NotoriousBRK

It's a D1 camera, which makes sense given the current state-of-the-art in megapixel sensors just beginning to get into acceptable low-light specs. Considering some of the development done by companies like Noble Peak that never really came to bear in terms of an actual product, this isn't a totally unbelievable announcement. Will be interesting to see what it is using for a sensor, and sensor size.

Not necessarily. The twisted pair Ethernet distance limitation is a factor of timing, not signal strength. The electrical pulses on twisted pair Ethernet can go far beyond 100 Meters. But, once you exceed that distance the back and forth propagation times can cause the collision-sense part of the stack to get confused (I am greatly summarizing here). So, a proper distance extender is also a protocol convertor.

Since RS-485 communication ports use a balanced signaling system, you'd really want to use a twisted pair cable instead of a basic wire (phone wire, lamp cord, etc.) in order to get maximum benefit from the noise immunity of the balanced signal. Any twisted pair cable from Cat3 on should work just fine though. A chunk of network cable is probably the most common approach since it's readily available in most cases.

Good suggestion.

For what you're describing there are a ton of options. The transceivers will be around $70/ea, and for more money you really won't get much of consequence in this case. Biggest decision is probably should you to Fo/Cu standalone transceivers, or do a switch with integrated Fo ports. I'd say find a transceiver option first, and then get a length of pre-terminated cable with matching connectors on it. Some options: http://www.newegg.com/Product/Product.aspx?Item=N82E16833156085 http://www.newegg.com/Product/Product.aspx?Item=N82E16833156005 http://www.garrettcom.com/es42.htm Your biggest challenge will probably just be running the fiber, and managing not to damage it. Which isn't really much of an issue if you a apply a little common sense and caution.

Depends, but I'm thinking of this as an IP-based system (so, NVR not DVR) using some kind of wireless network (mesh, 802.11 point to point, etc.) with a centralized NVR/video server somewhere. There are a ton of what-ifs, and possible outcomes with this scenario that will make it impossible to properly design without an on-site consult. My initial answers were to give him something to think about and establish pricing at least within an order of magnitude. I get a lot of dealers and end-users that come to us with grand dreams of CSI-grade security across a big subdivision or campus, and then when they realize the budget and practicalities recognize that it might be best to scale back a bit. So, in this case, it could end up being DVR's, NVR's, storage in the camera (my favorite ) or some combination. I've also seen residential systems like this where some of the components are mounted in or on the residences and the resident gets a comp from the HOA for power usage and convenience.

*Maybe*. You can get down to about $10/foot for a low pixel density (~8 pixels per foot) system, again being very rough budget. (I'm a manufacturer (VideoIQ), not an installer, so others can probably give you better quotes). You have 3 entrances, each would need at least 2 cameras (plus recorders, mounting hardware, installation, etc.). Frankly I wouldn't even try to do that with normal cameras (for license plates), I'd use specialized LP optimized cameras (I like this one: http://www.iluminarinc.com/LPRSSeries.html ) along with a megapixel camera as an overview of the gate. But that's going to set you back around $6000 just for the gates (cameras, encoders, recorder, etc.). Let's say that gives you $70K to spread over ~ 2miles of interior. Lots of small things come into play here... Can you mount cameras to existing poles, or do you need to install your own poles? Is power available for the cameras? What is the illumination like? Do you need to worry about privacy masking? I think that what you want is not really want you want. You are unlikely with even a million dollar system to get end to end coverage sufficient to actually catch people in the act. And in most places, if you don't have video of the suspect actually committing the crime, then the other video is just "helpful" it's not "prosecution grade". I'd do something like this: 1) Add high resolution megapixel (1080p) cameras and good lighting at the gates to log all vehicles coming in/out 2) Add dedicated License Plate capture cams at the gates to log plates. 3) Put up high resolution (1080p) cameras at 9 or 10 key points along the roadways so that if a vehicle comes into the property you have a rough idea of where it went and when. This might be a $20-$40K system, depending on lots of factors. But, if something happens you'll have a pretty good log of suspicious vehicles and timestamps of when they were in the area. The police can use this in a typical investigation, and while it might not lead to an immediate arrest, it will greatly optimize the investigation process. Most likely you won't get much incremental evidence quality by spending another $200K, the next jump would be the full prison-state lockdown setup with cameras end to end, and most residential areas don't really want that.

