months ago - they had a network outage that lasted for 3 days. Needless to say - he didn't get much video over that time period.

 

Like people have said - use the right tool for the job. Sometimes it's analog, sometimes it's IP ... but I agree with someone else's statement that IP has a looooooong way to go yet.

 

I pretty much agree with all that.

 

We install in a lot of schools. We started looking at IP cameras and software because the cabling costs of running coax through a building as large as a school (some are K-12) can get very expensive...in fact, sometimes 50% or more of the quote.

 

Throwing in IP cameras sounded like the way to go. However after gathering IP camera prices and sketching out installation plans, I find that installing an IP based system ends up costing more.

 

Like mentioned above...network drops are rarely located in convenient places for camera mounting. I thought going to wireless would solve that problem, but then you still need power for the camera...so now you are still running a cable for some distance...and usually at the price of an electrician doing the work and standard union rates.

The selection of cameras for IP is pretty limited. Sure you can use a video server to convert an analog to IP, but that just adds to the price of the camera $150.00 or more.

Network bandwidth is an issue. Even tho most schools have a fast network, it is being heavily used already...taxing it with video is bad.

To me, the proper way to do it would be to install a separate network...now your talking new cable runs, new switches and PoE devices, all of which = $$$.

Network outages are another issue. If the outage is near the camera, that's no big deal, but if the outage is near the recording server(s), then you are not recording anything

Backup power - we always design a surveillance system with the DVR and cameras running on battery backup. To battery backup a large network is difficult and expensive.

Security wise I dont think either system has much advantage over the other. If someone hacks into your network, they can see all your cameras on the DVR or they can see all the cameras on the NVR. Wireless IP cameras have an additional (minor) risk of someone using a laptop to hack into one.

 

The advantages I see for IP cameras are the wireless quality, decentralization (no head end required), and their sometimes easier installation.

At this point, there are a few installations where IP shines. One example would be - we have a client that owns 40 small offices around the country. He wants 1 camera at each location. Putting a DVR in each location would be cost prohibative. However, each location has broadband, so we proposed installing a wireless router in each office and 1 wireless IP camera. Back at his HQ we install a PC with IP camera recording software connecting to each office and recording.

We quoted it both ways DVR+camera at each location vs Wireless router + IP camera at each location. It was twice as much to put in even the smallest DVR.

 

I am guessing its the ratio of Analog/DVR to IP Camera solutions is about 98% analog/dvr to 2% IP.

As IP standardizes, compression and quality get better, software gets better, and camera selection increases, we'll probably see those numbers go to 50/50 over the next 10 years and continue that way until IP is clear dominant player.

My 2 cents.

[rory 0]

The advantages I see for IP cameras are the wireless quality, decentralization (no head end required), and their sometimes easier installation.

 

Yep i do see some advantages in this type of app .. but they are rare.

I see it going more hybrid, Analog & IP together (like VI) ... eventually.. so the DVR guys can target the IP cam market ... and vice versa ..

[murphomatic 0]

If someone hacks into your network, they can see all your cameras on the DVR or they can see all the cameras on the NVR. Wireless IP cameras have an additional (minor) risk of someone using a laptop to hack into one.

At least w/analog cameras - you have the option of not putting the DVR on the network. Granted you lose the remote-access features, but the head-end still records all the same w/no network connection. Try that with an NVR...

 

Anyway - your example of the 1 IP cam per site for your client w/40 sites is a PERFECT scenario where IP would be the beneficial route to go. I'll keep that kind of thing in mind.

[Thomas 0]

Um, you wire a seperate network, segrate via switches or use any other of a dozen methods. And the NVR is not on the same network as the users.

[murphomatic 0]

Um, you wire a seperate network, segrate via switches or use any other of a dozen methods. And the NVR is not on the same network as the users.

 

Generally, it's not the users I worry too much about (you're generally talking about people who think a "port-scan" is taking a nice stroll down the dessert-wine aisle at the local grocery store). It's IT/IS... Unless IT/IS is subordinate to, or part of the security organization, you have more people than necessary w/access to the surveillance system. It also adds a layer of bureaucracy -- as any decisions made regarding the system also impact the network, and now require the blessings of another dept. Very few organizations that I've worked within have been willing to set the Security dept. up with their own autonomous network that IT/IS plays no management role in. Getting the seperate network set up itself is not too big of a deal ... but getting IT/IS to turn loose of it is next to impossible.

 

Don't get me wrong - I don't think IP is bad or unvaluable - I just think the marketing hype out there is a bit overblown...and this causes people to make the wrong decisions on occasion.

[kao 0]

Yep i do see some advantages in this type of app .. but they are rare.

