MaxIcon

The spec shows it as 83 degrees. Whether that's more accurate than the Dahua and Hik specs is hard to say without testing it, but I'd trust ACTi more for having good specs. http://www.acti.com/download_file/Product/datasheet/Datasheet_D72.pdf I measured my Hik bullet with a 4mm lens, and got these FOV numbers: 1080p: 68 degrees 3MP: 62 degrees Spec is 76 degrees, but they don't say which resolution it's for My Dahua 2100 with a 3.6mm lens: 720p and 1.3MP: 63 degrees Spec is 70 degrees

Yeah, it would be pretty much like a regular light shining in the same area. Like with a white light, if it's shining too directly toward the opposite camera, you'd get a lot of glare. It's unpredictable how the beam spreads; some fixtures have a flood effect, others have more of a spotlight, depending on what kind of LEDs they used. This is another thing you can't tell much about until you get it, since they don't say. Also, when vendors spec a distance range for IR LEDs, you can usually cut it in half to get a more realistic estimate.

MaxIcon, i got the manual on unbrick from dahua. I dont have the rs232 adapter yet, probably be in better hands if someone with Linux skill trying figuring this out before i should. http://www.sendspace.com/file/6i3m28 here is pdf, forum wont let me upload pdf here. thanks Thanks, that looks useful. It's a modified version of the one they have available for the DVRs, but the extra information looks valuable. I'll see if it helps things make more sense.

This would work just like a visible light in the same location would. It doesn't need to be where the camera is, but needs to illuminate the area the camera's looking at, just like a white light spotlight. Two things to consider: - You probably want to verify if the first one with the standard E27 connector is rated for outdoor use, unless your motion detect fixture is enclosed. It doesn't look outdoor rated to me. - 940 nm has much less red glow than 850 nm, but cameras are a good bit less sensitive to 940 nm, so for the same IR power output, you'll get less effective illumination, and it'll cost more as well. Nobody specs the actual IR output of their devices, which is not necessarily related to the LED power. That is, the 18W LED power of the 140 LED fixture doesn't mean it puts out 6 times the IR of the 3W LED power of the E27 120 LED lamp assembly. You can't tell until you test them, unless the vendor can tell you the actual IR output of the LEDs.

I looked at panoramic cams for the front of my house, and decided they were either too expensive or too proprietary (or both). I ended up using a pair of Hik mini-bullets aimed to give similar coverage - not so elegant, but effective and affordable. I've had issues catching raccoons on cam lately, but this depends on your motion detection. With Blue Iris, I've had to turn off object detect/reject to get it more sensitive to small critters (which increases my alerts from shadows and such). As long as they're in the main IR beam, it's no problem, but when they're around the fringe of the FOV or further away, it's tricky to get good triggering. I get around this by recording cams 24x7 as well as with motion detect, so when there's activity, I can always go back and see what happened when motion detect wasn't triggering.

Mine came with the cameras. Each cam had a bag with the matching connector parts in it. For spider webs, I use the Webster, which is soft enough to not move the camera and designed for spider webs. I've only got one floor, though. You can get them at Amazon.

There are generally 2 approaches for this: - Regular camera with a super-wide-angle lens - this is easier if you use one that takes CS or C mount, as fisheye lenses can really highlight quality issues. The fisheye distortion can be corrected either with the lens (expensive) or with image processing (less expensive, but has to be designed into the camera). The Vivotek FE8172 is an example of the latter. - Multi-sensor cameras, like the Arecont AV8185. These have better resolution across the entire FOV for similar MP counts, but aren't always compatible with general purpose NVRs or recording software. Some treat all the sensors as a single stream, others as multiple streams.

You can measure TVL on any camera. Getting an absolute TVL result is tricky, but getting a relative TVL result to compare between multiple cameras is fairly easy. Typically, 1 and 2MP cams will run somewhere on the order of 800-1200 TVL, but like with analog, it depends on many things. - Sensor resolution - Lens quality - Software processing and compression - Aside from the camera, there's DVR compression as well Generally, when you see an analog camera that says 600 tvl or whatever, they are referring to the capability of the sensor and processing chip, and don't take into account the lens quality, which is usually poor on inexpensive cameras. Marketing and difficulty of measuring actual TVL has made this a meaningless number.

I haven't tried this with cameras, but it worked with some industrial RGA controllers that I support at work. Take a laptop or PC that's not connected to a network and has Wireshark installed. Connect the camera to it, either directly with a 12V supply, or through a POE switch. Start Wireshark Select the wired network card under the Start icon Click Start There should be no network activity in the windows if there’s no outside network connection Power cycle the camera Wireshark should show a variety of messages, and one will be similar to one of these. The XXX address is the camera's current address: “Gratuitous ARP for XXX.XXX.XXX.XXX”. “Who has XXX.XXX.XXX.XXX? Tell 0.0.0.0”.

