Help with a large NVR set up from scratch

[ProposalmanJ 0]

Hi,

 

First post, most likely of a string of amateur question. You guys seem extremely helpful and I am glad to have found this forum.

 

I am trying to design a system of 500+ PoE cameras for multiple campus residences. My basic understanding is that NVR will make it much easier to manage such a large number of cameras.

 

The primary requirements I have come up with for the indoor and outdoor camera short list are:

 

PoE

Internal storage (is SD or CF a factor?) - as a fail safe in the event of network issues

H.264 - mainly to save on bandwidth and storage

Motion detection with event triggering (is this dependent on the camera or the software or both?) - to switch from low fps to high fps, improved monitoring/response and searchability

 

Outdoor Cams:

Operating temp -30 to +40C (-22 to +104F)

Day/night capabilities (not sure if i need IR/ICR/low lux)

 

1. Are there any flaws in my reasoning or other things that need to be considered?

 

2. Could you please recommend some NVRs that would accommodate this system and briefly explain what the critical features are and why?

 

3. Is 1 NVR per building advised?

 

4. What would the best way to manage my network be?

 

5. Aside from storage, are there any other components that I will need for this set-up?

 

Functionality comes before budget.

 

Thank you

[thewireguys 3]

This is a huge system and if you do not have any experience with IP cameras I would not recommend going IP.

[tomcctv 187]

ProposalmanJ. hi please dont take this the wrong way. just on the questions you have asked says you have not done ip before. i would say not to do it yourself. just look at the money involved. best of getting a company in to do the work. that sort of money you want it right. or let people know were you are and if you have someone close to you from the forum see if they will help you manage the install of your system.

[thewireguys 3]

ProposalmanJ what is the budget for this system?

[destro_23 0]

Some assumptions we can make....

 

Can we assume the network infrastructure is already there? since it's some sort of campus? Also speak with your networking crew... it's basically adding 500 mini linux boxes to network that will stream video over the network.

 

this is one of those systems that if not done right.... with a correct monitoring center (think room with multiple large stations kinda like NCIS crime shows) there will be a lot of design involved just in that, let alone the network issues and design.

 

just some things to think about cuz this is a project that is on or over $500k and you don't want that a problematic system that was your design and baby. Again i'm one of those people that try and list the worst out first.

 

Systems Design is a MAJOR part of this... picking cameras and putting them up is the easier part. As a software engineer myself... largescale networking design is over my head.

[ProposalmanJ 0]

Yes, its obvious that I have no experience with IP cameras. My job is to research potential products for this system and offer comparison for the proposal. I wont be handling the implementation. So I am just trying to further my understanding in order to do my part correctly.

 

Thewireguys, Destro is correct, we are going IP because the network is already there for the most part. IS has assured that B/W will not be an issue. He was also in the ball park for the budget.

 

Would it be possible to get some answers for my low level questions? or at least some explanation as to why they are the wrong questions and what i should be concerned about?

 

Thanks for your patience, I'm sure that you can appreciate my need for help.

[thewireguys 3]

You need to build a new network for this setup. If they have spare (dark) fiber between buildings you could use this and it would save alot of money but do not run this system on there network.

[tomcctv 187]

Yes, its obvious that I have no experience with IP cameras. My job is to research potential products for this system and offer comparison for the proposal. I wont be handling the implementation. So I am just trying to further my understanding in order to do my part correctly.

 

Thewireguys, Destro is correct, we are going IP because the network is already there for the most part. IS has assured that B/W will not be an issue. He was also in the ball park for the budget.

 

Would it be possible to get some answers for my low level questions? or at least some explanation as to why they are the wrong questions and what i should be concerned about?

 

Thanks for your patience, I'm sure that you can appreciate my need for help.

 

 

 

Hi you are going to have to learn a lot on here before you can judge what sould and cant be used on your install.

which will take months.

 

if you are not doing the install (jusy over seeing) then you will be best getting quotes for your install and then posting the options on what equipment is to be used and them others can help.

 

are you in america ??? am in the uk and i know the regulations for cctv install in a campus/ school. in the US.

 

you say budget is 500k miles away for a 500 ip camera system you will need a secure network for the whole system. you will also need a control room and a good door accsess system to run along side the cctv. you are looking at a full central station set-up here

[Guest]

Yes, its obvious that I have no experience with IP cameras. My job is to research potential products for this system and offer comparison for the proposal. I wont be handling the implementation. So I am just trying to further my understanding in order to do my part correctly.

 

Thewireguys, Destro is correct, we are going IP because the network is already there for the most part. IS has assured that B/W will not be an issue. He was also in the ball park for the budget.

