HDCCTV Vs IP Video, Large Legacy Analog System With Long Runs, Conversion

I almost dread bringing this up again after reading the last one at migrating an analog system to IP, but need to, because that situation was a little different. This is a larger system with different considerations, IE, a small amount of the cable could be re-pulled. Or maybe that system had the same considerations, but weren't discussed.

This system is in a large distribution center. Perhaps 10 of the runs are 100 or so feet in the office area. However, the other 179 runs are of varying distances through the wrought iron of a large distribution center. The desire is to 'upgrade.' They think they want IP video. I think there's room for cost-benefit analysis though of IP vs HDCCTV.

For the short office runs, if IP, I could just re-pull the cable, cheaper than converting for short runs. For all the warehouse runs, it'll have to be a conversion for IP. For HD, I just re-use what's there.

Probably next week, we will replace two analog cameras that are down right now with SOMETHING, and provide a recorder for that something, then we'll do probably half the rest at the beginning of the year, then the leftovers by the end of the year. I want to make sure I've really considered things before making my recommendations for what we do next week. I know a lot changes fast with technology.

My perceived pros for IP:

  • Resolution
  • Interoperability with different systems
  • More feature-rich recorders
  • IT people know how to interact with it
  • IT security, mac address locks for cameras, etc
  • Lower (general) head end cost than HD (they seem pretty expensive to me, and only 8 channels?)
  • More than 8 channels, less rack space

Perceived pros for HDCCTV:

  • Less latency
  • Don't have to change existing equipment (maybe power supplies, but no cable changes)
  • Don't have to pay for converters, seems like the going cost now is about $300 per cable
  • Better live viewing (no compression)
  • No switches
  • Less labor

What do you all think about the two in this specific situation? Unless there are technical issues to consider that I haven't mentioned, I think the most cost effective solution will win out, if the cost difference is significant. If it's not significant, there is appeal to being able to have some cameras at 5 or 10 MP.


I'm assuming the warehouse runs are RG6, not RG59U? If they aren't RG6, you're going to need to rerun those cables anyway. You can only get an HD-SDI signal ~100 feet on an RG59U.

What is of value to the customer?

You make two pro HD-SDI points (are you specifically looking at hdcctv, or hd-sdi?) about latency and live viewing, yet most systems are not viewed live very often, and I would say latency is rarely an actual issue in IP systems. Are you listing specific "Pros" the customer values, or just general items?

Have you tested any of the existing cables or otherwise checked to make sure they would work as expected?

Related to the cost of Ethernet over coax converters, and total re-wire costs, you can carry more than 1 IP camers over a coax link, though it requires you adding a switch at the other end and connecting multiple cameras to that switch. It may or may not actually be easier to better to architect the system that way.

Topology things aside, the down-side to HD-SDI, IMO, for something like this is the general lack of diversity in the equipment, and the lack of probable new form-factors. You're not going higher than 2MP anytime soon. PTZ choices are severaly limited. Things like panoramic or multi-sensor choices are limited. I'd be hesitent to recommend that a customer with that many channels invest in HD-SDI, for fear that future upgrade and expansion options might be forever limited and they would "blame" you for painting them into a corner in that regard.

You're wise to ignore previous political rants and discussions on this topic and pursue the best possible option for your customer, but it's hard for us to make recommendations just off the data you provided. My summary analysis of HD-SDI is still that it's best suited for small jobs and trunk-slammer installs.

Nick, beyond the fundamental technical tradeoffs of the two, the big thing is the lack of major manufacturers offering HD SDI offerings. It's mostly companies with limited US support / coverage or low end brands like Speco.

Btw, the other major issue specific to a large system (e.g., ~180 cameras) is reduction in the number of recorders/servers one needs. Maybe there are 16 channel HD SDI recorded but even still that means 11 to 12 of them versus 1 or 2 VMS servers. This should primarily be a cost reduction but it likely also makes external storage easier as well.

