Soundy

Well, I'm talking box cameras with C/CS mount in particular, since it IS easy to change out for a wide variety of lenses with those.

Apparently, rory, everybody but you, me, and Alex. Seriously... I could count on one hand the times maximum lens aperture is mentioned when low-light performance is discussed. Chip size, chip type, chip brand, chip TVL, chip flavour, chip crunchiness, post-processing, IR/non-IR, phases of the moon, and even the shape of Oprah's hairdo this week are all covered to death when it comes to things that affect low-light performance, but how often has anyone (besides me) suggested trying a faster lens? And it's not just in here - when's the last time you saw marketing and advertising touting the lens's f-stop and specifically suggesting widers ones for better low-light operation?

I'm constantly amazed at how the lens is almost completely overlooked in this equation. The vast majority of the lenses in use that I come across (whether new ones, or on existing installs) are between f/1.4 and f/1.8... yet I do have some that are f/1.2 and I've seen them as far down as f/0.95. These don't seem like very big differences, but f/1.2 is a full stop faster than f/1.8, which means it allows TWICE the amount of light. In fact, from f/2 to f/1 is TWO stops, or FOUR times the light passage. When low light performance is a concern, simply changing the lens could make a HUGE difference, with ANY camera.

Christopher thx for pix but I was talking about CCTV cameras only by the way Avigilon uses the same type of sensor as the 1DmkIV Granted, it doesn't have dual-Digic-IV processors behind it... but the light-gathering capability is greatly improved with the MUCH larger sensor.

I believe eSATA is distance limited as well. You may need to look at network-attached storage instead. I'm not that familiar with Geo's capabilities, but I've never seen anything in one that suggests it can write to two disks simultaneously. I'd think it far more likely that the machine in question had some sort of mirroring set up, whereby the system itself maintains the second copy of what's being written to the drive, perhaps using something like these: http://www.usbgear.com/USB-RAID-System.html

The reason you had to set up port forwarding in the DSL modem is because it's actually a combination modem/router (I'm guessing it also has four LAN ports and wireless?). The SB5101 is *just* a modem; no router or firewall functionality. I would STRONGLY recommend that you add a router to the cable modem setup... not just for the firewall, but because the NAT layer adds some "physical" security, effectively separating your local systems from the open internet.

How do you think they are placing the text over the video? Huh? With analog, same as always. Doesn't work any differently. My point above is, text-overlay-on-video of POS logging should go away. There's no point to it when you can log the data right beside the video and make it searchable.

Via remote client: Direct on DVR:

Got it. But I do want to take advantage of my cameras as far as I can, so a 540 TVL does allow for higher captures, resulting in lower FPS. Nyet. Two problems with that logic: One: maximum vertical capture resolution for all these cards is 480 pixels. Thus, anything over 480TVL is largely wasted... and the difference would be virtually unnoticeable anyway. Two: If the card only does D1 resolution at the expense of total overall framerate, then you need to record at CIF to get maximum framerate, thus further "wasting" a higher-TVL camera. Is it possible that after paying for this card, at this price, I would have to pay for each CCTV cam additionally? Not on any system I've ever seen. Vigil, VI, and GV at least, if you buy a 16-channel card, you get a full 16-channel-capable system. Period. The only way I've ever seen per-camera licensing on this type of a setup would be when you go to IP cameras, and not all of them use that sort of licensing structure either (some license it in blocks of 2, 4, 8, etc.)

Last one is also one of the longer runs. Near as I can tell, the DVR is set for D1 on that channel, but I pulled all these off the remote client, so it may be scaling it down to CIF for transmission, then back up again for display... not sure, I'll have to check directly on the server when I get the chance.

I wouldn't confuse MP and HD at all, personally. "HDTV" defines a set of specific resolutions, frame/refresh rates and aspect ratios. "Megapixel" includes pretty much anything over 1,024,000 pixels. Technically, 720p isn't megapixel (1280 * 720 = about 0.9MP), and 1080p is barely over 2MP (1920 * 1080).

Why not just set up all four quad groups with the same camera layouts? Sure, it's a bit of a kludge, but...

Umm, yeah, right here. Don't have sample footage, but I have installed a couple. It's no different, image-wise, than having a small board or bullet camera mounted beside the door about 5' off the floor (which I've also done)... just a different enclosure.

I can think of one major problem with your plan right off, and that's the limited range of WiFi. You're talking "walls and floors" - how many? What distance? What sort of construction? If there's concrete (and hence, rebar), your range will be severely limited indeed.

Likewise with the Vigil DVR/NVR - POS data can be captured, cataloged, and searched directly in the main interface. Text overlay should have died with the time-lapse VCR.

How about once you add in the PC for monitoring?

Normal lighting, I'd say they're pretty close... I haven't tried them with any sort of severe backlighting yet, though, so I can't say how they'll compare to the SuperDynamic function.