I see it going more hybrid, Analog & IP together (like VI) ... eventually.. so the DVR guys can target the IP cam market ... and vice versa ..

 

Yep. I couldn't agree more.

 

So...I wrote Geovision a while back and asked them when they were going to step-up to their competition (Avermedia) and offer the capability to integrate IP cameras into their system. They said its on the way.

 

Anyone know when that is going to happen?

[Thomas 0]

The marketing hype of IP side is rather bad, and alot of it borders on lies. But at the same time, the knee jerk reaction of the analog world to IP is just as bad.

 

These are tools people. You don't use a hammer when you should use a screw driver.

[murphomatic 0]

So...I wrote Geovision a while back and asked them when they were going to step-up to their competition (Avermedia) and offer the capability to integrate IP cameras into their system. They said its on the way.

 

Anyone know when that is going to happen?

I've been asking GV this same question for some time now ... and they keep telling me it's coming... V7 came.... then V8 .... and there's been no IP support yet!

[woodyads 0]

I have spent many years dealing with poorly supported IT systems and Video on IP is not on its own in this area. The problems are spored from many areas. Lack of IT savvy from managers with influence over the IT department is the biggest. If IT is subordinate to other management structure then it tends to be poorly run. Then the age old problem of the customer believing the vendor. Rule #1 never trust the vendor. Rule #2 where is the reference site. Rule #3 when talking to the reference site talk to the guy who supports it technically. Not the guy who placed the order and forced the system on everyone.

 

Because of the above reasons, I do agree with you that in most cases analogue is the better alternative. Other mines are keen to implement the system I have, but I put them off due to poor IT support. One thing I tell other mines is if you cant achieve synergy then you are not technologically ready for IP on your video.

 

What do I mean by synergy. We have a slope radar system that cost $3m per year to run. There are 3 trailer based radars that scan the pit walls and warn of any movement. The system saves lives and if it is down we can't dig. Cost of standing up a truck is $5000 per day multiplied by 50 trucks. The system is on a hideous 500Mhz radio modem link. So I put in a proposal to install 3 camera trailers, 2 cameras on each trailer, 54Mb/s links at 5.8GHz to each trailer, Auto failover and a 2.4GHz link between the Radars and the cameras, converting the Radars to IP. We also run two other slope monitoring devices on the 2.4Ghz channel 11 network. By doing this I have eliminated nineteen 500Mhz radio channels. and two 2.4GHz links that was competing with my Heavy Vehicle Real time system health system worth $4m. All the slope monitoring devices share channel 11 while the Heavy Vehicle system uses channels 1 and 6.

 

The big advantage with IP is centralisation of management and control as well as synergy with other systems. For instance I can now install SNMP services over all the devices including the cameras, trucks, shovels, drills, servers, switches, wireless bridges a total of about 160 devices. I can also use Spanning Tree Protocol and other types of redundancy to ensure availability. These types of management just aren't available to an analogue system. In the big picture IP is far better than analogue. I installed the DVTel system and I am impressed with the software engineering and the maturity of the system overall. It is much better put together than my $4m Vehicle Monitoring system (granted it must be backward compatible to a 15yo legacy database) or the $3m Slope radar system with crappy telemetry.

 

Bandwidth:

Here is a hint. Group your cameras by network port address. Use Layer 3 switches to control port flow. Block any video traffic travelling away from the desired viewing areas. Split the output traffic to the desired machines on the switch servicing the control room. Find all 10Mb and even 100Mb devices and eliminate them or port filter them from camera traffic. Cisco 1410 wireless bridges do this perfectly. I would imagine most layer 3 switches particularly Nortel and Cisco would have some very powerful filtering functions. We are a Cisco site here but in general Nortel have been the bomb in switch technology. Do look at the backplane bandwidth of your switches.

 

Redundancy:

Learn about Spanning Tree, Put in redundant links

 

Qos:

Quality of service. you can use this to ensure your video stream is given priority over other traffic.

 

Synergy:

Get the IT department on side and get them to sell redundancy, Gb to the edge, Voice on IP, SNMP, and other advances.

 

Serviceability:

Now your IT department has a legitimate reason to hold the spare they have always wanted but been deigned.

 

Availability:

Now your IT department has an argument for putting in manageable switches, decent UPS's and backup Gensets.

 

Servers:

Your IT department should build and support the server. The corporate IT department should supply the following services for your server. SNMP to monitor the server send out auto notifications of alerts. Constant service pack updates and virus updates on a daily basis. These should be tested in a lab not on your server. The correct build with Logs and Database on separate physical drives, (I see this one stuffed up all the time). and Domain integration. To supply all this the server support team most would need at least 5 dedicated staff. The server should be a stock standard build. It they aren't already supplying these services for your other servers then they are probably to immature to handle the networking or support the IP system. Most CCTV integrators are incapable of offering this level of support. If this is the case, and you have non corporate build servers, then they should not be on the corporate network. As is the case with many of our 3rd party build servers, they are kept on the other side of a firewall. We then standardize their databases and move the database to corporate build database servers while the collection and telemetry services stay on the other side of the firewall.