Cut one of those "high quality" thick cables in half sometime. 95% of the time they have the exact same amount of metal as cheap cables, and about 300% more insulation to bulk up the cable and make it appear to be more robust. Years ago when I was starting an A/V store, I was visited by a local rep of one of the premier higher-end (marketed above Monster) cable companies. Came in giving a big pitch about the need for ~10% of the home theater budget to go to cables that can deliver superior sound and image quality and so on. So I played along with him, nodded my head, etc. I asked him if his cables really made that much of a noticeable difference, and of course he told me they did. So, I took him to the lower floor of my store and showed him where we were building some reference home theater demo rooms. I told him I'd take my $25,000 theater room and duplicate it exactly. One room would use all the default/free cables, and the other room would use all his high-end cables. That way, customers could clearly see and hear the difference between the two setups in a true side-by-side comparison. Then he starts backpedalling, telling me that's not a good idea. It would just confuse people, they are are coming in for my "professional advice and design" and yadda yadda. Never saw him again after that. Good cables are necessary for a proper system, but the bar between "utter crap" and "more than sufficient" is actually pretty low in most cases. If the output isn't what you're expecting, it's rarely the cable. Terminations however, especially hack easy DIY connectors, CAN be problematic though.

Does the picture not look good now? With analog video, better cables generally *will* show an improvement in the video, but you also have to be careful with the cheapie all-in-one/DIY camera kits. Many times the cameras themselves are relatively low quality, and the issue is not the cable, it's the camera.

Check fencing supply places as well. There are a lot of options if you're just looking for a post. However, there is only so much aesthetic variation you're going to get out of a basic post...

A very rough budget, but if you want to get faces and plates across that entire area, figure about $75 USD per linear foot of coverage with that level of detail.

What you need is an IP encoder at the camera end, and an IP decoder at the DVR end. I believe Axis may have the most cost-effective options for you.

2 cents. Also, the power outputs varies somewhat significantly device to device. So, not all 30 cheap IR LEDs are created equal.

The cameras with built-in IR usually just have a few cheap T 1 3/4 sized LED's. These are effectively unfocused, they just spit IR out the front (mostly). The distance is going to be very limited (almost never more than 30ish feet, despite what might be claimed). Dispersion will suck also, so you'll get bright spots in front of the camera and very dark edges. Also, IR LEDs create heat. Heat attract insects. Insects attract spiders. Spiders build webs, and now you've got a mess in front of your outdoor camera. A separate IR illuminator (Illuminar and Raytec being my favorites) are a purpose-built illumination device. They use high-power SMT LEDs, and will arrange the LEDs so that there is a controlled or defined (predictable) dispersion pattern. You purchase units of various size/power to match to your application and to the field of view of the camera. You can mount the illuminator away from the camera, so that insects are not drawn to the camera, and so that your illumination is closer to the target (when possible). Higher-end illuminators also have options to vary the power/intensity of the LEDs, either through network control, a potentiometer or a serial control.

Keep in mind these are approximate, as the final answer will depend on exactly what equipment you use: 500MB/Day = 320x240 recording at 2fps, encoded at ~40Kbps, h.264 Video at this size/rate will show obvious compression artifacts and loss of detail, but would be generally sufficient to prove timeline and chain of events (eg: vehicle 1 collided with vehicle 2, passenger in vehicle 1 exited under their own effort). 1.2GB/Day = 640x480 recording at 5fps, encoded at ~100Kbps, h.264 Roughly the same results as above, but with more frames and larger image size. 8GB/Day = 640x480 recording at 15 fps, encoded at ~750Kbps, h.264 Video is relatively fluid and shows little apparent loss of detail from encoding. 16GB/Day = 640x480 recording at 15 fps, encoded at ~1.5Mbps, MPEG4 Video is relatively fluid and shows little apparent loss of detail from encoding. 20GB/Day 640x480 recording at 5fps, using MJPEG Video is less fluid, moderate detail. 40GB/Day 640x480 recording at 15fps using MJPEG Video is relatively fluid and shows little apparent loss of detail.

