Soundy

Got the following query from a friend: Memocam are the only products I'm even slightly familiar with (and that, just through other people's suggestions). Found some various hits online at assorted online stores, but have no idea how good or bad they are. Any more thoughts, anyone?

Setting aside for a moment the fact that both are cheap junk... You don't NEED 30fps for surveillance purposes. You're not producing a TV show here, you're just recording security video. 8fps is more than sufficient for most purposes, particularly retail. Give the limitations of the Q-see, of the two, I'd say you're better off with the Clover, running D1 at 8fps on all channels.

Yeah, functionally it's no different than 8 individual baluns... it's just all neatly tucked inside one snazzy case

Yes, you can extend your coax runs using baluns and UTP (it doesn't have to be Cat5e, or even twisted pair necessarily; any sort of cable pairs will do, even Cat3 or station-Z wire). Check out using a pair of these to simplify things: You can use Cat5e/Cat6 with RJ45 connectors, to keep it clean, or any other sort of wire pairs via the screw terminals. Most other balun manufacturers (MuxLab, NVT, GVI, etc.) have similar products.

^Agree with tom, see how the included drive does. Remember if you're on motion detect, it's only going to record when something happens, so if you have cameras watching low-traffic areas, they'll use very little drive space. You also don't need to run full framerate - standard video is 30fps, but for security purposes, the difference with 15fps is almost indistinguishable, and even 7.5fps will give fairly smooth motion.

It also depends on the DVR - all of them SHOULD automatically do "recycle" recording, where it just overwrites the oldest video, but a few do default to just stopping once the disk is full. How much you can get on 1TB varies widely as well - it could be anywhere from a few days to several months, depending on the number of cameras, the types of cameras, the resolutions used, the framerates used, the codec used, the quality level it's set for, the amount of movement, whether you use constant or motion recording, pre/post-motion settings, whether you use alarm/triggered recording, whether you record in color or B&W... and on and on and on. Some, but not all, support scheduled recording. I agree with tom, you don't usually want it to ONLY record during the scheduled times, but more advanced scheduling can be useful at times (for example, I have the paypoint camera in one store set to record constantly during business hours, but only on motion outside those times). Without knowing exactly what make and model of DVR you have, it's impossible to give any more than generic advice.

Analog... how quaint

If you're comfortable with an all-IP system and figure your network can handle the load, why not look at a standalone NVR/NAS solution like those at http://qnapsecurity.com/ (just for one example)? They don't have a dedicated Android app, but QNAP do have an online web demo that you could test with an Android browser. Synology have similar products as well. If you still want to build your own PC system, take a look at NUUO and Exacq; they have analog cards available, but also work as IP-only systems. I've toyed a little with demo installations, but others here are familiar with both, and both seem to be pretty well-regarded. Or check Video Insight for an all-IP server (they use external video-to-IP encoders for analog support), they have a really nice browser interface that works on every browser I've ever tried it on - http://www.demovi.com Cost-wise, yes, analog cameras cost less than IP, but you'll also need analog capture hardware, as opposed to JUST software with an all-IP setup. And again, if you figure your existing infrastructure will handle it, you can avoid the cost of cabling specially for analog.

I have a car. Will it go fast? Can I put six people in it? Your question is impossible to answer without at least knowing the make and model of the card and the software you're using.

Braided copper shield is the best for CCTV. I've had nothing but trouble with foil-shielded wire. Most here will agree.

