MaxIcon

Most NVR software does most or all of those things one way or another, but the implementation and effectiveness varies dramatically from vendor to vendor. Best bet is to download some demos and start testing. It'll clarify the concepts a lot, and give you a much better idea of what's important to you in terms of software interfaces. Most software packages have demo or free light versions available.

That's a physical security vulnerability, vs a data security vulnerability. To tap into an analog cam, you have to physically be on location; to tap into a network backdoor, you just need network access, from anywhere in the world. Forward the camera port for external access, and the backdoor lets anyone in. A good rule of thumb: Once the adversary has physical access to your equipment, your security is compromised.

I've got a couple of Y-cam clones (LTS flashed with Y-cam firmware, virtually identical to the Y-cam), and they work well for me. One's in a Y-cam Shell outdoor enclosure and has been online for several years now, and one's an indoor cam that's used intermittently. They've both been very reliable and easy to use, and work fine with IE (and other browsers with some loss of functionality), Blue Iris, and the Aver NV6240 recording software. The main downsides are the 640x480 resolution (all my other cams are 1MP and better), the reliance on ActiveX for audio, and lack of an IR filter, which causes colors to be off in the daytime if there's a lot of IR present (like in outdoor shots). If you need PT, of course, they're no good, but the software's robust and support is good. I generally disable the audio, but sometimes forget when viewing through a browser window. The sounds of cars passing by reminds me. I've heard way too many horror stories about Foscam and its clones to ever want to go that way.

This is definitely reflected IR. There are typically 2 causes of this. Is this a dome or bullet? 1. Internal reflections from the glass or plastic cover that get in front of the lens. Domes are particularly prone to this. Most cameras have a foam or plastic sleeve that presses up against the cover and isolates the lens from the IR LEDs, but this is not always effective, especially if it's a dome that's tilted near the end of its travel. 2. External reflections from something close to the camera back to the clear cover or the sunshade, if it has one. These take re-positioning the camera or painting the sunshade (if that's what's reflecting). It's easy to tell which it is. If you take the camera and point it straight up at night, out in the open, you'll either have IR halo or not. If you do, it's coming from internal reflections. If not, it's reflecting off of something near the camera. As Owain said, one sure solution is to disable on-board IR (if you can, not all cams allow this) and use an external illuminator. Decent ones start around $30-40 and go up quickly from there.

Great info and screenshots! The two different 3.6mm lenses are almost certainly 1/4" (with the dark ring) and 1/3", since 3.6mm in the same size would give the same zoom factor, while 3.6mm in different sizes give different zoom factors. This is due to the light from the lens covering a different area around the sensor. If you took the pic with the dark corners and selected a square that excluded all the dark areas, it should look a lot closer to the 1/3" 3.6mm lens field of view. In any case, how much quality can you expect from a lens that costs a few dollars? The $14 lens is clearly the best of the lot, with much better detail even at the wider field of view, and is a pretty small investment for that big a quality increase. It would be interesting to see if the stock lenses are plastic or glass - that factory 6mm image is really not very good. I may try an upgrade on my Dahua with a stock 3.6mm lens - if it gives that big an improvement, it'll be great.

This is a problem in the jungle of board-cam lenses. You can't even assume that a $15 ebay lens isn't just a marked-up $3 lens. If image quality matters, best bet is to find a vendor who specializes in lenses and applications, like ragecams on ebay or m12lenses.com, and pay a few bucks more. My experience with inexpensive lenses has not been great so far, though I still buy them on a whim now and again. I'm a big fan of DX and have bought many things from them, but their quality control and spec dependability can be weak...

The plus of a PC based system is that you can change the NVR software relatively easily if one solution doesn't work for you. The dedicated NVR you mention will be maxed out from the outset, and if you want to move on, you'll need to start all over. PC systems generally take more fiddling with, but are much more flexible and configurable. Blue Iris with a recent I5 or I7 processor will handle the 7+MP you have listed without any trouble, and still have room for more, though it can start to bog down on 3+MP cams. There are lots of other NVR software packages that you could test demos of, as well. I have a 16 channel Aver NV6480 running about 24MP with 14 cams on an older i3-540 mini-ITX box, and it's handling it just fine. BI running on an identical dedicated box bogs down when I go much above 6MP, though my i5-2500 box has a lot more headroom for BI.