Depending on compression method and settings, you will use between 500MB and 20GB per camera per day of recording.

It's too bad that more people don't realize this...

So, you want someone to do the design work, and then you will most likely purchase and install the recommended equipment on your own? Nothing wrong with that. That's why it is called consulting work. It is done all the time in new construction of major projects. A lot of times you don't have to touch a thing other than giving the client a set of drawings and specs of materials used. That is why I gave the original poster my rock bottom rates. I didn't say there was anything wrong with it. I was just trying to figure out what exactly he wants. I have 4 good integrators in the SE Michigan area that I could potentially refer for him, but I know that at least one of them wouldn't be interested in a design-only job.

So, you want someone to do the design work, and then you will most likely purchase and install the recommended equipment on your own?

You should take this advice to heart. What are YOU bringing to the deal? Who are YOU?

The first one doesn't list the pan and tilt specs, which is usually an indicator it is limited. From the looks of it, I'm guessing it doesn't do a full 90 degree tilt, which can be a problem in many cases. The bottom one looks average. It's all a matter of price. I can get either of those cameras for around $200-$250, depending on how much I feel like arguing with the manufacturer about quantities and shipping costs. Note, I'm saying that as a reference, not as an offer to import them and sell them (not worth the trouble). Personally, I've used some similar units for my own houses and some basic demos, but I'd be hesitant to actually sell them to a customer.

That's actually good. There are a ton of analog cameras out there that are IR sensitive up to around 1000nm. Most of the analog cameras based on the latest Pixim chips should give you good low light, good dynamic range, and IR sensitivity. I'm not sure you *really* want to do 940nm though. It takes a LOT of 940nm light to give good, even scene illumination. I'd say roughy you'll spend 1.5-2x more on 940nm over 850nm IR all things considered.

Your camera (any camera) has a fixed number of pixels. If you spread those pixels over a very wide area, then each pixel is taking a large amount of square-footage area and storing it as a single averaged pixel/color. If you spread those pixels over a very small area,then each pixel is storing a smaller amount of area, which yields better detail. This is usually discussed as "pixels per foot", or how many pixels are spread over 1 foot of real-world area. To get good object detail (faces, license plates, etc.), you usually want at least 70 pixels per foot, and more like 80-100. You also need good light at night. This usually means a real light source (floodlight or real IR illuminator), not 12 cheap LEDs sprinkled around the lens. The pixels per foot math is easiest to do on the horizontal axis. The camera you quoted has 768 pixels on the horizontal axis. We want to have at least 80 pixels per horizontal foot. So: 768 / 80 = 9.6 As long as you keep the shot less than 10 feet wide, you'll be getting very good detail. As you go wider than 10 feet, object detail will start to fall off. More than 20 feet wide and it will be very noticeable. You can then use a standard lens calculator to tell you that at 4mm the shot becomes 10ft wide 10 feet from the camera. At 9mm it becomes 10feet wide about 22feet from the camera. So, keep the camera within 20 feet of where you want to detect faces and you should be able to zoom it in enough to get good detail.

Handbrake will do it. Or, you can learn ffmpeg and then the sky is the limit.

Why does the zoom need to be motorized? There are very very few lens manufacturers that do motorized zoom lenses right now, and those that do are mostly intending it for backfocus adjustment (eg: very very infrequent changes) and not as a quasi-PTZ setup. For a 5 MP camera you're going to be looking for a lens with probably a 250 lp/mm or better rating. You're going to have a hard enough time finding that, much less with motorized zoom.