 

Would it be possible to get some answers for my low level questions? or at least some explanation as to why they are the wrong questions and what i should be concerned about?

 

Thanks for your patience, I'm sure that you can appreciate my need for help.

 

Sorry Proposalman,

 

Your are starting one third down the line of the design process stream.

 

Before you you even starting to think technology and specification, you need to identfy the operational requirement. You need to consider what the end user need to see live and record, what the regulatory and policy compliance requirements would be, who will use the outputs of the system (in this case the video recording and their requirements), the level of (technical) understanding of the operators and managers of the system, the environment in which each part of the system is to operate, future requirements during the operating life of the hardware, integration options (with associated systems such as access control, building alarms, vending, environmental control, PSIM systems, etc) and so on.

 

This process will then offer a guide to the next stage. Which, for example would likely to include a review of the existing infrastructure (phyisical and management, consideration of video frame rates, resolution, encoding techniques.

 

Don't forget issues such as network timing and very important network security compliance (ie have I designed a secure network and how do I evidence this).

 

Only following stage is when the technical design and specification process starts.

 

Also, do not forget to examine the likely long term viability (5 - 10 years) of the hardware manufacturer and software vendor.

 

Look for open protocol control/recording/display systems to maximise opportunities for equipment selection through the years.

 

Starting by considering the camera and recording technology will ensure that any system delivered will not meet the end users requirements.

 

Does this help?

 

Ilkie

[destro_23 0]

i agree with the wireguys...each cameara lets take a guess is about 1meg a second (wild gess and could be completely wrong) 500megs a second is a pretty big load on any school network. Imagine 500 computers(cameras) on the network streaming a live movie off the internet. Just to put it into perspective for the network guys.

 

wireguys is very knowledgeable so just going off his idea... each building is going to need it's own switches that are in turn connected to main system via fiber backbone?

 

wireguys this a good extremely general setup?

 

 

And as for the Proposalman i understand what you are looking for now... just some general info on every aspect so you dont go into a meeting without doing some research... good ip cameras go from $500-700 each for fixed ones and up to $3000 for PTZ(pan, tilt, zoom) megapix cameras.

 

30+ POE(power over ethernet) or if there is power 30 regular routers + install time per camera/building... A LOT of labor will be involved for the possible extra cabling.

 

Since this is a big project you may have to do it in stages... and test the results one building/area at a time.. so lets do one building and it's outer areas... then step back and review the results make changes to what you want.... analyize the network, then move on to next building/area when happy.

 

Get several proposals if you can. Compare and take your time!

[ProposalmanJ 0]

Tom, I am in Canada, the cameras are just to cover entrances and exists but a lot of the buildings are connected on all floors via tunnels or sky-walks. I meant 500k for the non-network or storage related equipment but that was just to give thewireguys an idea. What I stated in my first post about performance or functionality coming first still stands. Meeting the requirements is what matters.

 

Ilkie, The end user needs to be able to quickly identify who has entered/exited a specific area during a given interval. All such events are to be recorded and stored for a given amount of time. Campus police would be the end users. The system should be able to use facial recognition to check ban lists and take appropriate action. The systems maintenance will be the responsibility of the IS department. By environment, what factors need to be considered? temperature? traffic/congestion? Some cameras will be outdoors, Canadian temperatures range from -30C to +50C. outdoor lighting is not the greatest but artificial light is always present. Traffic, it is a busy university campus. The system should be integrated with an automated energy saving system that controls lights, HVAC, building access ect.

 

IS is also taking care of the network security, my only job is to present options for the components that can meet these requirements.

 

May you recommend some open protocol systems for this operation?

 

Thank you. Responding to your questions has given me a better understanding of this project.

[ProposalmanJ 0]

Destro, Wireguys, it will be on an independent network.

 

The cameras that I believe meet the requirements are ranging from $1000-1500 with virtual PTZ.

[destro_23 0]

also don't forget since we are doing doorways the cameras need to handle backlighting very well... "WDR" is keyword here... Wide Dynamic Range

 

Here is a sony video of a megapix camera with "WDR" function they call it View-DR

 

heavy backlighting on a doorway could lead to no face shot at all..

 

 

And as for software for the ip cameras... what are you thinking?

 

exacq

 

milestone (check out their customer story section) the copenheagan casino has 200+ cameras

 

I have never used either one of these but a phone call to either company will give you a LOT of answers.

[tomcctv 187]

Destro, Wireguys, it will be on an independent network.

 

The cameras that I believe meet the requirements are ranging from $1000-1500 with virtual PTZ.

