|
After some interest in the Watch Dogs. thread in games, I figured I'd start a thread for those interested in knowing more about security. First, a little about myself: I design security systems for a living. I focus on physical security (cameras, door locks, badge readers, biometrics, etc.), but I also handle quite a bit of network security (firewalls, IPS, two-factor authentication, etc.). I've designed both very large (2,000+ camera), and very small (local gas station) systems, and I specialize in IP integration (putting cameras and other security devices on the network). Have a question about what kind of system to get for your house? Want to know if last night's episode of CSI was based on real technology or total bullshit? (It's probably bullshit.) Wondering how a camera works? Don't know the difference between IP and analog? Curious about just how much big brother is watching you? Ask away. DISCLAIMER: I am not a lawyer. There are lots of privacy issues regarding surveillance in particular and while I may know enough about the subject to offer some insight, I am not qualified to give legal advice. Nothing I say should be taken as such. My general rule for designing surveillance systems is "Don't be creepy." If it's creepy, it may be illegal; check with a lawyer. KillHour fucked around with this message at 01:00 on May 1, 2013 |
#
¿
May 1, 2013 00:54
|
|
|
|
| # ¿ Apr 23, 2024 22:25 |
|
|
Cameras in 5 Minutes - A Primer All cameras, from the most expensive movie studio camera to the one in your cellphone are variations on the original pinhole camera.  A pinhole camera is extremely simple. It selectively allows individual light rays through in such a way that they form a coherent image on the other side. This works because all of the light rays going through the pinhole are focused through a single point, which means that if they are all of the same length, they describe a sphere. Each light ray can be described as the coordinate on this imaginary sphere it comes from, and will cast an exact copy on the other side (except smaller, upside down, and backwards). The quality of a pinhole camera can be increased by making the pinhole smaller, and therefore being more selective in what light rays you let through. This leads to two problems: less light means a dimmer image, and there is a physical limit to how small the hole can be before physics makes things break in horrible ways (This mostly has to do with diffraction and wave-particle duality, which I won't get into here). These both can be sidestepped by using carefully cut pieces of glass or plastic (called a lens) to refract the light through an imaginary "pinhole". This is called a focal point.  The accuracy of the lens determines how sharp of an image can be captured, so more expensive lenses are needed for higher resolution cameras. (Incidentally, this is why an 8+ megapixel camera on a cellphone with a tiny plastic lens captures terrible images compared to a 3 megapixel DSLR with a proper lens. More advanced lenses can adjust individual pieces forward and back to make the field of view larger or smaller by moving the focal point closer to or further from the sensor. These are called varifocal lenses, and are made of several smaller lenses. After the image is focused, it is projected onto a sensor. For security cameras, this can be a CCD (Charge Coupled Device) or a CMOS (Complimentary Metal-Oxide Semiconductor). CCDs use an older technology and are expensive to scale up to higher resolutions, while CMOSs are more susceptible to noise, but lend themselves well to high pixel densities and consume much less power. Most IP cameras will use CMOS sensors, and CCD sensors will be found primarily on analog cameras. KillHour fucked around with this message at 01:46 on May 1, 2013 |
|
#
¿
May 1, 2013 01:37
|
|
|
1: What is this system primarily being used for? (Are you trying to deter thieves? Have footage to give to police if there's a break in? Find out which neighbor's dog is digging up your flower bed?) 2: What is your budget? What is the value of the property you are trying to protect? 3: How long would you like to keep recordings for? 4: How dark does it get at night? 5: Do you leave your computer on 24x7? 6: How large of an area are you trying to cover? (Is your front yard 30 ft wide or 300 ft wide?) KillHour fucked around with this message at 02:03 on May 1, 2013 |
|
#
¿
May 1, 2013 02:01
|
|
|
Absolutely. I have a fair bit of networking experience myself, but a second set of eyes is always welcome (Get 10 network engineers in a room and the only thing they'll agree on is that you need a larger budget). My weak spot is really alarms/fire systems, actually, so you'll probably complement me well. As for the privacy concerns, that's going to get very interesting in the coming years. As far as I know, most laws regarding privacy (especially recording audio as part of surveillance) haven't been thoroughly tested in court. I'm a very pro-privacy person, so too much surveillance creeps me out (and I'm always aware of it, since I know where to look). That being said, it's important that we feel safe when we're out, and surveillance can absolutely play a huge part in that. I live in Buffalo, and the city put up over 100 cameras to deter and monitor crime, and I think they did a great job with it. The cameras are very overt, and don't really seem like a privacy concern. I'd probably feel differently if they put one outside my bedroom window, though. For smaller installs, I like to recommend QNAP NVRs, since smaller environments don't usually have servers that can handle the load. For very small installs, you can't beat the Toshiba ESV16's price. When we start getting into the medium-sized solutions (24-64 cameras), I like to recommend Exacq or Lensec. KillHour fucked around with this message at 02:14 on May 1, 2013 |
|
#
¿
May 1, 2013 02:09
|
|
|
$1000 budget for 6 cameras is going to put you squarely in the "I picked this up from Sam's Club" category. Cameras like that are typically not going to tell you WHO stole your stuff, just that he was wearing black. Or was it blue? Maybe purple... In order to get a rough idea of resolution needed to identify someone's face, take the horizontal size in ft of the area you're trying to cover and multiply by 150. That's the minimum horizontal resolution for a camera to cover that area. For instance: 10ft * 150px/ft = 1500px, which would require a 1080p (1920x1080) camera to cover. If you're only worried about getting faces at your door, this would be more than sufficient: http://www.newegg.com/Product/Product.aspx?Item=81-122-032&ParentOnly=1&IsVirtualParent=1 If you're worried about someone breaking in through a window, I'd recommend putting at least 1 camera inside your house at a choke point that they're likely to pass through (a doorway or staircase). Indoor cameras are less expensive, and generally need to cover less area.
