MaxIcon

To check auto-switching, I either turn off the lights in the garage as I set them up, or drape a towel over it. For manual switching, it's on the image settings tab. I believe the actual location changes depending on if you have profile switching turned on or not, but I'd have to check.

My US supplied, CCWR serialed Hiks came in plain brown boxes. The retail style box may be something a vendor can pay extra for if they want.

Assuming the camera is not completely bricked, you can recover it with TFTP, as mentioned. The solution to the Chinese country code issue is to apply the CBX fix. This fix appears to have a 100% success rate, based on user posts. All IP cams have their bugs and issues, but if you want to get good performance at a reasonable price, it's hard to go wrong with a Chinese Hik with the CBX fix applied. However, you do lose the local support you'd get from a local supplier, but this is why local suppliers cost more. This applies to warranty service as well as firmware support. Likewise, going with someone like Axis will get you better support at a higher price. The Chinese cam savings are big enough that a user buying, say, 6 cams will be better off buying a spare for replacing a dead cam than trying to deal with warranty service. This is for buying from Ali, not ebay, where the middleman eats up some of the savings. Unfortunately, Hik is battling with the grey market resellers, and the end user can be the one who loses out, as you've found. I'm not sure what they're thinking.

Agreed. When I was testing cams a while back, I mounted them to wood planks, which I could move around easily to different locations. In this case, you could mount it directly to a piece of wood with a hole drilled for the cables, and mount a junction box on the back of the plank to protect the cables if you wanted. Either way, mounting the plank to the railing would be easy.

You can also put it into night mode manually just to check that they're coming on at all.

Hard to say without knowing more about the cameras and NVR. The cams are 3x zoom with auto-focus, which can be handy and adds more to the cost. NVR costs depend a lot on things like maximum bandwidth, and this one has high power POE, which also adds value. Without knowing more, it sounds like a piece of quality gear, but that could all be marketing. I'm not sure anyone figured out who makes the AvertX gear originally.

You can use uncrimped connectors, or cover the ports with tape.

You can zoom in to get the full resolution. On most cams, the difference between 1080p and 3MP isn't resolution in terms of pixels per foot, but in the area covered. Likewise, the cost difference between 3MP and 1080P is usually negligible. That is, 3MP gives more pixels horizontally and vertically for more degrees of field of view, but you don't get any more pixels per foot, just more feet of coverage. So, for the part of the image that's the same in 1080p and 3MP, the resolution is exactly the same. Hik 2xxx series cams are different; the 3MP image is scaled and gives a smaller horizontal FOV than 1080p, but most cameras don't do this. That said, most affordable NVRs only support 1080p viewing, regardless of the recorded resolution. To get more, you'll need to go with a PC and a higher resolution monitor.

I have an Aver NV6480 box (same basic software as NV5000) as well as a Blue Iris box, and have never been able to get BI to talk to it. Aside from capture cards, you can look at 4 channel encoders, like Grandstream or Axis. You might want to go to the BI community support forum at cam-it.org and search on terms like encoder, capture, and grandstream. Once you're read what's there, post any questions, as there are some people there using both encoders and capture cards.

If you use fixed bit rate, the bit rate determines how much space recordings take. Resolution, frame rate, and quality don't matter unless you use variable bit rate. For example: 4Mbps (bits)/8 = 0.5MBps (bytes) 0.5MBps * 60 = 30MB per minute 30MBpm * 60 = 1.8GB per hour So, 32GB would hold about 18 hours at 4Mbps fixed. For variable bit rate, it depends on the level of visual detail, quality setting, and amount of motion in the image.

Use the CBX fix! viewtopic.php?f=19&t=41225

This was probably written when 8fps was the max for 5MP cams. What's your current bitrate setting and frame rate? I'd suggest increasing bit rate until the the closer areas get more detail. The bricks on the next building back should improve as well. The bricks in the midrange look decent, because they're high contrast, and the firmware decided to compress those pixel groups less. The lower contrast areas that are closer to the cam are compressed more, so those should show improvement as you increase the bit rate if that's the problem. Based on the ugliness of these images, I'd double the bit rate for starters. If you're using VBR, be sure image quality is set at the highest setting. If the bit rate ends up too high for your needs, you'll need to reduce the frame rate or increase the i-frame setting. You should also play with the other settings, as BW suggested.

I agree. Using standard mount holes opens up a huge range of mount possibilities, and should only add a few dollars to the cost, which I'd gladly pay. I have several older cams mounted on alternative mounts in difficult positions.

I'd agree. What are your bitrate and frame rate settings? It does OK on some stuff, like the brick wall, but other areas are seriously over-compressed. Hik has a PDF of recommended bit rates for different resolutions, quality, and frame rates that's a pretty good general guideline: ftp://hikfirmware:Hikvision123@ftp.hikvisionusa.com/Technical%20Bulletin/2014/IP%20Camera%20Recommended%20Bitrate%20v2.0_20140219.pdf

2.1 x 5.5mm is the de facto power plug standard for CCTV equipment, though some vendors (notably Foscam) use different size plugs. You won't know for sure until you try, but this is the size that all my cams and IR illuminators use.

Also, if you use cat 6 cables, you'll have less power drop than cat 5 due to thicker wires. This is the real issue with POE. If your cams have low current draw (most don't draw the 803.2af maximum), you can go further with no issues until you run into the 300' ethernet limitation.