 

 

 

 

with this being such a big job and the money involved you will have no problem having a manufacture also involved in your design. this will save you time and money.

 

i would look at Avigilon. they are the best at this type of equipment and its used world wide.

http://www.avigilon.com/products/samplefootage/

[destro_23 0]

+1 to what tom just said... these software companies have done this before and know what you will be needing storage wise and network wise. They should be able to assist greatly and give you better idea on the big picture.

[Guest]

+1 to what tom just said... these software companies have done this before and know what you will be needing storage wise and network wise. They should be able to assist greatly and give you better idea on the big picture.

 

Sorry, I do not agree.

 

Consider this, go to a software vendor and say this is project, and guess which software they will design into the project?

 

Go to a camera manufacturer and guess which camera will be suggested as being the best fit for the application?

 

From the brief description of the requirement I could make a good guess at which systems might be appropriate given 20 odd years exprience in large complex CCTV systems, but I would not offer an opinion until the full operational requirement was devloped. Otherwise the guess that would be made would almost certainly not be appropriate to the end user's needs.

 

My suggestion is that you need a good independent consultant to ensure you will not be embarassed.

 

Ilkie

[tomcctv 187]

+1 to what tom just said... these software companies have done this before and know what you will be needing storage wise and network wise. They should be able to assist greatly and give you better idea on the big picture.

 

Sorry, I do not agree.

 

Consider this, go to a software vendor and say this is project, and guess which software they will design into the project?

 

Go to a camera manufacturer and guess which camera will be suggested as being the best fit for the application?

 

From the brief description of the requirement I could make a good guess at which systems might be appropriate given 20 odd years exprience in large complex CCTV systems, but I would not offer an opinion until the full operational requirement was devloped. Otherwise the guess that would be made would almost certainly not be appropriate to the end user's needs.

 

My suggestion is that you need a good independent consultant to ensure you will not be embarassed.

 

Ilkie

 

 

 

 

 

sorry ilkie. i have known you a long time but sorry not a consultant. i was in london last week with a company who also brought in a consultant he was usless ..... and had never installed in his life. sorry i think a true consultant is a guy who has been on the tools most of his life and is getting to old to get up the ladder. someone who has been through everything not a guy who just reads spec sheets.

 

and this type of install is not down to the customers needs it down to the state regulations and the stricked codes.

 

from what i can remember from working in west virginia everything is linked together accsses control / cctv / face recognition / and body scan

 

this is why i said Avigilon. they have been the leaders of this type of equipment for some time now even in the uk they are helping companys win contracts at airports because they have a system that does everything

[destro_23 0]

I understand what your saying about the companies telling you their software is the best....but the whole point was to call multiple Software companies and gather info from all of them on and what they think may be needed design and networking wise.

At the end of the day the specs aren't going to be to far off from each other. You really think the developers of the IP software no know nothing about networking? I bet you a beer that if you call up and say i need a 100 camera system can you help me spec it out with what networking equipment i will need and such... you seriously think they will say Sorry we can't help you... no idea about this networking stuff? They will help you because they want to sell you their software. Even if it's an added fee they will help you... and if one company doesn't help then go to one that will. Again i have never called any IPcam company and asked these questions so i really don't know... but i would be in complete shock if an IP camera software company couldn't help in this regard. It's also free to ask questions, so never hurts.

[thewireguys 3]

I understand what your saying about the companies telling you their software is the best....but the whole point was to call multiple Software companies and gather info from all of them on and what they think may be needed design and networking wise.

At the end of the day the specs aren't going to be to far off from each other. You really think the developers of the IP software no know nothing about networking? I bet you a beer that if you call up and say i need a 100 camera system can you help me spec it out with what networking equipment i will need and such... you seriously think they will say Sorry we can't help you... no idea about this networking stuff? They will help you because they want to sell you their software. Even if it's an added fee they will help you... and if one company doesn't help then go to one that will. Again i have never called any IPcam company and asked these questions so i really don't know... but i would be in complete shock if an IP camera software company couldn't help in this regard. It's also free to ask questions, so never hurts.

 

Your correct they will help you spec out THEIR software not the best solution for your customer. Call Trendnet and ask them to spec the system

 

I am doing a 36 camera multi building system right now and all I can tell you is if you don't do all of your home work before you quote the job you will lose a lot of money. If I were you I would spend a week at this site looking at every little detail before you quote the system. This job your working on will be 1.5M+ easy for everything cabled and installed.

 

Also you have a big decision to make with your VMS software. You could go with Milestone or ONSSI but keep in mind your gonna have recuring fees for your customer at the tune of $50 per camera per year plus the server fees so your looking at over $250K over 10 years in software fees.