First, I want to say that I truly appreciate your input.

I really just want to help create the best solution for my customer.

They're in a siuation where they have a lot of cameras from their perspective, and they just want to get something better.

They are a large, very cost sensitive company. To say the value of the company, I'd basically say I could make my quota for the next ten years this year. Not on this site, but they have many sites.

I've never been someone that "sells stuff." That's not really my gig. It doesn't fit my personality or desires. I'd rather make less money providing the right solution.

I've looked at different solutions. I can do it IP with 2 servers. Maybe less, if someone educates me. The challenge I have is the coax / 18/2 that is existing. A lot of it is old, some as old as 15 years. So, right now I'm torn between " Hey dude, we'll install it, and some of the runs are going to degrade the quality and we'll have to re-run some of them but we'll just have to wait and see," and "Let's just re-run everything at a major cost."

Couple things to consider with re-using the coax:

1. You don't really "lose image quality" with EoC converters... my experience, what happens is that they either work, or they don't. I've run into two installs where a bad/damaged/iffy cable run led to an intermittent camera, but when it was working, it looked fine. As Brian points out though, going to IP gives you a LOT more flexibility in infrastructure: in both these cases, the original setup had two cameras in the remote location, the coax running through a sealed conduit, with two sets of EoC adapters; in both cases, we added a small switch at the remote end, and then fed both cameras over the other, intact coax.

2. The technology is getting cheaper, rapidly - I have one site, again with non-replaceable coax in sealed conduit - running three cameras over a single coax with a small "SOHO" grade switch at the far end, and a single set of bargain-basement, $75/set EoC adapters to connect it to the DVR. Since there's no line power at the remote end in this case, I simply used a small 3A board regulator, split off the 24VAC from one of the camera feeds, and dialed the output to the switch's required voltage (5VDC, I think this one was).

Obviously these are much smaller scale than you're talking about, but the concepts translate well. On the whole, I've found EoC to be extremely tolerant of poor coax condition and type.

The other major consideration, as Ari notes, is that HD-SDI is still very distance-limited - even on top-quality coax, I believe it maxes out at something like 100m (although some future HDcctv rendition is supposed to fix this... Real Soon Now<tm>). Even cheap EoC adapters will do a couple hundred meters, and some claim 1500m or more.

We've recently been using Altronix's eBridge line and been very happy with them - I think a set of the basic models wholesales at aroung $120 now, and they have options to send analog video over the line along with the IP (hey, I found it useful once to run a composite monitor feed back to a PA monitor!), as well as a "16-port" model that basically combines 16-port switch with an eBridge tranceiver (neat thing is, this unit can be powered with 24VAC).

Hi Mat, Thank you for a clear informative post and also including the equipment supplier’s link.

Jim

Matt, thank you. That post was extremely helpful.

I ended up proposing a couple IP options for the intial fix after everyone's input here, and a couple other people's input which was basically the same ("Go IP").

I realize we will probably have some runs that need to be replaced, and I've communicated that to the customer, but that's also a much lower expense than running 200k feet of cable through warehouse space if we do it on an as-needed basis.

I am very interested in the altronix that you're happy with. That's much lower pricing than I've been seeing from other manufacturers, and it's helpful to know that you're happy with it.

Do you have model numbers or a link for the other EOC offering you wrote about?

Glad I could help, Nick :)

The "cheap" unit we've used is this one. We've actually used a couple dozen of them - they work fine in a pinch, but I'd hesitate to recommend this model just because I've had a few iffy ones... and because they use F-connectors (really!), and the spring-contacts for the center conductor tended to be fairly weak, leading to poor contact over time.

I've also bench tested and then implemented one of this unit and found it solid, but ultimately, the Altronix models come in close to the same price, I've never run into a bad one (yet), and I can just pick them up locally at ADI, making them just plain more convenient.

BTW, in the two cases above with the one bad coax, one site was using the cheap EnConn models, the other was using eBridges, so there was no real issue with one over the other there.