Why not just a single 32-channel card? Less space, less complexity, and probably lower cost. Vigil and Video Insight have them - I've used both and they work great. Vigil even has a 32-channel/960fps (full 30fps on every channel) system. Naturally, something like that will be spendy - for the card, AND the PC necessary to handle it. But as you've noted, even 10fps looks smooth in most cases, and you certainly don't need a beast like that. Here's another thought too: instead of assuming that you need the full 10-15fps on ALL channels... identify channels where you can do with less. Lots of times you may find something even below 5fps will suffice. Nobody says all channels have to use the same framerate, and when you're looking back at recorded incidents, rarely are high framerates NEEDED. Sure, they look nicer, but they don't necessarily give you any more information.

Okay, got those pics from the other VCM-20VFs. Same site, same DVR... I think I know what part of the problem is now with those patio cams, though: we used some ultra-cheap baluns (because that was all we could get on short notice), and the patio cams are a good 4-5 times the distance of any of the others (on the order of 250-300'). Now I'm thinking that the cheap baluns may not be faring so well on the longer runs (some of these are only about 30-40'). The shorter runs, however, seem to be just dandy:

That could be said of any camera position, though. One site, they had a camera on an 8' ceiling watching an under-counter safe... the cleaners rolled in a cart with a stack of cleaning supplies and neatly blocked the camera... very innocuously, of course. (We added a second camera with a reverse angle on the safe, from a 12' ceiling... hard to block BOTH and claim it's accidental!) The whole point is to get around guys who hide themselves from typical ID-shot positions with ball caps and hoodies.

Yeah, the camera itself doesn't have to be IP for this to work (unless they need megapixel, which doesn't appear to be the case): any analog camera can be plugged into any "video server" box that will allow it to be plugged into the network. Get one with an audio input if they really want to go that way. Frankly, if they're only listening and watching, and not recording, I see little reason to have a whole separate PC just for monitoring the feed. It adds a lot of expense and complexity (not to mention the extra power consumption and noise generated by yet another narrow-purpose PC) and provides almost NO benefit - the only possible reason to use a network(ed) camera and audio in this case would maybe be the ability to easily access it from any PC on that network, but I don't think that's a capability that would be desirable in this instance (from a privacy standpoint, would you want YOUR counseling session to be accessible to anyone in the building with the proper information???) This may be an instance where you need to put your foot down and TELL the client what's best for them, instead of letting them rely on what they think they WANT (as opposed to what they NEED). From the sound of it, someone wants to pretend like he's James Bond and be as high-tech as possible, with no real thought to the benefits or drawbacks involved.

I've tried different refresh rates, yes, but most LCDs only support 60Hz, and if they do have more than that, it's only 75hz, so there's not usually a lot to play with.

holy moly batman. LOL Also about the cat5 VGA extender, I saw one at a location and it was okay ... the client said he had nothing but issues with it. I didnt install it and I didnt change it. He was not my client. We've been using MuxLab VGA-over-Cat5e extenders (50010/50014 pairs, if memory serves) on a number of sites with the Vigil HD Viewer systems... they work well, but some resolutions can be a little twitchy, depending on the video card and the monitor. I did discover, though, that the MuxLabs at least are a LOT happier with 568A wiring than 568B - could be the way the individual channels are balanced for the varying capacitances of the pairs. Don't know if this applies to other brands. They do recommend SHIELDED Cat5e, but we've been fine so far with just standard stuff and crimping on our own RJ-45s.

Can't say that I've ever seen an S-to-component dongle like that. I know where I used to work, we had a line of Matrox cards that came with a VGA-to-S adapter; we almost never used those. I also have a number of ATI cards kicking around with S-video ports on the back. I'm just looking at two Nvidia cards sitting on my desk, one of which has DVI, VGA and composite (single RCA) outputs , the other with VGA, S-video and composite outputs. S-to-component would NOT be that easy without some sort of upscaling circuitry.

Welcome to the board! The first thing you need to do is find out (from the store owner/manager) what the "areas of concern" are - where do they have their most problems/losses/incidents? Do they experience employee theft from food storage, or from cash drawers? Is there a lot of trouble with shoplifters? In general, you want to start with a tight shot on the main entrance, to get clear facial shots of people coming in. This is usually the camera you want to spend the most on, especially if there's bad backlighting (bright sunlight outside) - you want something that can handle the dynamic range. Next, plan on at least two wide shots of the whole store are common, usually in opposing corners of the main floor area. On 1/4" or 1/3" cameras, something in a 2.8mm lens will get you close to a 90° field of view. Don't expect a lot of detail in these shots; the main idea for them is to get an overview of what's happening in the whole area. From there, you can concentrate on the particular areas of concern. If they have a high incidence of shoplifting of particular products, consider a camera or two near those areas, with views just tight enough to capture those areas with as much detail as possible. If there are issues with employees skimming from the till, perhaps a covert camera over the paypoint will be required. That should be enough to get you started.