[destro_23 0]

i'm mostly responding to the origional posters "general questions" and giving ideas and tips that i think will help educate him... calling 3-4 IP companies and getting some FREE input from them is invaluable. 3-4 different tools that will get you the same end result you are looking at.

 

Even from the avigilon site it says their installers are also trained in network design and system design. Thats all i was trying to convey... is to talk to as many people as possible. Thats all no more no less... i'm talking "in general" terms... not literal.

 

If i go to build a house on my own for the first time... i call a few general contractors and ask them for an estimate and ask them questions... I ask all the 5 plumbers i called to give estimates questions and all the drywall people questions... but only 2 of them say you need the "green" drywall in the bathrooms because of the moisture... soo i just learned something on the way.

 

this is all i'm trying to get at. The OP is actually learning as he goes as am i here. Having 3-4 software vendors fighting for your business is a good thing... and you pick at their brains for info.

[Guest]

I understand what your saying about the companies telling you their software is the best....but the whole point was to call multiple Software companies and gather info from all of them on and what they think may be needed design and networking wise.

At the end of the day the specs aren't going to be to far off from each other. You really think the developers of the IP software no know nothing about networking? I bet you a beer that if you call up and say i need a 100 camera system can you help me spec it out with what networking equipment i will need and such... you seriously think they will say Sorry we can't help you... no idea about this networking stuff? They will help you because they want to sell you their software. Even if it's an added fee they will help you... and if one company doesn't help then go to one that will. Again i have never called any IPcam company and asked these questions so i really don't know... but i would be in complete shock if an IP camera software company couldn't help in this regard. It's also free to ask questions, so never hurts.

 

Sorry destro, in that case you will be shocked.

 

The OP has stated that this is a full integration project that includes access control, door entry, HVAC, environmental controls, face recognition and everything els. Most CCTV VMS systems would not be able to manage all of this. When the software comes from the camera manfufacturer they integrate with their camera systems (hopefully and maybe limted integration witha few others. If you purchase a vMS system from an open manufacturer say Milestone, the system will interface at various levels of integration with a vast rage of CCTV systems.

 

But after all VMS = video management software. Integration with other non CCTV systems would be very limited.

 

I would suggest tha the project would need PSIM type software to integrate all and could then include a Milestone type VMS solution in the next tier of command/control.

 

The skill s required to design networking a large amount of MP cameras in a WAN/MAN environment with a distributed architecture with other shared users on the same infrastructure would not be offed on a camera manufacturer training course.

 

In fact, the vast majority of in-house IT engineers would not be able to do this either.

 

Ilkie

[frank3 0]

I agree with others that you should investigate Avigilon. They have a unique and very scalable, flexible offering. For any system requiring outdoor surveillance their high-end 16 megapixel cameras are just fantastic, yet their economical 1080p H.264 cameras integrate seamlessly in the whole design.

 

They're also Canadian so I suppose their cameras are able to withstand extreme winter.

See their online case studies, they designed a lot of college campus systems : http://avigilon.com/products/documents/casestudies/

 

I'll try to answer some of your questions... As far as I know all their cameras have built-in motion detection to trigger recording. They don't have internal recording but their HDSM technology works great across wireless links or simply in a distributed architecture across many buildings so you can spread the NVR servers around. Their cameras are day/night and become monochrome (infrared-sensitive) at night. They offer rugged outdoor domes with built-in heaters for the 1-5 MP cameras and outdoor enclosures for the 11-16MP cameras. Third-party integration is quite decent too so you can add many other brands later on, including PTZ cameras where useful. It also integrates with a bunch of access control systems. Searchability is amazing with their pixel search feature, letting you go through days of footage in seconds. I don't know how it works, but it's just instant! Maybe thewireguys can comment on that?

[ProposalmanJ 0]

Thanks for all the responses guys, I am still researching the VMS options, but I am finding it difficult to compare them via data sheets as many options use different names for the various functions and I also don't even know which attributes I absolutely require. It seems as though after device compatibility and basic functions, it really just comes down user preference for interaction. Correct me if I am wrong.

 

Are there any advantages to standalone NVRs compared to the same enterprise software on a PC?

 

Also the milestone VMS says it has Built-In Motion Detection, does that mean that the software is not dependent on the camera having motion detection? that would reduce my camera costs significantly.

 

Thanks for helping me learn,

J

[destro_23 0]

even the more basic dvr systems have "motion detection" and it's all software based.. so the motion based recording is kinda just a standard anymore done via software. This isn't something you need to think about to much when picking cameras.. as it's the software that decides/programs this.