Soundy

What is a "hybrid DVR card"? One with BNCs and a network port? With a Vigil DVR, if I want to go hybrid, I just enter the IP license keys and plug the machine into the network. Many (most?) other PC-based hybrids work this way as well. No need to change hardware or even software; no need to learn a different interface; no need to make two pieces of software get along. Apples to apples, sure... except a lot of PC-based DVRs are a complete fruit basket. Like I said, you just haven't been comparing the right systems.

A gigabit switch will let cameras and server communicate efficiently - a 10/100 switch would probably choke on the total load of those cameras. Your internet connection itself, though, will be very limited in bandwidth unless you have BIG money to spend on it. We once priced a dedicated 100Mbit connection for a customer who wanted to set up for offsite recording... it was going to be on the order of several hundred dollars *per month*. A typical residential or small-business broadband connection, at $40-$60/month, won't usually give you more than 10Mbit of upstream (ie. from your system, outward), which would be enough to view one or two cameras smoothly, but no more.

Depends on the PC systems you're comparing to. I have yet to see a standalone that comes anywhere close to Vigil's interface and feature integration. And yes, I *like* having access to the desktop and underlying system to make tweaks/customizations/repairs/etc. The ability to take a drive from a dead system and access it directly on another. The ability to add a massive iSCSI storage array. The ability to add other support software. The ability to actually FULLY access the system remotely, rather than just the few options the remote client gives me. I may start with a standalone system for all-analog, but what if I want to add IP cameras later? Do I toss it out to replace it with a hybrid? If I'd started with a PC-based machine, I can just add the NVR software or module later.

For that many cameras, I'd definitely recommend a gigabit switch. However, your remote viewing speed will be limited by your internet connection speed, which will typically be far below even 10Mbit, and the switch you use will have no effect on that.

Are they still around, when i go online to see what they are about, i see many sites that speak of their products being discontinued... There was a thread about this in the Dealers forum a while back. If memory serves, Samsung kicked GVI out of bed, and GVI pretty much folded after that. Should be able to find more info by googling.

This is the case with built-in IR *in general*, not just with domes - again, IR has its place, but for the most part, it's just used as a cheap workaround for a crappy camera.

You're probably comparing poor domes to good bullets, then. There's nothing inherent to a dome design that makes it more susceptible to internal fogging, rain, distortion, etc. Many IR bullet designs actually put the lens and IRs behind the same glass as well, with nothing more than a foam gasket around the lens for light isolation. Better designs have the lens protruding through the glass over the IRs, or put each behind separate pieces of glass. Either way, there are still issues with ANY cameras with built-in IR - a common one is how the cameras attract bugs, which then attracts spiders to build webs in front of them. Separate illuminators are a better way to go... or better still, motion-activated flood or spot lights combined with proper true day/night cameras, so you can actually get COLOR images of your prowlers.

It sounds like a ground loop typical of using a central PSU and common-video-and-power-ground cameras with baluns... if so, using separate adapters for each camera would clear up the issue. What make/model cameras are these?

GVI used to be an affiliate/offshoot of Samsung, so I'd expect it to be a pretty decent camera. Most of the prices I've found for it are around $1200, and since you get what you pay for, to a large extent, in this business, I'd say that's a good indicator as well. The specs on it look good - TDN, 1/3" CCD, etc.

So is there a point of using a pc for a IP CCTV Surveillance System ? Even if I use a server grade components, would that still not do the trick ? You have to remember that Rory lives in the tropics - the heat his systems have to deal with year-round isn't relevant in most of the developed world. PC systems offer some major benefits, not the least of which is how expandable they are. Not a lot of standalones will let you extend your storage with USB, Firewire, eSATA, or network drives, or using RAID storage arrays. Not many support offsite storage, something that can be easily configured on most PC-based systems. Usually, the DVR software doesn't even have to support it; as long as you can give additional storage a drive letter, most software can use it for writing video. Of course, if you are using IP cameras, a PC-based system will allow you to install the necessary manufacturers' configuration utilities, something you can't do with a standalone. You can also install and run backup software, time-sync utilities (if, say, an NTP server isn't an option), mail-relay utilities, etc. You can directly access the file system to extract or backup video files, something almost no standalones will let you do. For that matter, you can remove the drives from a dead system and usually extract video from them or even plug them into a working system and still read the data, something that can be a sketchy proposition on a standalone (some will immediately format any new drive you insert).

