Soundy

Vigil will do much (most) of what you're looking for, including a client that can connect to and view multiple sites on a single screen. I don't know about doing it on four screens - you could probably expand the single client view across four monitors with the appropriate video card, or just use one really big monitor. It will also record in a higher quality than it streams at, and can adjust the stream for your available bandwidth. Keep in mind, at some point, the downstream speed at your monitor station will start to become a bottleneck - eight simultaneous streams at 384k each is 3Mbps. You'll want to make sure your station has AT LEAST that much bandwidth (factor in other regular internet traffic as well).

Seriously... 16 video streams on a 384k uplink is going to be beyond painful. No DVR can overcome the laws of physics.

He's asking how you're exporting from GeoVision to that AVI file. You say you're recording at 720x480, but that clip is only at 360x240.

How far is this troublesome PTZ from the power supply, what size wire are you using, and what voltage is the power and camera (12VDC or 24VAC)? PTZs tend to draw a lot more power than fixed cameras, and if it's too far away or the wire is too small, the voltage drop caused by the higher current draw could cause the camera to "dip" in and out.

Wowzers... and it was just three years ago that we put in a FLIR Wide-Eye 180-degree view that listed at something like $25k

Ummm, not really... this forum is for security and surveillance video cameras, not consumer still cameras. I'd recommend sites like http://www.dpreview.com, or http://www.photographycorner.com/

Each shutter actuation does not produce a frame of video. You're looking at it backward: the video's framerate is fixed. It will be 30fps (assuming NTSC analog video) regardless of whether your shutter speed is 1/100,000s, or 1s. Are you familiar with photography at all? That may be a simpler way to look at it: every time you press the button on your camera, you take one image, completely regardless of what the camera's shutter speed is. The faster the shutter, the less "motion blur" you'll get, but the less light you get as well.

Move away from Alberta?

Good sales pitch. Did he also explain how shutter speed affects exposure? You'd need a ridiculously bright light on an object to obtain a usable exposure at that speed. No; the two are unrelated. Not necessarily. For one thing, NTSC video is limited to 30fps anyway: the camera itself is still only snapping 30 pictures per second, but for each of those, the shutter is open - and the sensor is collecting light - for only 1/100,000 of a second. See previous paragraph. The camera is still only sending 30fps. Double-shot of Crown. Helps make everything clearer.

Whoa, don't blame me, you started this! Some interesting stuff going on here, but let's not forget the original query: dejota works for a manufacturer and wants to sell his products to a distributor/installer company, and was looking for selling points to highlight that would be attractive to installers... Things that he can say, "Look, as installers, I know you'll appreciate these features...."

Unfortunately, no, not really. Neither is a model name, and HikVision makes a number of different standalone units. Maybe if you have pictures of them, someone (scorpion?) will be able to identify them. Scorpion has manuals for just about every DVR ever made

Auto-iris allows the camera to control the lens iris to adjust for varying lighting conditions. There are often two related options as well, DC and Video - DC means the camera actually controls the iris, Video means the camera only supplies power to the lens, and a sensor in the lens controls the iris. Backlight compensation, to simplify, brightens the image so that dark objects can be seen in strong lighting conditions. This is most often used in instances where, say, the camera is facing a door where bright outside light will cause it to dim, and the lower indoor lighting means people (and particularly their faces) will simply be dark blobs. White balance allows adjustments for different "temperatures" of lighting. Incandescent light, for example, tends to have a lot more red and yellow than other lights, while fluorescent tends to be more blue/green-ish. That's why you'll often see fluorescent light over the fresh veggies in the supermarket, but incandescent over the meats. Your eyes and brain process light to compensate for such color shifts, but because camera sensors see all colors equally, they'll show those color castes, unless you adjust the "white balance" to allow for it. Slow shutter provides better low-light images by reducing the shutter speed below the normal, which allows more light to be collected by the sensor... however, that also allows motion blur to occur. Because of this, slow shutter is not normally useful for CCTV; there are instances where clear picture is more important than lack of blur, however, so this option is provided on some cameras. The Mirror option simply flips the picture left-to-right, as if you were looking in a mirror. Some cameras also allow you to rotate or flip the image vertically (top to bottom). This can allow you to compensate for a number of different installation situations.