Hahahah, spam fail: Now then... I heartily agree with Ted's suggestion... DIY is great for home setups, but in a retail situation where proper coverage and operation of your system could save you thousands of dollars, you should at least CONSULT with an experienced professional who can come *on-site* to help determine your requirements. Liquor stores especially are prime targets for shoplifting, robbery, break-ins, fraud, and yes, employee theft, so you really don't want to skimp on a security system. That aside... a typical system we'd put in a liquor store would include at least two environmental domes inside the coolers (may more, depending on the layout), to watch for employee theft and accidental breakage. They need to be fully sealed, and ideally have internal heaters, to avoid fogging... having an open wire path from the camera interior out through the cooler wall is just inviting warm, moist air and thus condensation. We'd have two on the paypoint from opposite angles, both to catch customers' faces and actions, and to watch transactions - these are often used in credit-card fraud cases. Megapixel cameras are particularly useful here. You'll also want enough to cover all the liquor aisles and cooler doors, ideally feeding a monitor or two hanging in view of the paypoint, so counter staff can monitor activity throughout the store. Quality analog cameras with a wide view is usually sufficient for this - you want to see people slipping bottles into bags or coats, but you don't need to read the labels on the bottles. Domes are preferred because they can't be easily knocked out of whack by a clever shoplifter. To get a clear facial shot of everyone entering, you want a WDR camera (or ideally, a Panasonic SuperDynamic, although those are pretty spendy) to deal with the usual harsh backlighting, with a tight shot watching customers entering the front door. We'll usually add a "customer awareness" monitor facing the front door, and displaying the front-door camera, so customers entering can see immediately that they're on camera and that the camera has a clear shot of their faces. Alternately (or better, in addition) to the standard entry shot, you could put one on the exit to catch faces when people are leaving. The height-strip cameras are good for this (this is one example, there are many others out there) - for one, shady characters are usually well aware of entrance cameras and often wear a hat or hoodie and keep their heads down to hide from them... but the height-strip cameras are face-level and covert, so it gets a straight-on facial shot that they won't be trying to hide from. Most are angled or adjustable, so you can also angle the height strip for better viewing from the counter. Most of our liquor stores also have a camera in the office - again, just in case of employee theft. We've also installed the odd covert camera directly over the paypoint drawer for employee skimming, and in one case, over the drop safe to catch an employee who was fishing moneybags out with a coat hanger. Now here's the thing: depending on the size of the store, you PROBABLY don't need a full 16 cameras. Most of ours make do with 8-10 (we've even got combined liquor store/pub systems coming in at under 16 cameras, with full coverage). The key, again, is defining the proper locations, and spec'ing the proper cameras for each location, rather that just running out and buying a bunch of gear and tossing cameras willy-nilly all over the place. That's something a professional can help you with, and potentially save you money.

What sort of "delay" are you talking about? Live local viewing? Live remote viewing? If you use a hardware overlay DVR card (like a Hikvision, GeoVision, etc.), there is no delay; local live display is instantaneous. Remote latency will be more a factor of the connection bandwidth than the codec used. It would be easier to recommend something if we had a better idea of just what you're trying to do...

"Fixed dome" as opposed to PTZ - just means you have to adjust the camera position manually, and it remains fixed after that. Most dome cameras you find with "automatic" varifocal will be PTZs.

Why not just test it on the bench? 800m on a single spool? Put a connector on each end and put a camera on it, see how it looks on the other end Hard to tell from the picture, but does it have foil shield? Does it look like the core is solid copper? There are no end of stories here of problems with aluminum or foil/wrapped shields and copper-clad steel cores - if it has any of that, you probably do just want to toss it (or better yet, take it to a metal recycler and get a few bucks for it!).

DVI spec allots some pins in the connector for analog VGA video... if your video card supports DVI-A, you just need the proper adapter. Any electronics or computer store should have them; most video cards these days come with them.

In about 99.9% of Windows-based software... Help menu -> About.

That thinking is fine if all you're dealing with is "immediate" events like robberies and such, where someone's going to go back and look at the video within an hour or two of the incident... a good portion of the time though, nobody even KNOWS to go looking for the evidence until a month or more after the incident. In retail especially, "book" losses don't usually show up until monthly reports are done. With credit card and payment fraud, it can sometimes be two or three months before things make their way through the chain and someone actually wants to look at the video. You have to still HAVE the "perishable" video a month or more later to search through so you can FIND the evidence.

Yeah, those aren't actually serial ports. Serial ports on computers tend to be 9- or 25-pin male connectors, for one thing... what you're showing is very common for DVR cards and use 15-pin female VGA-type connectors. What software works will depend to a large degree on the exact card and what capture chips it uses.

