What Should I Consider When Moving From Analog To IP?

Roger, thanks for your first comment!

There's lots of factors, a few to start:

(1) Who is the existing manufacturer of your recorders? Do you prefer to keep that manufacturer or switch to someone else? Related, do you want/need to watch all the cameras on 1 user interface or can the 'old' ones be on the 'old' system and the news ones on the new?

(2) How big is the facility? How hard is it to pull new cable? This could impact whether you go for HD analog or IP with baluns or new cat cable, etc.

(3) Do you have existing PTZs? Do you plan to keep PTZs? I ask since you mention multiplexers and historically multiplexers are often used with PTZs.

Some common approaches:

• If you want to keep all the analog cameras but add a VMS, put in racks of encoders and 1 server with the VMS software.
• If you want to keep the cabling, swap HD analog cameras for your existing ones, replace old DVRs with HD analog DVRs (which can connect record a mix of old and HD analog cameras).
• If you want IP cameras, it depends on whether you have coax or cat cable in place. If coax cable, either replace with cat cable or add EoC adapters. Then trash the recorders, add a VMS server.

Roger,

Unless you have a immediate need to replace the system quickly I would take my time and use this opportunity to plan.  I would consider the strategic plan first then get to the technical/logistical plan.    

Some strategic plan questions to consider.

What are the common characteristics of a video surveillance system for your type of environment?  (Are you a school or campus, private corporation or public facility, manufacturing, office etc)    

What do you want the new video surveillance system to accomplish?  (mitigate risk, loss prevention, perimeter security etc) 

Who uses the system now and will other departments need it in the future?  Who are the major and minor stakeholders in the project?  

Which budget(s) will pay for it?

After writing the strategic plan I would put a budget number on it and decide if this is a phased project over a few years or one time capital expense.  Either one of these approaches works well however the technical/logistical plan is dependent on it.  

Roger - Will Dohertys advice is spot on.  Get as much feedback  as you can from the people who will be using the system. Then apply Johns technical advice when you take this out for bids.  Every minute spent on proper planning will pay for itself many times over.

there's a great article in Security Sales & Integration Magazine - just read it over the weekend. https://www.securitysales.com/surveillance/10-steps-avoid-ip-video-issues/

One tricky thing to be sure to ask the customer is how they will want to view the camera feeds.   With analog, the signal gets amplified and multiplexed, and in general you create many copies of the original signal and the 'viewer' simply shows them.

In IP space, the Client viewer application will need horsepower to view the streams and there will be limits on what the hardware can support.  To make it even more difficult is the fact the each VMS will have its own way for doing this and thus a system that can support  XX streams of RR resolution at FF frames per second will support YY streams on a different VMS.

Consider using a VMS that has "no license fee" per channel therefore freeing up a considerable amount of cash for future IP Cameras as many of the top suppliers have wonderful VMS applications without the financial burden of licensing.Compare features and decide what is really 'needed' and what is considered 'bells and whistles'.

Just saying......

I haven't spent much time in either side of an orange jumpsuit, but these hotels have some unique issues.

Typically these resorts are built in pods and each pod has a special viewing room with multiplexed views of the appropriate cameras to make sure the guests get plenty of attention.  Cameras are often run to a matrix which I don't see listed.

Many times these systems are interfaced with a PLC system that manages the guest entry points.

Usually there is a retention requirement that's a little longer than the average hotel so multiplexers and tapes were used or very low resolution dvr's to attain that retention.  These guests have access to lawyers and law enforcement on a regular basis.

While cabling may seem easy, the last mile per se can be riddled with time wasting guest management.

John Honovich lays this out pretty well. You could literally overhaul your system for a fraction of the cost of going IP if you go the route of HD Analog. 5MP vandal cameras are incredibly inexpensive. You can use a HD DVR and that basically becomes your 16-ch encoding device and use a high quality VMS & storage, or fill the DVRs up with HDDs and use an interface to tie the DVRs together. Almost any major brand has this option. 

Being a prison, there is really no need to go IP, especially if you already have the analog infrastructure in place. And by the way, I just saw a couple factories release DVRs that push power/video/Data over the coax directly from the BNC connections with built in power supplies on recorder.

You won't be allowing any video access over the WAN. Save a lot of money and/or greatly increase profit margin. I'm a manufacturer of PoE switches and the Ethernet/power-over-coax converters. I really don't know why this solution isn't used more.

Don't get me wrong, I love selling switches and converters, but in your shoes, I'd go the easy route. Oh, and not having to deal with the difficulty of networking in general...Big time win if you are not an IT expert. Good luck! 

Typically camera call-ups are being driven by the PLC system in jails/prisons. Often the HMI for facility control can call up cameras. Door opening or breach would call up that camera automatically. It's likely driven by dry contact type relays, I/O. You need to investigate how this was configured and cabled. You may end up requiring some PLC programming outside the realm of a typical security intergrator's skillset. If you go to a typical IP server/NVR system, you will need to consider how you are going to integrate all those I/Os.

Roger I recently completed a surveillance design for the US government in which they initially wanted to use analog cameras at lower resolutions and I convinced them (their proxy actually) to allow me to use IP based cameras. My reasons were as follows:

There is not a one to one correspondence in using analog cameras and high resolution cameras. To prove my point I recommend you use the IPVM calculator to develop a surveillance model for each approach. Its one thing to sit and pontificate about the merits of high resolution but something entirely to put a surveillance mode using each of these technologies in front of a client (customer) and show them the difference using REAL devices. Its an awesome tool. Its easy to see significant differences in power distribution and infrastructure requirements. Yeah... I'm a big fan.... So more expensive cameras but lots fewer of them.

Second, while high resolution cameras are more expensive, they come packed with more analytics like virtual fencing and slew to cue capabilities with an intrusion detection system. Each individual camera has more inherent capability that should be utilized. Pushing processing to the edge of networks is a big trend that will certainly continue, this edge computing will reduce the communications bandwidth requirements and processing requirements on the DVR / surveillance server.

For Indoor applications (I cannot yet recommend this or outdoor, pole mounted types of applications personally) IP cameras using POE or POE+ is a big deal in savings on retrofit costs. For outdoor applications I am still old school using MMFO OM3 or better both for EMI and distance purposes. 

In a application like the prison you mentioned, the best solution will probably be to go with a hybrid approach using IP where it makes sense. Large viewing distances at high resolution enabling both digital and optical zoom is a great feature. In the no mans land between the inner and outer perimeter fences, I cannot imagine a better intruder detection solution than using virtual fencing or motion detection analytics at the camera and a buried volumetric presence detection system.

With technology today i think it is better to think of security as a single, comprehensive, integrated solution of surveillance, access control and intrusion detection in addition to operational procedures. 

PLC's are great tools and I would look for ways to integrate that functionality into a common user interface for operators. Applications like Tridium Niagara are able to communicate with many disparate HMI and SCADA applications to create a single user environment for all those systems.

Best wishes on your project.

Tim