 

I cannot fault the system I have put in and I am sure many IP vendors have systems that are just as good. I just don't think many of us can supply the level or service or understanding required to intergrate such a system into our corporate network. In most cases it is the customers fault for not understanding and manageing IT and its function.

 

After installing a full IP system to replace an analogue system I don't think its a case of IP is not ready for us. Its more a case that we are not ready for IP.

[rory 0]

The marketing hype of IP side is rather bad, and alot of it borders on lies. But at the same time, the knee jerk reaction of the analog world to IP is just as bad.

 

These are tools people. You don't use a hammer when you should use a screw driver.

 

Hmmmm ... with that said, where does it stop? How about cleaning the clients pool, fixing the water pipe, trimming the bush .. ?

[Thomas 0]

Yes Rory, providing IP based cameras is a completely differant industry. Change out IP cameras in that analogy with Cat5 and Baluns and see if you still feel that way.

[rory 0]

cat5 and baluns are still closed circuit .. its not a network. .. its still CCTV .. though baluns cheap and I wouldnt recommend them on my worst enemy ..

 

Id like to ponder a question, how many people in the IP industry have actually been in the Security industry, or even the CCTV industry, before they started doing IP video systems ... from what i have seen most of the manufacturers dont know much about CCTV let alone the security industry, and alot of the users are mostly IT guys ..?

[Doug 0]

Every tool is a hammer, except a screwdriver and thats a chisel

 

Doug

 

These are tools people. You don't use a hammer when you should use a screw driver.

[kensplace 0]

I think most of the people in many IP related jobs know very little about the sector they target, most of IT is marketing these days. Does not matter how well something works, as long as they can tick that box to say it has the feature......

 

I would be suprised to be honest, if many people who make analogue cctv equipment actually had any real experience of using security equipment under fire, as most things I have tried (tecton multiplexer, bbv matrix+telemetry controller,honeywell DVR (new one just got it), bewator telemetry and various bits and bobs all seem to have annoyances that just would not have been missed at the design stage if they had a clue what they were selling.

 

The cctv industry is surrounded by hype and marketing , more so in the IP world, but the analogue world is full of it too, with incorrect lux ratings, lighting claims, resolution claims etc.

[rory 0]

The cctv industry is surrounded by hype and marketing , more so in the IP world, but the analogue world is full of it too, with incorrect lux ratings, lighting claims, resolution claims etc.

 

Alot if not most of the incorrect specs though are in the OEM sector.

But yeah there are definately more sales guys than actual techs ..

[kao 0]

So...I wrote Geovision a while back and asked them when they were going to step-up to their competition (Avermedia) and offer the capability to integrate IP cameras into their system. They said its on the way.

 

Anyone know when that is going to happen?

I've been asking GV this same question for some time now ... and they keep telling me it's coming... V7 came.... then V8 .... and there's been no IP support yet!

 

Don't know if the belongs here or in the Geovision post....I'll let the mods sort it out

 

The link below is to a product announcement showing the Geovision video server.

I believe they are demoing it on their current tour. They will be in Chicago this Thursday, but unfortunately, I won't be able to make it.

Has or Is anyone going to see this product? Feedback would be great!

 

http://www.taipeitradeshows.com.tw/NewProduct/Product.aspx?ID=1632&Lang=E

[murphomatic 0]

So...I wrote Geovision a while back and asked them when they were going to step-up to their competition (Avermedia) and offer the capability to integrate IP cameras into their system. They said its on the way.

 

Anyone know when that is going to happen?

I've been asking GV this same question for some time now ... and they keep telling me it's coming... V7 came.... then V8 .... and there's been no IP support yet!

 

Don't know if the belongs here or in the Geovision post....I'll let the mods sort it out

 

The link below is to a product announcement showing the Geovision video server.

I believe they are demoing it on their current tour. They will be in Chicago this Thursday, but unfortunately, I won't be able to make it.

Has or Is anyone going to see this product? Feedback would be great!

 

http://www.taipeitradeshows.com.tw/NewProduct/Product.aspx?ID=1632&Lang=E

 

Hmm... that looks like it'll take analog cameras and put them on an IP address like any IP video server currently on the market.

 

What I'd like to find out is if the GV-Main System they're showing in that diagram is capable of picking up an IP address ... or if they're just tying to the main system using the analog outputs on that box.