|
|
#
¿
May 1, 2013 02:42
|
|
|
There's a few ways to accomplish that. The most straightforward (but not necessarily the cheapsest) way is probably something like this: This camera: http://www.axis.com/products/cam_m2014e/ Plus this decoder: http://www.axis.com/products/cam_p7701/index.htm Plus a small PoE switch Plus an HDMI/DVI monitor Alternatively, any IP camera can be viewed with a tablet or laptop (or Chromebook) from the same network through a web browser or app. Pick a cheap one and create a shortcut to the IP. Edit: If you skip to 4:30 in this video, you can see the quality of a typical analog system you can pick up for less than 1k. https://www.youtube.com/watch?v=Ukpvn1BijzU You're not going to identify anyone in the video, but you'll at least know what happened. KillHour fucked around with this message at 03:05 on May 1, 2013 |
|
#
¿
May 1, 2013 02:53
|
|
|
I haven't installed too many biometrics systems other than testing for compatibility (I do design work, and don't have much installation exposure day-to-day), but I have a general understanding of how they are installed. Most access control systems (including biometrics and card readers) are tied to a central controller using structured cabling. This controller is what makes decisions regarding authorization and what actions to take. A controller can typically handle between 1 and 8 doors, and multiple controllers can usually be connected, either through a serial interface or through the network. If you want, I can go over the specific protocols these devices use to talk to each other. As for the specific types of biometrics, here are a few of the more common ones: Fingerprint - These can be either optical (takes a picture of your finger) or capacitive (works the same way as a touch screen). Optical readers are generally more accurate and aren't affected by temperature, and capacitive readers are less expensive. Capactitive readers are generally found on laptops. Handprint - Same thing, but larger. Generally optical. Hand Geometry - These measure the shape of your palm, and compare specific ratios (length to width, finger lengths relative to each other, etc.). Disney World used to use these before they went to fingerprint. Vascular Pattern Recognition - These use infrared light to take a picture of your hand. The twist is that skin is transparent to IR light, but the hemoglobin in your blood isn't. This essentially leads to an "X-Ray" type image, except showing blood vessels instead of bone. Retinal Scan - These work similarly to VPR, but project a beam of IR light into your eye to read the blood vessels on the back of your eye instead of in your hand. Vendors are very careful to point out that this beam is not a laser, but there isn't much practical difference. Very low (effectively 0) rate of false positives, but expensive and invasive. Not to be confused with Iris Scan. Iris Scan - These take a picture of your iris (the colored portion of your eye), just like an optical fingerprint reader. Very simple for the user, and non invasive. Some Iris Scanners can read irises from several feet away as a person is walking, so they don't have to stop at a terminal to be scanned. Not to be confused with Retinal Scan, and vendors will point that out about 3 times every 5 minutes. Other types exist, but are far less common. Let me know if you want me to go into more detail about anything. KillHour fucked around with this message at 19:43 on May 1, 2013 |
|
#
¿
May 1, 2013 19:37
|
|
|
Ah, yes. I had completely forgotten about Ubiquiti. It's a company we don't carry (but are working on getting), but I've heard good things.Solomonic posted:In your opinion, what's the most common mistake people make when buying/setting up/maintaining/etc. surveillance networks? For someone who's unfamiliar with cameras, it has to be infrastructure. People think that storage, power, networking and installation are all free, and don't plan properly for the costs. They also tend to think that all cameras are more or less the same (and also magic), which leads to people demanding that I build a system with no more information than "8 outdoor cameras, 1 month recording, $2000 budget" and then proceed to get mad when I tell them that I need more information. And more money. (Not picking on Cosmic Charlie, here. These are people with real businesses and real high expectations that should have real budgets to go along with those.) TATPants posted:I've got a question about the practicalities of the 'walking retina' scanner. In order for this to work well, wouldn't you would need an insanely high resolution camera or one with great magnification and an equally high quality tracking/positioning system to take a (let's say) 100x100 px image of an eye? Plus all the crazy fast calculations the computer would have to make in ~3s...this system has to be ridiculously expensive. I would really appreciate any more details about this if you have the time - it seems fascinating They're obscenely expensive and bleeding edge. I don't have a ton of information about them, since I've never worked with anyone that wanted to plunk down for one. Here's an example to oogle at, though: http://www.sri.com/engage/products-solutions/iom-passport-portal-system Edit: They have one that will capture your iris in a drive through.  http://www.sri.com/engage/products-solutions/iom-pass-thru-drive-up-iris-recognition-system KillHour fucked around with this message at 03:52 on May 2, 2013 |
|
#
¿
May 2, 2013 03:32
|
|
|
Not an Anthem posted:I have a Synology NAS that supports IP Camera capture/logging and remote interface. I don't have a need for a security camera, but it was more of a "oh, I can do this now, maybe it'd be fun to play with" for either identifying people outside my building or just for security. Cameras are actually far better as deterrents than they are as reactive security measures. The reason for this is simply that people greatly overestimate the capabilities of a camera. People assume that cameras are all like the cameras they have on CSI, and have 11 billion megapixels and magic "identify people from their reflection in a window" software. Even if you know that most cameras are junk, it's still extremely hard to tell how good a camera is just by looking at it. That, and 90% of people would rather stop a crime than solve one, so the relative value of a camera as a deterrent is greater even if it's equally competent at both. Besides, what's the first thing you do when you notice a camera looking at you? You probably look straight at it for 2-3 seconds. Are you going to break into that house after you did that? As an anecdote, I had a customer that wanted to return some cameras I sold him and replace them with more discrete cameras. Apparently, when he installed them, people stopped vandalizing and stealing from his store. He didn't want to deter them from doing so, he wanted to catch them. Edit: Although, if the thief is smart/gutsy, he may just try to steal your DVR instead. Make sure that's in a lockbox. KillHour fucked around with this message at 18:03 on May 2, 2013 |
|
#
¿
May 2, 2013 17:49
|
|
|
Agatha Crispies posted:Thank you for responding so thoroughly.  Spergy Post Ahead!Wikipedia has some good diagrams for this, but I can't access image hosting sites from work, so I'll try to explain it as best I can. The "brains" of an access control system is the controller. A controller pretty much looks like a bare circuit board. Sometimes they're in a cabinet, but most of the time they're mounted directly to a wood board on the wall. This GIS should give you a decent idea: https://www.google.com/search?q=acc...iw=1920&bih=955 These boards connect to devices and each other with phoenix connectors (screw down terminals), so structured cabling (Think Cat5 without the ends) is typically used. The first connection you need to think about is the one from the reader to the controller. A reader has no direct control over the door. Instead of sending an "open door" command, the reader just repeats the information it was given to the controller, and the controller makes all the decisions. It's similar to a restaurant - the waiter takes your order, and the cook decides what do to with it. There are several different types of connection between the reader and controller, but 3 are the most popular: Wiegand - By far the most popular, this is the original connection type for readers. The term Wiegand is actually very vague, since it used to describe an entire type of access control system. Now it's typically used to describe one of three things - the connection between the reader and the controller, the data format stored on the card, and the transmission protocol between the card and reader. Of these, only the first two are still in common use. The way Wiegand works is actually very simple. It requires a minimum of 4 wires - D0, D1, +12VDC and Ground. In order to send a message, the system will send pulses (or bits) down either the D0 or D1 wire. A pulse down the D0 wire is the binary equivalent of 0, and a pulse down the D1 wire is the binary equivalent of 1. So in order to send the following message: 0110, you would send pulses down the wires in the following order: D0, D1, D1, D0. This eliminates the need for a common clock between the devices and predefined "header" and "footer" messages. Clock and Data - C&D is very similar to Wiegand, except that instead of D0 and D1 wires, it has C(lock) and D(ata) wires. The C wire carries a pulse at a set interval. If the D wire has a pulse on it at the same time, the controller considers it a 1, else it is a 0. So, to send the same 0110 message, you would need to send the following: C, C+D, C+D, C. RS485 - RS485 is a serial communication standard that is far more complex than either Wiegand or C&D. The advantage of this is that unlike the other two, RS485 is bidirectional, and can send information both from the reader to the controller and from the controller to the reader. Wiegand and C&D can only send data from the reader to the controller. The downside is that these systems are typically proprietary, and readers and controllers of different brands will not talk to each other, while all Wiegand/C&D readers will work with all Wiegand/C&D controllers. I'll get into communication between card and reader, card formats and inter-controller communications in another post.