Look at it this way: you're going to pay more for an 8-channel capture card (or standalone DVR, for that matter) than you will for a four-channel unit... and you'll pay more for a 16-channel card/DVR than for an 8-channel. So whichever way you go, the more cameras you want to record, the more it's going to cost. The benefit of a per-camera license structure is that you only pay for exactly how many cameras you use. For the sake of easy math, assume both IP or analog will cost you the same per channel: $100. A four-channel card will be $400; four IP licenses will be $400. But now what if you want to go to 5 cameras? With the IP solution, you pay another $100 and add the fifth camera... with the analog solution, you either pay $400 for a second four-channel card, or you pay $800 for an 8-channel card to replace your now-useless $400 four-channel card. Or in the case of a standalone DVR, you pay $800 for the eight-channel machine and the four-channel box becomes a doorstop. Yes, I realize that actual pricing for analog devices doesn't work that way (cue Rory with his $300 Dahua toys) - as I said, the numbers were used just to make the math easier. Point is, with analog units, at some point you end up paying for inputs you're not going to use, driving up your average per-camera cost... with most IP license schemes, you only pay for as many cameras as you need. Yes, IP cameras cost more... BUT, part of that extra expense is the cost of the video-to-digital-stream encoding hardware. With analog cameras, that would be the capture card. So if you factor four VGA (640x480) IP cameras, vs. a four-channel D1 (704x480) DVR + four analog cameras, you'll find the costs aren't that different - it's the same process, you've just changed the location of part of the hardware. The other thing you pay more for is resolution... and this is also true of analog. A DVR that does D1 on all channels will cost more than one that does only CIF on all channels. Likewise, if you want to go even higher in resolution, that will cost more as well. Except, well, analog is limited to D1. Period. D1 vs. 1.3MP (click for full size):

Actually, the majority of IR domes you see will have the IRs inside the dome, and then usually have a foam gasket around the lens that snugs up against the bubble to help block reflections from entering the lens (it's never 100% effective). The thing to remember is, IR is built into cameras in the first place, usually to compensate for poor low-light performance - it's cheaper to take a junky camera and add a bunch of 2-cent IR LEDs and sell it as a "night vision" camera, than it is to actually build a camera that works well with little ambient light. At that point, it's also cheaper and easier to just slap the LEDs behind the same bubble and isolate the lens with a gasket, than it is to put the LEDs "outside" the bubble, whether it's really the best idea or not. You can still sell it as a "night vision" camera, and if there IS a problem with that particular design aspect, you can always blame the installer or user (well, you must have moved the gasket/scratched the dome/etc.)

That's why the links get edited out of the posts.

I've been editing some of her posts from the very beginning. But threads like this are entertaining

My ASUS netbook's LED-backlit screen does quite well even in direct sun.

I've only ever seen one camera damaged by lightning... in that case, the lightning hit the power mains and also fried a 16-port switch and a video card. Amazing that those three things are all that was damaged, actually.

Yeah, the HDMI on your laptop is output-only. Seems to me though, HDMI should be telling the DVR what resolution the monitor supports, and the DVR should be adjusting its output accordingly. Unless the DVR has a lousy HDMI implementation.

Why does it have to be wireless? How about a simple desk button? If the doors have accessibility buttons on them, just tie into those.

Actually, that's pretty middle-of-the-road (around CDN$79). One hDVR/NVR I work with, they list MSRP$150/camera unless it's their own branded cameras, bought from them, for which they list $50/camera. Some others charge per block of cameras - $X for up to 4 cameras, $Y for up to 8, and so on. GeoVision will let you use as many of their own branded cameras as you want for no extra charge, but charge per-camera for other brands (if memory serves, it starts at $100/camera and gets cheaper if you buy in bulk). If these companies simply charged £49 for anything up to 16 cameras, they'd never make any money - the bulk of NVRs are already 16 channels or less. If there's no money coming in... there's no more software development. Well, you can go to a DIY Linux-based system... be prepared to spend a lot more time futzing with it, tweaking and coercing, to make things work properly, or at all... maybe you'll come to understand some of the amount of work it takes to create a ready-to-use software package and realize why they charge so much for the software. He's saying some cameras - such as certain Axis models - have the ability to write directly to a network share, and thus don't need an NVR or the associated licensing. Of course, that will just give you files on a disk... it doesn't necessarily give you a software front-end for searching and exporting video.

Just takes a setup like this: viewtopic.php?f=19&t=11322&start=82 viewtopic.php?f=19&t=11322&start=108

I would bet on faulty or damaged cable as well. Obviously if you get picture on your test monitor AT the camera, power is not an issue. It's possible there's a bad input on the DVR as well. Have you tried a different input, or tried connecting the DVR end of the run to your test monitor?

If they're REALLY insistent on this idea, maybe they should look at just streaming one or two cameras off-site, rather than all of them...

Don't forget that DSL is usually optimized for downstream-only, since your typical internet user only cares how fast the web pages come to him, not how fast things go out (plus, there are technical aspects that usually limit the upstream bandwidth). Their service may be 4Mbps downstream, but may not be anywhere near that upstream. POS, timesync, employee tracking, and other such office software will take very little bandwidth on their own.

Looks like Tavcom is strictly a brick-and-mortar school in the UK... do they offer any online courses? Anyone know of any similar online training? I got a couple co-workers who could really benefit from learning IP basics