And you're expecting the average storefront electronics salesdroid actually knows what he's talking about? That's a crap shoot at the best of times. Sure, but from their perspective, there's nothing WRONG with their product. If you bought the wrong tool for the job... that's not their problem. To invoke an analogy: you want to turn one piece of wood into two, so you go to the hardware store, tell them that this is what you want to do, and they sell you a hammer instead of a saw. Both will do what you ask, but only one in the way you want to do it. As long as the hammer isn't broken, the manufacturer of the hammer has no business in the transaction. Why? Did you ask? I bet the software won't make toast, either, but they didn't tell you that. There are some things you have to think of to ask for yourself. Agreed, but that's the store's job, not Q-See's. Yeah, the thing is, if it was a high-end brand, there'd be a higher expectation that the salesdroid would know the product and the technology. As it is, he's selling a cheap consumer brand that's available from a bazillion eBay vendors, or from the local Costco. I'd have no more expectation that he'd know his stuff, than a TV salesman at a big-box furniture store would understand the intricacies of NTSC video spec. In his position, he's not "required" to know, just to sell a mass-market product. Given that the store has suddenly up and disappeared... I'd say this is a good guess. He told you whatever you wanted to hear to make the sale. Wrong. I'm not aware of any CCTV software that works this way. In some circumstances, this would result in massive numbers of very small files (some file systems have limitations on the number of files), or in a few excessively large files if there's lots of continuous motion. Once again, since editing and splicing IS NOT a requirement of typical CCTV markets, and in fact is generally not desirable for CCTV uses, this is not something that would generally be a design consideration. I'm familiar with three different methods Windows-based DVRs use for storing video: one that creates proprietary "bank" files and then writes the video data into those; one that continually creates files up to a user-designated size (200MB by default, meaning it will keep writing one camera's video to the same file until it reaches 200MB, then start a new file), and one that creates individual files for each minute of the record time.

^What's the video link?

No idea - never had to actually buy one individually. I might be able to find something in a google search...

I like the plug style on the dual-voltage CNBs... near as I can tell (haven't actually tested the theory), they're the same as the connectors on the Pelco box cams we use, as well as the IQEye box cams... interchangeability between manufacturers is a big plus for me. The screw terminals and recessed pins on the supply side makes them easy to connect and leave hanging without worry of them shorting against anything.

Unfortunately, you're partly to blame for your own predicament - you're trying to accomplish a certain end without fully understanding how to get to that end, and your purchase is based partly on not asking the right questions. What you're looking for is not specifically a function of the card; it's a function of the software. The only way the card comes into play is in the software it supports (or the software that supports it). In short, with the proper software, it's POSSIBLE (though not likely), that the CARD will do what you want. Of course, the trick there is that many cards won't work with software other than what they're shipped with, and most software will only work with the specific cards that it ships with. This is usually by design, intended to lock the two together. Beyond that though, a CCTV system is really not the best way to achieve this end. Since surveillance video must remain unedited to be of any use as evidence, the ability to edit is not generally high on the design priority list for CCTV manufacturers. Also, for it to look "normal", and to work properly in most video editing applications, you really need to record at a full 30fps, and at a full D1 resolution to match DVD quality... and it must do it on all used inputs simultaneously; that sharply limits the hardware that will fulfill your requirements.

I generally use the blue pair for video (blue +, white/blue -), orange and green pairs for power (orange pair for hot, green for neutral on AC; or orange for +, green for ground with DC), and leave the brown pair as a spare (future additional video, RS485, or whatever). The brown can always be separated and added to the power lines if a little more thickness is needed there, although I haven't yet run into a need for it. Seeing as the twists are designed to aid noise cancellation and reduce cross-pair capacitance in the upper-MHz range, I really don't see them having any significant effect at 50/60Hz (I doubt it would even be measurable with any but the most sensitive gear), and none at all at DC.

I was talking about stand alone DVR. So was I.