|
|
#
¿
May 3, 2013 22:29
|
|
|
Sorry I haven't been updating the thread. Just got back from vacation. I'll go deeper into card communication tomorrow, but to answer your question quickly: Looking at the EM4102, that looks like a 125khz card. http://www.trossenrobotics.com/store/c/3433-EM4102-125-kHz-RFID-Tags.aspx Any 125khz reader should work with any 125khz card unless the reader has a controller built into it. If the reader has an integrated controller, the controller needs to be able to understand the format of the card. You can think of it like a bar code - any 2d barcode reader can read any 2d barcode and spit out binary, but the backend system needs to be able to parse the data. The short answer is likely that he HiTag2 system is designed to be more open than the other system you have in place.
|
|
#
¿
May 27, 2013 03:32
|
|
|
Warm und Fuzzy posted:When are we going to start seeing 3D laser scanning technology in security systems? I have to imagine there are prototypes in place somewhere. If not, do you want to create a startup? It's possible to do a 3d scan of a person, but the resolution required to actually identify someone coupled with a computer powerful enough to do it would probably be too expensive to consider. Current techniques work very well for the majority of use cases. If I was going to make a startup, it would be using commercial grade security cameras as video conferencing systems. Seriously, you can get an HD PTZ IP camera for $2000 bucks, but an HD codec + camera from Cisco or Polycom costs 10-20 grand on the low end, and the protocols are all the same (H.264, G.711). Warm und Fuzzy posted:Not for access control, but to replace ptz cameras. Think an Xbox Kinect. Having a 3D point cloud of a perpetrator would give you extremely precise data, like height, shoe size, walking gate, etc. The range for something like this isn't good enough. They'd need to be less than 10 ft or so from the sensors, and it would make more sense just to take a picture of their face than try to identify them by gait. Laser rangefinders attached to a camera can pinpoint their exact position, but that's generally only needed if you're trying to shoot them.  themrguy posted:Do those gadgets to spoof security card readers/biometrics/etc exist outside of movies or the inventories of places like the CIA? To put it another way, would a competent professional criminal (someone who may not be Jason Bourne, but has more equipment and professionalism than a crackhead or delinquent youth) have access to stuff like that? Or would they be more likely to say "gently caress it" and kick the door down/hold a gun to someone's head? Kind of, sort of, not really. Some headway was made in cloning DESFire RFID cards awhile back, but that hole was plugged. RFID is like encryption - implemented properly, it's mathematically impossible to crack in a reasonable time frame. 125khz cards are NOT secure for reasons I will go into in my next post. Don't use them for real security. As for biometric spoofing - it's probably possible, but can be very difficult. Most modern systems have anti-spoofing mechanisms to help detect forgeries, and getting someone to stand still while you take an imprint of their finger isn't exactly easy. Also, the CIA doesn't have anything you can't get on the open market with enough money. When a 3 letter agency wants to do something they saw in a movie, they generally call me to ask if it's possible. General rule: With access control, it usually is. With surveillance, it almost always isn't. It may sound counter intuitive, but access control is leaps and bounds ahead of surveillance in terms of capabilities. As a species, we've been making locks for centuries, after all. I'd stick with kicking the door down, personally. (Unless they have this door: https://www.youtube.com/watch?v=FD2paju93aI. Maybe try breaking in through a window? https://www.youtube.com/watch?v=nRtDc2_SI40) Gromit posted:Thanks for the info, but you need a free hand to operate your car as well and keyless entry seems pretty popular with those. I would also get endless amusement from my front door going "boop boop" and having the porch-lights flash when I unlock it. How about with an app on your phone? http://www.crestron.com/about/press_room/press_releases/show_release.asp?press_release_id=1780  Hope you have 5 figures lying around. KillHour fucked around with this message at 06:21 on May 28, 2013 |
|
#
¿
May 28, 2013 05:54
|
|
|
Kemper Boyd posted:About surveillance, there's lots of things that you can do to improve your (kinda fuzzy on the english-language word here) metrics i.e. what data you get out from your equipment. Like magtriggers on every door, pressure plates, that whole 3d laser scanning thing and so on. Exactly. A good term for this is "project scope". The first part of every design is to figure out why the customer needs a security system in the first place. I can talk all day about something that's really cool on a conceptual basis, but I won't sell any of them if it won't help accomplish a customer's goals.