The lens length will depend on the distance you need to read plates at and the conditions. You can't get much more than 25mm on an M12 lens camera, especially bullet cams. Longer lenses are physically very long, and the f-stop drops at longer lengths. What you need is a certain number of pixels per foot. In the US, a car plate is 1' wide, and if you measure the character stroke width, you'll find you need 40 pixels per foot if everything is ideal. Everything's not ideal, of course, so you'd want to go to 60 ppf for overhead. At night, you'll need more, and you'll need to worry about exposure time, as BW mentioned. You'll find plate illumination varies quite a lot, including lights off, plate light on, taillights vs brakelights on for rear plate, and headlights on or off for front plates. Also, those numbers hold if you're getting the image straight on. If it's at an angle, like if your watching the street from the front of your house, you need even more resolution, as the virtual plate width is lower. This also extends the distance you need to be able to read at. BW's post above looks great at 45', because he's getting 90 pixels per foot. If you can get this in a straight-on capture, you should have few problems. I'm testing a similar setup, using 1080p for more pixels per foot on a Hik 2032 with a 1/2" 25mm lens, and at 100' and about a 30 degree angle, I can read plates fine in the daytime, but as soon as the lighting drops, the resolution fades. At night, they're unreadable on the Hik, because its low light image quality isn't good at the exposure times you need to avoid image blur and avoid plate washout. A cam with better low light performance would help, but if you want to do it with a high likelihood of success, you'll need some dedicated hardware. BW has some good posts on how to do serious LPR with serious hardware.

An interesting, if non-technical, article about CV dazzle - facial camouflage that attempts to confuse face-recognition software in surveillance systems: http://www.theatlantic.com/features/archive/2014/07/makeup/374929/ So, the cameras may not notice you, but the people around definitely do! William Gibson fans, and surveillance hackers in general, will already be aware of this, but it's going more mainstream as user awareness of surveillance increases. Here's a site with a more technical approach, including info on the algorithms used: http://cvdazzle.com/

You should be able to tftp a good version on it. iTuneDVR has a simple procedure here: viewtopic.php?f=19&t=42028 and there's a more detailed version, and software for it, on the Hik FTP sites.

In 5.1.2 firmware notes, Hik said Chinese cams would no longer work with IVMS-4200. You should check with CBX to see if the region change will fix this, but I'd guess it will.

There are some 4300 links here: viewtopic.php?p=246998#p246998 You didn't say if you wanted a specific version, and I'm not sure what the V2 means. As always, be cautious with Dahua firmware upgrades, as they can brick your cam if they don't match. If you read further in that thread, there are some reviews on the firmware version listed.

I got my 25mm 1/2" lens and mounted it. It was quite a pain, and in general, I wouldn't recommend the Hik 2032 for lens conversions. I'll post some screen shots and such later, but here's the overview: This lens should give a longer zoom than a 25mm 1/3" lens. I made the FOV measurements but haven't crunched the numbers yet. - The lens is a little too long at the focus point, and I had to replace the sensor board screws with slightly longer ones to re-mount it. These are M1.0 screws, and I keep a bag full of tiny screws scavenged from old hard drives and such for just this reason (or repairing Macs). - Depth of field is not great, as mentioned before. The longest distance I can focus from the bench is about 50', so that's what I used, but the distance to my plate capture area is about 100'. Since the Hik can only be focused when completely disassembled, this is a hard thing to do, and may have affected my results. - The Hik captures images fine at this distance if the lighting is good, like direct sun on the plate. In the day, if the plate is in shadow, it starts to blur, even at high shutter speeds. At night, if everything is perfect, you can get a readable plate image, but if anything's not right, it's a blur. Whether this would be helped with a 100' focus instead of a 50' focus remains to be seen. I've got a Dahua 4300S that's easier to deal with lens adjustments on, and I'll probably switch over to it for further testing. In the meantime, I turned on ROI for the plate area, and will see if that helps.

Mine were bought from a vendor who can't be named, and have different model/serial numbers than the OEM Hik cams, but the US firmware loads on them fine. Here's an example: Model: NC-KBU32 US SN: NC-KBU3220130607CCWR42423xxxx From above: Ch SN: DS-2CD2032-I20140515CCCH464808456 Here's the SN breakdown as I see it: NC-KBU32 or DS-2CD2032-I - model # 20130607 or 20140515 - manufacture (or initial code burn) date CC - ?, same on both WR or CH - region code 42423xxxx or 464808456 - serial # ETA: Here's another SN from the Swann version of this camera: Model SWNHD-820CAM Serial No. SWNHD-820CAM20130204CCWR41681xxxx - also has the WR region code.

Does Milestone display in Mb or MB/sec? Some packages, like Blue Iris, use MB instead of Mb, which often causes confusion. In any case, 12Mbps is some pretty serious overkill (ETA: I see you were only running 8Mbps before), and will cause your files to be much larger than necessary. I haven't seen a case yet where a Hik 1080p image improves at more than 4Mbps, but I have only tested at 10 and 15 fps. I don't run 3MP because of the smaller HFOV. Hik recommends 5120 kbps at 3MP and 15 fps for best quality. Even at 30 fps, they recommend 10240 kbps for best quality.

"Suspect identified in ski slope bear attacks - news at 11!"