Disclaimer: I COULD be wrong on the cost of the eBridge - I know they were more like $199/set wholesale when we first started using them and I was looking for a cheaper option than Veracity, but since then I haven't really looked at the prices; we just spec them because they've been 100% reliable. I do know they've gotten a lot cheaper since then, though, and I'm PRETTY SURE the last time I asked about it, they were down to the $120 range.

Either way, if a solid EoC product will do the job without breaking the bank, what's a couple dozen bucks between friends? :) If you can put a half-dozen cameras in a remote location and run them all over a single existing coax, you're saving a ton over having to re-wire!

Writing a proposal with the line "but don't worry, I'm sure that existing 15 year old coax I haven't inspected should be just fine for an infinity number of years, so let's just zero out the labor costs for cabling", implicity or explicitly, is a really, really, extremely, specifically bad idea.

Some slick guys in nice suits from Lockhead Martin sold the New York City transportation authority a multimillion dollar surveillance system, years ago, and it is not finished. It is not close to being finished. No one realistically expects it to ever be finished. And while the many problems would make a hilarious movie, possibly starring Matthew Broderick as a bumbling MTA security director and Eddie Murphy as a fast talking camera salesman, the comptroller's report is enough to make you sick. And while the problems with the project are many and varied, one of the biggest factors is that Lockhead Martin just figured on reusing cable runs which turned out to be- suprise!- mostly bad. And then they tried to rerun the cables, and that turned out to be a bear of a job (for starters, you can't just turn the trains off whenever you want to so you can send a couple of guys with ladders and boxes of cable to replace the bad wires, not if you don't want three or four million late for work New Yorkers burn down MTA headquarters).

So, quote a new system with new wires, and say something like "and we wanted to quote you HDcctv cameras rerunning the wires, but since we couldn't guarantee the integrity of the wires and we cannot tell you how long the cameras would last, we decided not to, but if that's the direction you'd like to go, we recommend you plan on your system going down entirely in three to five years from now", which covers you in case a less ethical company comes in and bids a really cheap system on the back of the existing cable.

Brian, I've not tested any of the existing cables. Service rates are horrendous from my point of view. I try to do everything I can without incurring cost.

Understood, but you should probably spot check at least a few of them. Seems like cable re-use is a big varible cost in your intended proposals.

Sure sure. Unfortunately, it was a "I need a fix by tomorrow" sort of thing for the initial problem with the down cameras. I plan to do some testing over the next couple weeks. The major change isn't until January.

Nick, let me provide you with the benefit of our experience as a manufacturer of Ethernet-over-coax devices (Veracity). We've been selling our range of HIGHWIRE adapters for seven years now (longer than anyone else in fact), and the installed base is now vast. These devices have been installed onto all sorts of coax cable types (RG59, RG6, RG11 + others) and all ages of coax and across several hundred meters of cable, depending on cable type (RG59 is the most limited, but should still give you 300m - although watch out for copper-clad steel coax if you want to transmit POE as well).

The feedback from our integrators is that in the vast majority of cases the old cabling works well and stays working. Typically if a connection is a bit iffy, re-terminating with new BNCs solves the problem. If the cables were getting any kind of useable analog signal across them before, then generally they'll give perfect performance for the digital transmissions. The thousands of HIGHWIRE installations which have been in for 5+ years and are still performing perfectly. In fact some installations have upgraded their cameras again since (from standard IP to mega-pixel IP), but just kept the EoC adaptors.

It's also possible to buy EoC adaptors for the camera end which have a built in switch, so you can run up to 4 Ethernet devices down a single coax - note that not all the devices need to be IP cameras - an extra Ethernet link can be very useful for other network connections sometimes.

Just be wary that the EoC solution you choose provides sufficient bandwidth for mega-pixel cameras, as there are low-cost solutions on the market with limited bandwidth. You need to be aware that even if a camera is running only say an average 2Mbps, the instantaneous bit rates can be many times that, and if the adaptors can't handle that, you'll see frame stutter and drop-outs, and it will be nothing to do with the cable.