|
|
#
¿
May 28, 2013 12:56
|
|
|
Gromit posted:Urgh, that looks like poo poo, with a price to match. How can they even justify five figures for something as simple as that? It's not five figures for the lock, it's five figures for the back-end system the lock integrates with. Crestron is for rich people who can't stand the idea of their home theater NOT having lights that automatically dim when you press play. (It does a bunch of other cool stuff that makes it totally worth it if you can afford it and go all-out, but that's definitely the target market.) This might be more along the lines of what you're looking for: http://www.smarthome.com/51890/MiLocks-RF-01P-RF-Remote-Controlled-Keyless-Entry-Deadbolt-Brass/p.aspx Edit: I promise I'll stop procrastinating on writing up the rest of the access control stuff. I've been really busy looking for a new house. Hopefully I'll have 15 minutes tomorrow. KillHour fucked around with this message at 02:55 on Jun 3, 2013 |
|
#
¿
Jun 3, 2013 02:46
|
|
|
themrguy posted:Interesting answers, thanks. I'm kind of confused as to why anyone would want to do this. Technically speaking, any form of trap is physical security (landmines, for instance), but typically you wouldn't want to hurt people, just keep them out. The only exception to this in the commercial space (outside of research facilities that can and do use lethal force) is probably a man trap, but that just keeps them there until they can be arrested (or let out in the case of a false positive). I actually do a lot of systems for power substations, and the alarm goes to both the police and the local hospital, since one of their big goals for the system was to keep people from getting hurt if they break in. If you really want to shoot bullets/teargas/beanbags at intruders, this is your guy: http://en.wikipedia.org/wiki/Samsung_SGR-A1 Slanderer posted:EDIT: Nowadays, you'd want to use an encoder to bring them on the network and treat them like IP cameras. You could probably do a no-frills encoding solution for around $100/channel. I still get requests for big matrix switchers and multiplexers on occasion, but I don't even care to stock multiplexers larger than a 4 channel. Those builds are a nightmare. KillHour fucked around with this message at 14:34 on Jun 4, 2013 |
|
#
¿
Jun 4, 2013 14:23
|
|
|
JohnnyRnR posted:In function it's similar to the automated smoke systems that discharge to obscure the area being attacked. Those are being widely sold. http://www.bandituk.co.uk/ Pepper spray or a fog machine isn't nearly on the same level as tear gas. I have a feeling that that pepper spray system could open you up to a lawsuit, as well.
|
|
#
¿
Jun 4, 2013 23:52
|
|
|
Look Ma, I stopped procrastinating! (Not really, I'm supposed to be doing laundry) RFID cards part II: The part of an RFID system that gives it its name is the communication between the card/tag and the reader. This communication typically happens on 1 of 2 frequencies: 125khz and 13.56mhz. A 125kHz card, also known as a "Prox" card, is extremely simple: The reader puts out a wireless pulse several times per second (5 is common), which is picked up by the antenna embedded in the card.  You can tell by the shape of the antenna that this is a 13.56mhz card, but prox looks similar, just with a bigger antenna This pulse is enough to power the chip for the fraction of a second it needs to transmit the stored credential over the same antenna, which is then picked up by the reader. For Prox cards, this information is in plain text, which is then sent to the controller to match it up against the database. For this reason, 125khz cards are considered inherently insecure. Anyone can clone a card just by getting a reader within a 6-15 inches of your wallet. 13.56mhz cards come in several different formats, but the general idea is the same. Before the card sends the data, there is a mutual authentication step, which works like this: Reader sends card a pseudo-random number. Card responds with an "answer" based on a cryptographic algorithm (each vendor uses a different protocol for this), along with a different pseudo-random number. The reader will then send the answer for the second problem back to the card, and the card will respond with the authentication information (still in plain text) back to the reader. The authentication information is secure, even if it is in plain text, because the card will refuse to send the information if the reader does not correctly authenticate. Likewise, the reader will refuse to accept the information if the card does not correctly authenticate. All 125khz cards and readers should be compatible, because the information is in plain text with no authentication. 13.56mhz cards need to match up with their readers (iClass with iClass, MiFare with MiFare, DESFire with DESFire, etc.), because the authentication mechanisms need to match. There used to be another communication method (confusingly called Wiegand) but it's not commonly used any more, unlike the other 2 Wiegands. Still with me? Good. Now, when the reader gets the information from the card, it doesn't know what to do with it. It has to pass the information down to the controller. Unlike with the reader, the information needs to be on the card in a format the controller can understand. This is a bit tricky, since there are literally 10's of thousands of different formats out there. Formats themselves are pretty simple, however. They are a series of binary numbers, generally between 26 and 64 bits long. Here is an example of the most common format (again, confusingly called Wiegand): PFFFFFFFFNNNNNNNNNNNNNNNNP Each letter is either a 0 or a 1, meaning Wiegand is a 26 bit format. The F's are what is called a Facility Code. In 26 bit Wiegand format, the facility code is a number between 0 and 255. This number is most often used to distinguish what location the card originated at (For instance, if you have 4 offices in your corporation, each office might have a different facility code to distinguish them). The N's are called the Card Number. These are between 0 and 65535. People typically start at 0 and work their way up as they need more cards. The P's are Parity Bits. Parity Bits allow the system to ensure that the message was received correctly. The system does this by counting all the 1's in the pattern, adding the parity bit, and determining whether the number is even or odd. The first Parity Bit (or Leading Parity) is applied to the first 13 bits, and will always make the number even, and the last Parity Bit (or Trailing Parity) is applied to the last 13 bits, and will always make the number odd. So for instance, if I had a card with a Facility code of 3, and a card number of 20, the encoding would look like this: code:When you add the first 13 digits together, the number is even, and when you add the last 13 digits together, the number is odd. Other formats may or may not have a Facility Code, and may have other properties, depending on who designed it. I'm sure this is clear as mud, so feel free to ask questions for clarification.