That is a very cool link! Interesting that my no-name Bluetooth adapter shows up with a Logitech-assigned MAC address! Don't know if it helps with chipset manufacturer though... my motherboard's onboad NIC comes up with a range assigned to ASUSTEK, but it's a Marvell chipset.

Interesting... I've never even thought about what ethernet controllers were being used in these things, or how 'hackable' they might be. Pretty much all IP cameras do support firmware updates, so it should be entirely possible to hack that for added/different functionality, but I don't know to what level you might be able to alter the network interface. What sort of modifications were you thinking of performing? I'm not a coder, so I can't really help in actually doing the mods, but I'm interested in this as a concept...

Yeah, touching the lens is a way bigger concern - nothing like getting in there with a microfiber cloth trying to scrub the fingerprints off a board camera's little lens element

And this, kids, is why a $1000 DVR is better than a $100 DVR... Sometimes you gotta look BEYOND the specs and numbers!

I've been in the business (officially) for just over six years, I've never seen ANYONE use a ground strap while installing a camera. Not saying you can't do it, just saying I've never heard of it being done.

Before you go tossing money at the wrong piece of the puzzle, what is this for? What is your budget? What is your existing equipment? If facial recognition is your prime concern, and you're not already locked into a particular camera, have you considered megapixel cameras? You can then use onboard storage (with some of them), or storage directly to NAS (with some), no DVR required. Even at the lower end of the market, 1.2MP is four times the resolution of 4CIF/D1. 1/3" chip, at 25m, you'd need a 60mm lens to get a field of view 2m wide. 9mm is WAY too wide to give you usable detail at that range.

"Rankmount"? Is that the new smell-recording technology??

I've never worried about a ground strap when installing or handling ANY kind of camera... never had a problem with it yet.

Framerate has nothing to do with facial recognition - all a higher rate does is give you more individual shots to choose from, it won't make them any clearer. 4CIF will give you four times the resolution and far better detail than CIF; 5fps@4CIF is better than 30fps@CIF if you're concerned with seeing details.

I understand now! Well... this is the tricky part: even in bright sunlight, you're not likely to catch plates going by at 50mph, at least not without very specialized cameras. After dark, there's no chance unless you have a flash attached. At this point, we're talking technology like is used in speed and red-light cameras - you'd need a camera aimed and focused at a specific point on the road, a tight flash beam aimed at the same point, and something to trigger both when the car hits that point. The technology is out there, but I don't expect you'll find it for anywhere near your budget... would probably be cheaper to just let them bash the hell out of your Jeep, then go buy a new one

Welcome, from Western Canada!

In aid of keeping everything in one place, I'll reply to a couple things from your other post: Ask yourself this: WHY do you want to catch license plates, especially of cars going by at 50+? They're not the ones that are doing the damage. If people are walking onto your property and removing bits of your Jeep, they're certainly not jumping out of those cars going 50+! I'd concentrate more on focusing on the sidewalk - if they're walking up, it's likely they're either coming along the sidewalk, or parking along the sidewalk. I'd say this is a good idea, and the only thing that needs to change is your parking habits. In short, pick a spot where you can nose-in or back-in your Jeep, then position your cameras to give a clear view down each side of it. Not only will that then give you good images of the Jeep itself, but should also give a good view of anyone approaching it from almost any direction. The only catch then is, you have to remember to always park in that orientation, in approximately the same space In that case, you could put out some planters or something else innocuous to delineate your parking space without giving away that you're creating a choke-point for approaching would-be vandals.... in other words, make THEM come to YOU.

Are ALL the cameras 12V? How many are using Cat5 and baluns? It's not uncommon to get ground-loop issues when you combine baluns with cheap 12V cameras. The ability of one bad camera to bring down others is a common failing of the GeoVisions... and I suspect, anything else that uses the same ConeXant-based cards.