I mounted an old Vivotek dome camera directly to a square junction box. I just drilled holes in a junction box cover to match the mount holes and cable, used stainless screws and nuts to mount it, and screwed it back on to the box. I needed room for a POE to DC converter, and this worked fine. I only ran it a few months before switching to an eave-mounted bullet, but I could have mounted it on the junction box as well. My cams are all IP now, so no suggestions except that it's time to switch!

1.3mp is the best quality. You'll see that the aspect ratio changes as you select the different settings, try it. Some other have made mention that yes, you can use other Dahua cameras with this NVR. Just to clarify a bit, on the Dahua, the quality is the same for all the resolutions, but the field of view and number of pixels changes. 720p is 1280x720 pixels, or 0.92MP (usually just called 1MP). What Dahua calls 1.3MP is 1280x960, or 1.23MP. D1 is 704x480, or 0.34MP. All of them have the same image quality for a given codec and compression setting, but they show a different field of view. For example, if you're looking at something in the center of the image, it will look the same at 1.3MP or 720p. The higher resolution just extends what you can see to the right and left of center, while retaining the same basic image quality at the cost of more disk space, higher CPU usage, and higher bandwidth. If you need that extra field of view, you've got it; if not, you can reduce it as you wish. If anything, the lower resolution images will have higher quality at the same bandwidth settings.

Is the DX lens a 1/3" lens? The opening looks quite a bit smaller than the stock lens. That would explain the dark corners. This one here. http://dx.com/p/3-6mm-fixed-iris-lens-for-webcams-and-security-cctv-cameras-44101 It doesn't specify the size, which usually means 1/4". Generally, if it's 1/3" or more, they specify it, since it's an improvement. Like all specs on bottom-end lenses like these, there are no hard and fast rules. This one specifies 1/3": http://dx.com/p/replacement-fixed-iris-lens-for-cctv-camera-3-6mm-45189

Is the DX lens a 1/3" lens? The opening looks quite a bit smaller than the stock lens. That would explain the dark corners.

I don't think the 2100 series (1.3MP mini-bullets and domes) have uSD cards.

Unless you're buying a high-end camera, super-low lux ratings tend to be the triumph of marketing over technology. One problem with these super-low lux ratings is usually that they don't spec the exposure time and lens f-stop. I have cams that will show very nice images illuminated by a PC monitor, but they are running very slow frame rates (like 1/5 sec, which also gives 5 fps max) and will show nothing but blur if something's moving. If you want to capture pics of someone walking by that aren't completely blurred, you'll need 1/30 sec minimum exposure. This is also the minimum speed to get 30 fps, if that matters. To get decent images at these speeds with very low ambient lighting, you're probably looking at a 1/2" sensor with a fast lens. Some 1/3" systems do OK, but still need a fast lens. I'd find out the full specs on the camera first, otherwise, you're better off with more ambient lighting or IR.

No, too busy last weekend. I'm hoping to get to it this week.

Another plus of IP cams is that you can upgrade the camera resolution if you need to, while analog will always be stuck at its standard D1 resolution. That is, if you start with 1.3MP cams and decide you need 3MP in one position, this is a quick and easy switch if your recorder will support it. Changing NVR software is also relatively quick and easy. IP cams tend to have less low light capability than expensive analog cams. IR can help get around that, but there are pros and cons to IR as well. I've replaced all my analog cams with IP cams from 1MP to 3MP, all with built-in IR, and the images and coverage are much better. If you go IP, you should look into the trade-offs between dedicated NVRs and PC based systems. A quick search of this forum will turn up many discussions on this. If I wanted to set up a 4 camera system from scratch with that budget, I'd consider 4 Dahua HFW2100 1.3MP cams, a modern Intel i5 based PC, an inexpensive POE switch, and Blue Iris software. If BI isn't pro-grade enough, there are a lot of other options, though they tend to cost more. Most software NVRs have demo versions available to check out.

I'm running a Dahua HFW3300C and an HFW2100 on an Aver NV6240-16, using the latest version of their CM3000 software. The 2100 works great, while the 3300 has some minor issues that I believe are related to camera encoding, as I also have similar problems on a Blue Iris system. Both cameras are recognized and recorded, though.

This camera probably has a fixed IR filter between the lens and the sensor, since the sensor is similar to the IR-capable Dahua 3200 cameras. These fixed filters are usually easy to remove, and are typically small squares of glass. It may be part of the lens, but that's pretty unusual. I removed the fixed filter from one of my Areconts, and it was pretty quick and easy. Now I get good IR sensitivity (and focus shifts if I use a non-IR corrected lens), but also get color shifts during the day when there's a lot of IR present.