One solution is to put an inexpensive IR illuminator outside and tape over or disable your onboard IR LEDs. There are a number of them available on Amazon and ebay, and the Amazon models have the benefit of user reviews. This won't help if your tenant regulation prevent you from putting stuff up outside, but I've seen some clever solutions to this: - Make a bracket for the illuminator that you can hang from your windowsill overnight, then remove in the daytime. This would work for a camera, too. - Put the illuminator or camera in a birdhouse or flower bed so it's not obvious from the street. Having an illuminator on the outside still leaves it open to vandalism, but if your camera's on the inside, at least you'll get images of the perps.

Someone else said the NVR crops the image to 1080p resolution.

The big 80MB or so files have the full set of software, including the files needed for TFTP recovery. Hard to say what all the rest are for, but you really only need the correct .bin file for a normal upgrade. You can change the lens, but choosing the right one can be tricky. There are few standards for M12 lenses, and with day/night cams, many lenses will run into the IR filter before they're focused. You'd need to know the back focus length of the lens and the height of the IR filter to be sure this won't happen, and few vendors provide the back focus length. Likewise, different lenses may have a different field of view, even if they're the same size, so one 4mm lens may have a different FOV than a 4mm lens made by someone else. Most people buy one that looks like it might work and give it a try, but you can go through a few lenses this way. The real solution is for the camera manufacturers to offer lenses guaranteed to work with their cameras, and give the FOV specs with them. Only the higher end manufacturers do this currently.

I looked around for user feedback on these and didn't find any yet, but it turns out these are used for homebrew touch-panel displays. One of the vendors published a handy guide that talks about thinks like connectors and other details. It's not very applicable to surveillance lighting, but is worth reading for anyone thinking about these lights: http://cias.rit.edu/~nmtp/179773/mousetrap/infrared_led_guide.pdf

The latest version for that family of cams is 2013-05-17. Do you have PAL or NTSC? Here's a link to the full PAL version - don't use the download manager! viewtopic.php?p=223533#p223533 and a link to the update-only NTSC version (the clickable link isn't complete on my browser; you may need to copy and paste it): viewtopic.php?p=226171#p226171 These are the supported devices: Supported Devices: HDB(W)3202,HFW3101C,HF3101,HFW3X00C,F728WP,HF3300,HDB(W)3300,HFW3300,HF3301,HF3500,HF3500A,HF3200,HDB3200C,HF3110,HDB(W)3110,HDB(W)3101,HFW3110,HF3100,HF3211F,HDW2100,HD2100,HFW2100,HF2100,HFW3100,HDB3100C,HFW3(2)105(2100S),HDW3100S,HDW2105S,HDW3105S,HFW3200S,HFW3202C

If you need to stay under $100, your best bet is the Foscam/Agasio/etc clones. They have their limitations, but are very popular, and there are lots of resources on them out there, including forums dedicated to the different brands and Amazon reviews. The budget-minded users at the Blue Iris forum, cam-it.org, also use them a lot. You'll definitely want to do your homework, as some models work much better than other, very similar models. At least when you have problems, you'll have lots of company! If you can afford more, you'll be happier in the long term with the Hiks that Buellwinkle mentions, in terms of video quality, ease of setup, and reliability.

Yeah, the best bet for these is to spend some time fiddling with the settings, and take a snapshot of each combination. I always rename the file immediately to help me remember which one has which settings. I'm systematic about these things, so I'll start at the defaults, then move various settings in jumps of 25, which gives you a good feel for how they affect the image overall. After that, I'll set one that I care about (WDR or whatever) at a new setting, and play with the others to see how they combine. WDR really increases the low-light noise, so a lower setting like Buellwinkle uses might be better if you don't have deep shadows in the day or need lower night noise more. There's a ton of combinations, but you'll find that you can ignore some of them. Once you're done, go back and flip through all the captured images and it'll soon be clear which ones you prefer and which ones you need to test more and fine-tune. Be sure to post whatever you decide on!

If you run all the cams at 4096 kb/s, which is a reasonable bit rate for 1080p at 10-15 fps, and record full time, here's what you'd get: 4096 kb/s = 512 kB/s/camera 8 cameras = 4096 kB/s or 4.1 MB/sec 4.1 MB/sec = 246 MB/min = 14.8 GB/hr = 354 GB/day Assume 1.8 TB available on a 2 TB drive, you'd get about 5 days with 24 hr recording. Bumping to 4 TB would roughly double that. With motion detect, it would depend on how often motion was detected and how long the recording continued after detection. If you dropped the frame rate, you could get by with a lower bit rate, and would get more time.