 

GV really needs to set up their main system so that each camera can either take a data stream from their gv capture card or a data stream from an IP address. When that happens - the main gv system will truly become a hybrid system.

[rory 0]

guys i imagine this is the hardware here: NVE Series

http://www.udptech.co.kr/products/products02_nve4000.php

[kao 0]

It looks to me like you would plug the analog camera into the mpeg encoder, the encoder would feed out an mpeg4 stream to the LAN/Internet and the Geovision server would recognize the mpeg4 stream. Its possible they could be using a decoder at the geovision end to convert the stream back to analog, but I suspect they will just read the stream with their software.

 

In my opinion, this is the smartest way for Geo to get in the IP game. Create their own box/encoder (or sub it out as it looks thanks to Rory's post) and engineer the software to work with it.

 

I like this way of doing it as opposed to trying to find compatible IP cameras. It would be different if there were some kind of standard or even a few different standards for IP software developers to deal with, but it seems just about every company that produces an IP camera does something a little different with the stream. I don't know how companies like Milestone keep up with all different IP cameras.

With analog it's easy...its just plug and play - you dont worry about software compatiblity or whether or not you should do a flash update, etc...

So if I am right about how Geo is doing it, it still allows the installer to choose any analog camera to best fit the job, plug it into an mpeg encoder and stream it to their server - anywhere in the world.

 

Rory - you got an inside guy at Geovision or something?

 

I have looked into Acti and IndigoVision for their encoder/decoders. A pair of Actis is running somewhere around $1.000.00 us. I hope the Geo encoder is less than $500.00 and offers a high quality solution like that of Indigovision.

[murphomatic 0]

I guess the problem their kao is that more and more of my clients are refusing the buy analog cameras, and are insisting on going IP. This little box doesn't help much when it comes to tying IP cameras directly to the GV main system.

[kelvin 0]

analog camera and ip cam both got their own pro and con. it only depend on your application which is more suitable for you.

[kao 0]

I guess the problem their kao is that more and more of my clients are refusing the buy analog cameras, and are insisting on going IP. This little box doesn't help much when it comes to tying IP cameras directly to the GV main system.

 

Yeah...I understand your pain. I am getting more customers asking for IP as well. I really try to educate them on the pros and cons for both systems so they can make an informed decision.

Most of the time this works fairly well. However there are some people that go against my recommendations and go with an IP solution when an analog one would be best - just because its the latest buzzword.

 

In the end you have to give the customer what they want.

 

When a customer tells me they want an IP solution...and I feel that an analog would be a better fit. I ask them specific questions about what they want and expect from an IP system. 9 times out of 10 I can show them how the analog will meet or beat thier needs.

If they want something very specific - like logging directly into the camera...or placing single cameras over large geographic areas, then an IP solution is the obvious answer. If Geo offers this mpeg4 box, then I can give the customer the IP capabilities they specifically ask for.

 

The one advantage I think IP will always have over any analog system (and someone correct me if I am wrong plz) is that you can get such high resolution pictures for certain situations. Some of the cameras I've looked at are up to 11 megapixel. The pictures they produce are incredible - the detail is unmatched.

[rory 0]

Kao .. G e o is a software company, U D P makes their cards ..

[rory 0]

The one advantage I think IP will always have over any analog system (and someone correct me if I am wrong plz) is that you can get such high resolution pictures for certain situations. Some of the cameras I've looked at are up to 11 megapixel. The pictures they produce are incredible - the detail is unmatched.

 

Thats only true for some of them .. and depends on the software and setup, most I have demoed have been low res. There is a thread on Motobix test images though, that one looks nice. He does however give you examples of the frames and image sizes. For most CCTV jobs though that kind of product wont be applicable or feasable.

 

Back to Morph's question on Geo, you would simply plug your IP cameras into their decoder which will then plug into the Geo system.

[murphomatic 0]

The one advantage I think IP will always have over any analog system (and someone correct me if I am wrong plz) is that you can get such high resolution pictures for certain situations. Some of the cameras I've looked at are up to 11 megapixel. The pictures they produce are incredible - the detail is unmatched.

 

Thats only true for some of them .. and depends on the software and setup, most I have demoed have been low res. There is a thread on Motobix test images though, that one looks nice. He does however give you examples of the frames and image sizes. For most CCTV jobs though that kind of product wont be applicable or feasable.

 

Back to Morph's question on Geo, you would simply plug your IP cameras into their decoder which will then plug into the Geo system.

So just to make sure I'm understanding this correctly ...

 

With the box GV is introducing - you could set it up so it takes input from an IP camera and streams that video to an analog output, that could then be plugged into the gv-capture card? So it goes IP -> Analog? Most of the products out there will do analog -> IP, but not the other way around.