|
|
#
¿
Jun 9, 2013 03:23
|
|
|
1550NM posted:What rules and regulations if any governs the deployment of cameras in your part of the world ?. Here in Northern Europe, it's fairly strict. All security camera installations have to be registered with a central government agency, signs warning about the installation must be provided and strictly speaking if your camera sees into public spaces that must be masked out, as only the state from counties and up are allowed to monitor that. No recordings can be stored more than 7 days, unless a crime has been recorded, then no more that 30 days, and so forth. In the US, recordings without audio are usually allowed in any area without a reasonable expectation of privacy. Reasonable expectation of privacy is usually defined as "I could undress or change here and expect not to be seen naked." There is no central registry of surveillance systems. You can go to Cosco and buy a (crappy) surveillance system and put it on the outside of your house facing the street, for instance. In fact, that's extremely common. Most states don't require that you notify people that you are recording them, and I've never heard of a maximum allowable retention time. Many states actually require a minimum retention time for certain areas (typically prisons). Audio is significantly trickier, as it falls under wire-tapping laws and notification is generally required, at a minimum. As an example: I have a camera outside my apartment window looking at the nearest intersection that I swap out when I want to test a new camera. I generally keep low frame-rate recordings for 2 months, or so (It overwrites old footage when it gets past 3TB). Again, I'm not a lawyer, and this is not legal advice. KillHour fucked around with this message at 01:17 on Jun 11, 2013 |
|
#
¿
Jun 11, 2013 01:09
|
|
|
1550NM posted:Quite the difference then, the reasons that we have these regulations over here is that back in the late seventies, just as digital databases and cross referencing started to be viable formed, an agency to overlook how the data was used. Most of the time it's a great bulwark against misuse of data, be that things collected from access control, surveillance, gps tracking or the other myriad of ways you could be traced. In the US, we tend to have a much different view on privacy. Honestly, I feel safer knowing that not all surveillance has government oversight, as I trust the government and the police force a hell of a lot less than I do the guy with a few cameras outside his grocery store. I have a feeling privacy laws like that are going to have to loosen up, because they will eventually become unenforceable.
|
|
#
¿
Jun 12, 2013 00:14
|
|
|
MisterOblivious posted:Just want to point out that pepper spray/Oleoresin Capsicum/OC and tear gas/CS/CN(Mace)really are on the same level. Wisconsin is the only state that I can recall that completely bans the use of CS/CN gas. A lot of the good "pepper spray" products are a OC/CN combo. I've never seen an OC system installed in a house. I think the legality issue would be because it's much harder to say you used it in self-defense when it gets set off automatically on an alarm system rather than by you personally. If you weren't home and it went off, there obviously was no chance of imminent bodily harm, but (again), I'm not a lawyer. Also, having it go off on a false alarm would suck. A lot. Baconroll posted:Are there any reasonably priced outdoor IP cameras which are sensitive to the no-glow 940/950nm IR illuminators ? I've seen some analogue cameras listed as sensitive, but it doesn't seem to be the sort of thing listed in the specs for IP cameras. Sensitivity to a wavelength isn't an on/off type thing, it's a curve. I don't think I've come across a camera that had good 850nm sensitivity, but couldn't see ANY 950nm light. At 950nm, the camera will be less sensitive, though, so you will probably need brighter IR-LEDs. There may be exceptions to this since 950nm is a pretty niche market, and I don't have a ton of experience in it. Most people WANT intruders to know they're being recorded. The dull red glow does a surprising amount for deterrence. the_lion posted:Ages ago, I read that back in the ties of dial-up some ISP (I think in America) was forced to use some sort of surveillance software by the government. It slowed things down too much or something. I've never heard of this, honestly. Google didn't pull anything up for me, either. LegatusP posted:You mentioned earlier in the thread that aside from finger/handprint, vascular, and retinal/iris readers, there were some other less common types of biometric security. Could you go into more details about what else there is? The only other one that springs to mind for me is possibly blood/DNA recognition but that seems unrealistic and invasive, given how long a DNA test takes in a lab setting. Ever hear of "Skin Spectroscopy"? Didn't think so. There are a ton of biometric systems that have never been commercialized. Most of them are behavior-based (Identifying someone by their specific walk gait, signature, speech patterns, etc.). I've run across systems inside man traps that measure your weight and make sure you're within 5% of what you were the last time you came through. Those are mostly to make sure you're not bringing someone else along with you or stealing equipment. There have been rumbles about using infrasonic waves to measure your "acoustic signature" (basically, how dense you are), as well as using microwave and millimeter-length scanners to measure your *ahem* body proportions under your clothes. I'm sure you've seen them if you've ever been to a major American airport: Millimeter scanner, possibly  for creepy 3d genitals.Microwave scanner, also  .Tons more, as well: http://fingerchip.pagesperso-orange.fr/biometrics/types.htm KillHour fucked around with this message at 17:10 on Jun 16, 2013 |
|
#
¿
Jun 16, 2013 16:57
|
|
It's mostly a curiosity thing.
|
Thanatosian posted:Heh, I was just coming in here to ask about proxcards. 13.56mhz systems (from HID, at least) are less expensive, as the cards are cheaper to produce (smaller antennas, less copper). Some companies charge a premium based on their added security, but companies that go by a flat margin would not. Here's an example I found on Google (Not my company, never worked with them): Prox: $3.66/card iClass: $2.76/card Prox: $82.84/reader iClass: $79.00/reader All HID cards have a minimum order quantity of 100 (cards are made to order and cannot be returned). 1550NM posted:For the systems that I have available, about 30% for the readers, and about 90% premium for the cards.  I hope your customers don't go window shopping. KillHour fucked around with this message at 19:46 on Jun 19, 2013 |
|
#
¿
Jun 19, 2013 19:37
|
|
|
Thomamelas posted:Hrm. Tyco bought Exacq. $150 million so they can drive it into the ground. I read about this yesterday. I carry Exacq and compete with Tyco, so this is going to be interesting, to say the least. Rolling your own SHOULDN'T be hard to do; controllers are pretty simple. The problem is that unlike IT, security integrators stick to names they know and are very stubborn fanboys. If you're doing the installs, though, it's not like the end user is going to know the difference. For a cheap prebuilt networked solution, I'm fond of Geovision: http://www.usavisionsys.com/mirror/english/3_1_AS.html If you shop around, you can probably get the GV-AS400, which handles 4 doors and 8 readers, for around 400 bucks. Software's free for up to 4 controllers, so you can do a pretty basic 16 door system for around 100 bucks/door plus readers and cards. If you only need ingress protection, the GV-AS810 will do 8 doors and 8 readers for around 650-700 bucks, bringing the maximum number of doors to 32 with the free software. You're a little mistaken about the way the cards are programmed, though. You don't need to have special authorization to order any cards you want from HID, you just need to get them through a distributor (if you're an integrator) or reseller (if you're an end-user). The exception to this is Corporate 1000, which does need to be set up by HID. Here's the HID ordering form for Prox cards, for instance: https://www.hidglobal.com/sites/hidglobal.com/files/resource_files/d00530-b.9.3-125_khz-physical-access-htog-en.pdf Technically, you CAN order plain prox cards and program them yourself with a 3rd party programmer, since HID's patent on prox cards expired. iClass can be ordered un-programmed as well, but the only way to program them AFAIK is to lease a programmer from HID, which is extremely expensive, and the programmer is locked down to hell and back. As an aside: HID does not make controllers for more than one door. They have their single door system, but they don't talk to each other. They don't want to step on the toes of the controller manufacturers. KillHour fucked around with this message at 15:10 on Jun 20, 2013 |
|
#
¿
Jun 20, 2013 14:49
|
|
|
Thomamelas posted:Yeah, we compete with Exacq and this is going to be interesting. Given the annoyance the caused by switching over to distribution only, I wonder how many more of their dealers they are going to drive away by being part of Tyco. It's going to be interesting. We're a distributor, so the distribution only move was good for us, but this Tyco thing is going to make it .Maybe they're just going to go full-on Honeywell and tell the channel to go gently caress itself. Then again, Dell made a token effort to appease the channel when they bought Sonicwall, so  .