I looked at a solution like this at one point, but didn't test it. They have pretty decent specs on their page, but unfortunately don't spec the IR output power, which is what counts. Input power and output power aren't reliably related on LEDs. Assuming a decent IR power level, what this would be good for is illuminating your walkway or house front evenly, without the hot spot (or long range throw) of an LED illuminator. You'd be trading off even, smooth lighting against the range and narrow FOV of a traditional illuminator. Best bet would be to get a strip and try it out.

The settings that worked well for me on my Hik bullets: WDR = 21 Noise reduction = 100 Contrast = 25 This lightened up the shadows in the daytime nicely without too much noise at night. The noise reduction softens the daytime shots a bit, but the trade-off for reducing the night noise is worth it to me. The IR looks pretty strong - it lights up the far end of the yard pretty well, but it's a spotlight. Most cameras do this, as you get better distance specs with a spot beam instead of a flood. With a zoom, it's especially problematic, since good coverage zoomed out means a lot of wasted IR when zoomed in. This could be fixed with optics (like a lens) for the IR that could zoom along with the camera lens, or LEDs that could be turned on and off for different area coverage. I don't know anyone doing either of these things in the value camera range.

OK, I used the 3.3V adapter I linked above to test this on the 3300C, and still no good. Like before, I get a lot of data during the boot, but it's all garble. It looks like what you get when the baud rate is wrong, but I've tried all the baud rates, plus the camera says it's set up for 115kB. Interestingly, when I connect the ground wire to the adapter, I don't get anything on the terminal. The RS232 adapter I used before was ok with the ground connected. Anyway, enough time chasing this for now. I'll have to think on it some more.

Yes, that's the one I believe is it, since it's the only un-connected connector when the camera is assembled, and it's on the CPU/memory board. The board above it is the power board, and there's limited access to that connector when the power board is in place. It's unlikely power comes in over the ribbon cable, so I'm assuming the CPU board is powered by the 10 pin connector between the two. Since this isn't on a cable, I can't see how to connect the two boards and still have access without a low-profile cable connected. I usually use clip leads for this when the pins are accessible. This is easy to get at on the 3300C: I believe this is the pinout for the 3300C, but would have to check my notes: Pin 1 - ground Pin 2 - Rx Pin 3 - Tx Pin 4 - 5Vdc

Here's a link to the NTSC version of the latest firmware. The readme for the PAL version says it support the HDB3100C. Note that the link in the post is broken by the forum software; you have to copy and paste the entire thing into your browser. viewtopic.php?p=226171#p226171

I haven't heard of anyone doing it on the Dahua yet, but it should be straightforward. My last attempt to figure it out on an HFW3300C was a waste of time because I was going straight from the TFTP connector to a serial port/USB adapter, and forgot that PC RS232 had a different voltage range than the camera's ports. As a result, I captured activity on the port, but none of it made sense... So, based on that, here's what you'd need. I'm a bit of a hacker on this stuff, so a lot of this is educated guesses, but I'm pretty sure I'm going down the right path. - TFTP server software - I use tftpd32 for managing my switches - A usb to 3.3V serial adapter like this: http://www.amazon.com/PL2303HX-RS232-Cable-module-Converter/dp/B008AGDTA4 - A full set of files for your cam, not just the .bin file. These are readily available and are usually around 80 MB. It wouldn't need to be the latest version, just a compatible version, since you'd be able to do an upgrade if you got it back online. The TFTP update would happen one of two ways: - Automatically checks for a TFTP server and updates the file with the name it's looking for if it's there (the Hik bullet does it this way) - Connects to the TFTP server and requires a manual update (the Dahua NVRs and my corporate switches do it this way) One problem with the 2100 is that the connector that I think is the right one is under another board, but all my work's been on the 3300. I got one of the 3.3V cables last week, and may have time to mess with it this weekend. Anyone with TFTP experience, please correct any mis-assumptions I've made!

Here's the Dahua HFW3300C varifocal lens, as a comparison. One connection is for the IR filter, which is part of the lens on this camera, and the other is the iris. Most lenses have the IR filter separate, but vendors seem to be moving towards integrated IR filters on varifocal lenses to save money, likely on the theory that the lenses are much less likely to be swapped. Finding a replacement for this would be tricky, as you'd have to match both the IR and auto-iris drive signals and connectors.

I've been working on a cheap and easy way to measure this on random lenses, and it's been elusive so far. Best bet would be to measure how high your IR filter is above the sensor, and get a lens with a BFL larger than this. Likewise, measuring the height of the filter above the PC board would give you a few extra mm for buffer. In theory, this would prevent it from hitting the IR filter when it's focused, and would only require the vendor to know the BFL for that lens.