I take Ari's point about the NYC MTA system, and we also have experience of underground transportation systems. These are not representative of normal industrial CCTV installations and indeed can be very difficult due to the length of the cables (up to 1000m sometimes), the cable types (often not 75ohm CCTV cable), their age (30 years + !) and condition (they've been in a very harsh environment for a long time). We do have special solutions for such applications, but that's not relevant here. Just don't be put off by such (very real) horror stories. Coax cable is an excellent and long-lasting transmission medium. Reusing it is smart, lower cost and environmentally sound.

Just a comment on HDcctv - whilst it doesn't require separate adaptors, these are built into the devices at each and you still pay for them. The flexibility of EoC versus HDcctv is that you can use any IP cameras - indeed any Ethernet signal at all. In another 5 years time, we expect that HDcctv will have faded away.

I hope you find the above useful.

Sure, coax that I personally pulled I'd trust for 15 years+, because I know I use good cable, I know I use good installation practices, I know that I use good connectors properly installed. Someone else's coax? You don't know if they got the cheapest cable they could find, you don't know if they wrapped it around a fluorescent light fixture or a hot water pipe, you don't know if they ran short in the middle of a run and spliced a line together with duct tape and prayer- all things I've personally discovered walking onto other people's jobs, usually to my sorrow and at the worst possible time.

OP, if you do decide to try and reuse the wires, certify them all. Pull the connectors on the DVR end, to get an idea of not only the type of BNCs they used but how well they installed them and how well they're holding up. Tone out, identify, and measure the resistance of as many cables as possible- ideally, every cable. Ask the guards or the security manager- the people who actually spend the most time looking at video- if any of the cameras seem a bit flakey, which could mean not just a bad camera but a bad cable. Look at little details to try and figure out how professional your predecessors were (probably "not very", because if they were, they'd probably be doing the job, not you, right?). Is everything labeled neatly, with a numbering scheme that seems to suggest they were planning on writing it down and saving it in a file later? Professional. Is everything scrawled on electric tape in marker with descriptive names rather than numbered zones? A little trunk slammery but still suggesting they expected to be called back in the future, so, okay. No labels at all? Amatuer hour. Or lightning suppression. Is all the equipment properly grounded and plugged into good quality surge arrestors? Good. No grounding and plugged into power strips? Less than good. Are tie wraps trimmed, even out of the way ones high up in the trusses? Someone takes pride in their work. Are all the tie wraps loose and ragged looking and still have the tails hanging out? Someone wanted to get home as fast as possible, and knew that no one was going to care enough to yell at them for doing a sloppy job.

Bottom line, resuing someone else's coax is a huge risk. There are steps you can take to reduce your rsik, and if the potential rewards are great aneough and you can sell the customer on a service contract, you may want to go for it, but the risk will always be there. I wouldn't do it but it's certainly doable, and Veracity makes probably the best equipment to do it.

Alastair,

We just purchased several Comnet converters and experiencing dropped packets quite frequently. Would it be possible to

obtain a demo unit of the Highwire for testing purposes? We have roll of 1000' ft. rj59 in our lab for testing. If this resolves

our issue we more than likely will take the loss on the Comnets and purchase the highwire.

thanks in advance,

Tony

Hi Tony,

Yes of course, we'd be delighted to let you have a couple of HIGHWIREs for testing. They'll work perfectly over 1,000ft of RG59. To allow us to take this discussion offline could you please drop me an email with your full contact details to :

info@veracityuk.com

and we'll take it from there. We can discuss whether you need just data transmission (HIGHWIRE) or data plus POE over coax (HIGHWIRE Powerstar) and get you the right units. We can also provide detailed technical advice and support for specific installations.

I look forward to hearing from you.

Thanks,

Alastair McLeod

Veracity