KillHour fucked around with this message at 18:04 on Jun 20, 2013 |
|
#
¿
Jun 20, 2013 17:59
|
|
|
Thanatosian posted:I'm guessing this would fall under "other security topics." Barcodes are pretty cheap. RFID can get pretty expensive. Unfortunately, that's more of a Data Capture / Point of Sale question, and I'm not hugely versed in that area. Maybe some other goons can chime in. Try this: Call CDW/Insight/Tiger Direct and ask for pre-sales tech support regarding the issue. I pretty much guarantee they will call my company (and try to pass us off as their internal team) and one of my colleagues in DC/POS will help you out.
|
|
#
¿
Jun 21, 2013 03:53
|
|
|
movax posted:Huh, I had no idea you could program Prox cards, interesting! They don't want you to know.  http://www.aliexpress.com/item/125KHz-RFID-ID-Card-Reader-Writer-Copier-Programmer-FREE-Rewritable-ID-Card-Keyfob-COPY-ISO-EM/576448257.html KillHour fucked around with this message at 16:41 on Jun 21, 2013 |
|
#
¿
Jun 21, 2013 16:38
|
|
|
LegatusP posted:What is the most ridiculous/paranoid security system you've ever had to set up, in terms of quantity of devices as well as expenses? Was it for a private owner or a corporate building? My more paranoid customers are more worried about their employees than thieves. I've had people put hidden cameras and microphones in employee cubicles and desks. they generally want software to capture screenshots and keystrokes, too. Most of my customers are businesses. I'm not licensed in my state to do installs, I just go design work. Jeesis posted:Have IR LEDs ever been a threat you had to account for? If so, does any company make cameras specifically designed to not pick them up? I'm not sure what you mean by threat. People shining IR lights into the camera to blind it? Most people would just use a regular flashlight for that. HDR helps with the blinding issue on cameras that support it. Most nicer cameras have an IR cut filter to block IR light, but that's just so you get proper color reproduction during the day. As for training, I went to school for network security. I just had an opportunity to branch out. If you want to do installs, you need to become licensed, and have some kind of security background (this usually means cop or military). A lot of cable guys used to go into cctv, but you really need networking background, now. KillHour fucked around with this message at 00:22 on Jun 24, 2013 |
|
#
¿
Jun 24, 2013 00:19
|
|
|
Machai posted:Have you ever turned down a customer because their proposed system was too creepy/invasive? Can you tell us about any of them? No, I don't really have that luxury. Fortunately, I've yet to have someone actually tell me they were using the system to spy on little boys in the shower or something like that. Since I don't do the installs, I can't tell if a customer is lying about what they're using the system for. I have built systems that I was uncomfortable with due to the agency/group using them. I'd rather not talk about those in detail to avoid derailing the thread with my political views, but I made absolutely sure to do my best work on those projects. I don't let my views affect how I do my job because lives are at stake - especially when my customer is controversial. Machai posted:You said you were fairly aware of security cameras, do you ever critique their placement/usage, and do you try and find ways to thwart them for fun (just in your head, not any serious attempt at theft)? Oh God, all the time! My wife and friends are all sick of hearing me talk about how crappy the system is in my credit union / dollar stores / malls. Games like Thief and Hitman have always been my favorite genre, and I always play the sneaky/rogue/non-lethal character whenever possible. I have many a fantasy about breaking in to places, and I honestly believe that makes me better at my job.
|
|
#
¿
Jun 27, 2013 20:03
|
|
|
movax posted:We used to have this ancient HID unit where you write down the memory locations you stored given cards in, it was terrible (and you had to wave the credentials in front of the reader in sequence to program it). The only digital interface was IrDa to an old PalmPilot Do... do you need a hug? Seriously, just buy this: http://www.everfocus.com/aproduct.cfm?productid=1&tab=specs You can probably get it for $350 or so if you shop around, and it has to be better than replicating the database between Edge Solo controllers. KillHour fucked around with this message at 03:42 on Jul 3, 2013 |
|
#
¿
Jul 3, 2013 03:38
|
|
|
movax posted:There was a guy selling BNIB RP15s for $60/pc on eBay, I can only assume they fell of a truck or something. drat, that's cheap. I can't say what HID sells them to disti for, but it's a LOT more than that.
|
|
#
¿
Jul 3, 2013 17:39
|
|
|
Dmaonk posted:Is the placement of hidden cameras and microphones for the sake of monitoring your own employees legal in the US? Hidden cameras are generally considered allowable in the US. Microphones, however, are very tricky. I ALWAYS tell my clients to talk to a lawyer, but very few of them actually do. They never intend on their employees finding out, after all.