They get busy this time of year, and the delays get longer.

That's worth checking out. All my PCs run MSE these days, for comparison.

This should be covered in the contract agreement, which should specify the number of dead pixels that are acceptable. This is standard with laptops, monitors, etc - each vendor specifies how many dead pixels and what the status of them is before qualifying for replacement. Here's Dell's LCD policy as an example: http://support.dell.com/support/topics/global.aspx/support/kcs/document?docid=414288 Otherwise, it's a matter of how happy you want to make the customers and at what cost. For finding dead pixels, you want to look at a solid white image and a solid black image. For white, I usually point it at a light bulb at a slow exposure; for black, I just cover the lens. Some pixels will be stuck one color, others will be dead. Bright pixels are usually more annoying than dark ones. Note that dust on the sensor or lens can look like a dead pixel. If it's on the lens, rotating it will move the spot. If it's on the sensor, a blast of canned air will generally move it around. Personally, I find that a few dead pixels really don't matter in a real world application, and I just ignore them once I find I can't clean them off.

I'm using the primary streams. I have secondary disabled on the 3300 for troubleshooting, but never use them anyway. Distance shouldn't really matter much. One of my PCs has 4 switches and quite a bit of cable between it and the camera, and there's no appreciable extra lag. I'd be surprised if all the extra added a few tenths of a second. If you ping the camera, your delay should be well under 100 ms on a local network, even going through multiple switches. If your network is adding tens of seconds of delay, you have a serious network problem. Another easy test would be to connect a laptop to the HTPC network cable and see if it has little lag. If your other PCs have less than 2-3 seconds lag, it sounds like the problem is in your HTPC. It may be underpowered; if C2D means Core 2 Duo, that's a bit on the slow end for MP cams, though it should be able to handle one cam ok, if there's not a lot of other stuff going on in the background. It could also be the decoding in the PC end, if it's not a high priority task or if there's something else going on that takes resources, or if there's something in the HTPC that is juggling the signal (like unusual codecs). Not an easy thing to track down, unless you can use task manager or Process Explorer to figure out what's active while you're trying to decode the signal.

I just tested my HFW2100 and HFW3300C, and both had a lag of about 1.5 seconds. 45 seconds sounds pretty extreme to me. I can't imagine what would cause a delay like that unless your CPU is pegged out, which you'd probably have noticed. This was on the IE web interface, VLC, and Blue Iris. My Vivotek FD8134 is about 0.5 seconds faster than the Dahuas. My BI is on a dedicated PC, and I ran the VLC and the web interface on a utility PC; both are modest i3-540 boxes, one with 2G, one with 4G RAM, and Win7Embedded 32 bit. Is it the same on multiple PCs for you?

PM sent...

At the risk of sounding like a broken record, Dahua really needs to get their act together on firmware and customer support. All it takes is a little reading on their "support" board to figure that out. Apologies to b44kwz for going off topic... I also have bought a number of car parts from ebay. It's completely revolutionized support channels for older cars, especially quirky ones where parts are relatively unavailable, or expensive imports where dealer costs for new gear are quite high. Same with IP cams, PCs, you name it. I like to experiment with interesting technology, and ebay is an excellent way to buy and sell such stuff at low cost. I'd really be mad if I had bought my 5MP Areconts at full price, but I didn't. Buying them new wouldn't have made them perform any better. My Aver hybrid cards ran 1/4 the price of new ones, and work great. My 4 Vivoteks average half-price over new, and also work great. My 24 port POE switch cost a fraction of what it was new. I understand when people don't feel comfortable buying on ebay. It's a tricky place, and you can definitely get the short end of the stick if you're not careful. That doesn't mean you can't get good stuff there at a good price, though. YMMV, as always...

Because that's where I buy most of my cameras, and that's where people were buying them at the time. Again, I usually buy used, so ebay's the place for that. In general, I've had good luck, since most vendors provide support on their websites. In the past, that was one of the differentiating factors between the no-name imports and the name brands - manufacturer support on the website. Finding the camera you want by Dahua model number is not always easy, since the re-branding vendors call them something else, and google searches come up with lots of alibaba links. Apparently there are official vendors now selling Dahua branded cams, so it's not clear where their marketing team is going with this...