|
|
#
¿
Jul 5, 2013 03:50
|
|
|
Thomamelas posted:Generally you'd use a DVR with analog cameras and an NVR with IP Cameras. Some cameras have an SD card option but I'd rather see those used to as a backup for network issues rather then a primary storage devices. Cameras can be vandal resistant but none are vandal proof. So you probably want to have them recording to something else. While I agree with a lot of this, I think cameras can have a tremendous value for home users. The other week, I talked to a guy whose car was hit in the street by a drunk driver. His security camera caught the crash, and the police recognized the guy's car (repeat offender). Sure enough, the car was parked in his driveway with a busted bumper. And in my experience, it definitely acts as a deterrent for thieves. It should get you a discount on insurance, at the very least. That being said, I agree that if you don't have good locks on your doors, that's the place to start. A dog won't do crap, and unless you plan on investing in crowbar-proof windows ($$$), planting really thorny bushes under your windows is the best way to keep thieves from coming in that way. We just picked up LG as a vendor last week, and their indoor box cameras are pretty much the best price/performance ratio I've ever seen. http://www.lgecommercial.com/security-en/products/network-product/box-camera/lg-LNB3100 MSRP: $300; Street Price: <$250 720p, HDR*, H.264, varifocal lens, what's not to love? *Real HDR, none of that "software HDR" poo poo. LG claims 80dB SNR. Here's a low cost outdoor bullet: http://www.vivotek.com/web/product/ProductDetail.aspx?Model=IP8332 Street Price: $300 If you're doing 8 cameras or less, XProtect Go is free: http://www.milestonesys.com/Software/XProtect-IP-Video-Surveillance/xprotectgo/ Load that sucker up on a cheap PC (don't cheap out on the PSU!) and throw one or two of these in there: http://www.newegg.com/Product/Product.aspx?Item=N82E16822136783 You don't have to worry about people getting into your cameras as long as you change the password on them to something people won't guess. Lots of idiots leave the default password or no password at all. That's pretty much the best you're gonna get for $2k or less. If you go much cheaper, quality drops like a rock (unless you go Ubiquiti, but I've never tried them, so YMMV). KillHour fucked around with this message at 03:09 on Jul 13, 2013 |
|
#
¿
Jul 13, 2013 02:24
|
|
|
Thomamelas posted:For every story one of us has about the thief who saw the camera and walked away, we've always got two for the idiots who got right up to the camera and looked directly into it to see if it's on. But when you look at the studies based around measuring CCTV's effectiveness as a deterrent, they tend to show a very negligible deterrence factor when they show any at all. Now every design goal you have creates a series of trade offs. Cameras positioned to catch employee theft may not be well placed to catch a brawl in the lobby. Placing cameras to catch both may mean increasing budget requirements. For a home user to maximize theoretical deterrence is going to compromise placement for forensic use, to chase a goal that shows up as statistical noise. There may be a one in a million chance it prevents a break in but it's not something to bet on. And while the gentleman in your story might have been pleased the drunk driver was caught, he probably would have been happier if his car hadn't been hit in the first place. I'd love to see any statistics you have regarding the effectiveness of camera as a deterrent. All the stats I've seen were marketing to sell cameras. I still think that cameras can be effective in a home environment, but they usually aren't the most bang for the buck.
|
|
#
¿
Jul 14, 2013 02:47
|
|
|
Spatule posted:I am looking for a motion detection camera system that could safely store (in the cloud thus) video clips / pictures of any break in, and alert me somehow (email/text on cellphone). Wifi / power network based, no cabled ethernet in the rooms. There are some cameras with PIR motion detection. The only one I can think of with wireless is this one: http://www.lgecommercial.com/security-en/products/network-product/cube-camera/lg-LW130W http://www.amazon.com/LG-Electronics-LW130W-Network-Camera/dp/B006MZFRSW It's 720p, so the resolution should be high enough for facial recognition if you put it in a hallway or something, but I've never seen the picture in person, so YMMV. From what I've heard, it's similar to an Axis M1054, which is a fairly capable home camera. A quick Google brought up these guys for cloud video storage: http://www.ivideon.com/ I've never used them, but they look cheap enough. You will need at least 2mbps/camera upload speed for 720p @ 8-10 fps. I highly recommend against storing the video only in the cloud. At the very least, you should put SD cards in the cameras as a backup. I'd actually rather see you get a Synology (or similar) NAS for recording. They're cheap enough, and they can do double duty as a DLNA server.
|
|
#
¿
Jul 15, 2013 15:01
|
|
|
Spatule posted:Thanks for the info. I don't need PIR if video motion detection is decent, but it should definitely be in-camera unless you know of a cheap appliance that could do that. PIR is definitely what you want if you don't want anything and everything setting off the motion alarm. The camera I linked has it built in (it's the white round thing in the middle). There are some Sony cameras that can ignore things like trees swaying, but they aren't foolproof and cost well over $1000 each. Pretty much every camera can push to an FTP server, but I've never used that particular LG one, so I'm not 100%. You might want to check the user manual. Also, you definitely want local storage. It is pretty much a given that your internet will poo poo itself when you need it the most, so I'd say having local storage on hand is a good idea. A NAS is pretty drat small, so it's easy to hide if you're worried about it getting stolen. As someone who sells cloud solutions all day, let me just say: cloud is ONLY good for a backup. If you keep your only copy of something in "the cloud", you're going to lose it - guaranteed. KillHour fucked around with this message at 05:15 on Jul 16, 2013 |
|
#
¿
Jul 16, 2013 05:05
|
|
|
Spatule posted:If any NAS is ok for this then I'll just get either http://www.amazon.co.uk/Synology-DS112J-DiskStation-Diskless-Enclosure/dp/B007E8DIO6/ref=sr_1_87?ie=UTF8&qid=1373972614&sr=8-87&keywords=NAS or http://www.amazon.co.uk/ZyXEL-NSA310-Network-Storage-Closure/dp/B005LDM09U/ref=sr_1_19?ie=UTF8&qid=1373972499&sr=8-19&keywords=NAS If you're just pushing to a network drive (FTP or similar) from the camera, you don't need anything security specific. If you want to use the NAS to centrally control them (I HIGHLY recommend it, for your sanity), then you want something like the Synology that's designed for it. Note - the Synology will only come with 1 camera license - you need to buy more: http://www.amazon.com/Synology-IP-Camera-License-Pack/dp/B001MJ0JAO The particular Synology you picked tops out at 5 cameras. Use either an AV drive or an enterprise drive, BTW. A green drive will die in a few months in surveillance conditions. http://www.newegg.com/Product/Product.aspx?Item=N82E16822136783 Spatule posted:Ok, looks like I found a camera that has what I want and isn't crazy expensive, and has good reviews on Amazon. I guess I wasn't clear with what I was saying. Not all cameras have PIR sensors. Most don't, in fact. The LG one I linked to just happened to have one. KillHour fucked around with this message at 14:06 on Jul 16, 2013 |
|
#
¿
Jul 16, 2013 14:01
|
|
|
Spatule posted:That does sound interesting, but expensive as gently caress (50 bucks per license, wow). Do you know of any supported camera that features 720p+ resolution, PIR and IR illumination ? The one I found isn't supported apparently. There's a Dlink that is, but reviews says it blows (poo poo firmware). 50 bucks per license is actually dirt-cheap, believe it or not. Most software is between $100 and $200 per license. This is the cheapest system I know of including licensing: http://www.toshibasecurity.com/products/recorders/prod_recorder_detail_esv4.jsp http://www.amazon.com/Toshiba-ESV4-1T-Surveillix-4-Channel-Recorder/dp/B00BBRRPMC Comes with a 1TB HDD and supports 4 cameras. The only downside is that camera support is relatively limited. This is the only 720p camera I can think of with PIR that it supports: http://www.amazon.com/Communication...&keywords=m1054 No wireless, though. IR illumination and PIR are going to be hard to find in the same camera. The Axis and LG ones have a bright white LED, however, which has the useful side effect of making the person breaking into your house look directly into the camera when it comes on.
|
|
#
¿
Jul 16, 2013 16:03
|
|
|
Spatule posted:Ok, so that's just 1000$ for 3 cameras and NAS, good. Something is unclear about the Synology though, does it come with a license or not ? Some reviewer complain that it doesn't, but it says it does on their website. The Synology comes with a single license. As for cloud backup, you can probably have it automatically push the folder to dropbox nightly, or something. Never set one up, personally.
|
|
#
¿
Jul 17, 2013 03:01
|
|
|
Gromit posted:It's funny you said this. Back when I worked computer forensics for the State police it used to come up every now and again that we could buy cameras that would hash each photo taken and write them to a CD so you could prove they weren't altered. The expensive ($4k+) DVRs/NVRs have, for the most part. The cheap ones? Eh...
|
|
#
¿
Jul 17, 2013 13:48
|
|
|
Thomamelas posted:So a practical note on the cloud backup for home users. If the goal is to have video in case the NVR is stolen then the backup needs to be pretty close to real time. If it delays more then a couple of minutes then you aren't likely to get what you need. Example: The reason this is done is because most security specific cloud solutions are aimed at businesses. Businesses don't have the spare bandwidth to reserve 1mbps+ of upload speed per camera during business hours, and they aren't particularly worried about someone stealing their NVRs. Companies that do care about this are willing to pay $15+/mo/camera. I hate to say it, but cloud providers don't care about the amount a home user is willing to spend. EssOEss posted:I have an apartment building in a relatively bad neighborhood - there have been some burglaries. Therefore, I would like to set up some video surveillance. My main goal is deterrence. It would also be nice to pass along to the police some useful info from the recordings if a break-in does happen. What do you recommend? Details follow. You probably don't want PTZ cameras for the garage ($$$). Will these work? http://www.axis.com/products/cam_m3007pv/index.htm http://www.axis.com/products/cam_m3007pv/video/index.htm Your options for darkness are to either brighten it up (white light or LEDs - White light is better for deterrence, especially on a motion detector) or to use cameras designed for difficult lighting conditions ($$$) Something like this is ideal for that: https://www.youtube.com/watch?v=x-XIlByytgk You may just want to reinforce the garage wall if deterrence is your main motivation. Or put cheap cameras in these: http://www.dotworkz.com/products/S-type-camera-enclosure/ (They're loving huge and scary looking) https://www.youtube.com/watch?v=MkPxjkgfJc4 Thomamelas posted:Killhour, I mentioned the studies showing a lack of deterrence, and while I don't have the .pdfs anymore, I can point you to which studies they were at least. Thanks for this, I'll do some reading. Something I'd like to note, though, is that the second article refers to city wide deployments concentrating on violent crime, which most private systems aren't concerned with. The first blurb didn't draw any conclusions, but I'll take your word on it. KillHour fucked around with this message at 20:07 on Jul 17, 2013 |
|
#
¿
Jul 17, 2013 19:45
|
|
|
|
| # ¿ Apr 23, 2024 22:25 |
|
|
7thBatallion posted:Why in the high hells is thermal imaging still so goddamn expensive? It's like you can't pick up a FLIR for less than ten grand.. Probably because FLIR (The company, not the technology) is in the business of making expensive cameras. The Axis ones aren't as bad: http://www.bhphotovideo.com/bnh/controller/home?O=&sku=714151&is=REG&Q=&A=details Still expensive, but not 10k. Thomamelas posted:It's Dropcam's core market. Alarm.com's core market. Same with the new system Comcast is rolling out. And Vivint. According to the Axis rep I was talking to, most of the AHVS customers signed up were home users. All of the big VaaS players want that market desperately. The problem they run into is high start up costs like Dropcam, or infrastructure issues. I talk to three or four integrators a week who think they've come up with the next brilliant industry shaking idea of leasing cameras to home users and getting the monthly fee. I should have qualified it by saying companies weren't interested in doing it properly. All the home "cloud security" companies cut corners to make themselves profitable, and I wouldn't trust them as far as I can throw them. Those same 3 or 4 integrators a week call me, too. They yell at me over the phone when I tell them how much it would cost to do it properly. You could probably do it if customers were willing to pay $10 per camera per month for D1 or VGA resolution @ 2-3 FPS with 7 days retention (This is roughly in line with what EMC charges for their AHVS solution). Anything more than that, and you start loosing money fast. As an aside: EMC's AVHS interface and feature set is loving TERRIBLE. I had to evaluate it to see if it was something we wanted to carry, and I told my director to run away far and fast. KillHour fucked around with this message at 14:34 on Jul 18, 2013 |
|
#
¿
Jul 18, 2013 14:12
|
|