NotoriousBRK

The short version is that probably won't ever work the way you expect.

In most areas you'll find your upstream Internet bandwidth to be fairly limited and unreliable. The practicality of streaming 4 cameras at a bit rate and resolution sufficient for any purpose is pretty low.

You'll also likely run into overage issues (even in most places that have "unlimited" Internet, it's really not) at one or both locations.

You didn't specify what resolution of IP camera, but let's assume a basic 720p camera. That would likely be at least a 2Mbps video stream to get decent resolution out of it. Times 4 is 8Mbps, most Internet connections have significantly less upload speed than that. Additionally, 8Mbps is going to be roughly 80GB/day of transfer. Many companies will start to harass you if you use more than 250GB/month.

Put the recorder on-site...

How big of an area are you looking to cover, and what generally happens in that area?

Outdoor PIRs can be serviceable in some locations where you are trying to cover a relatively small patch (like a standard residential 24'x30' driveway) that doesn't get much random activity.

Trying to cover a commercial parking lot, a large yard area, places with lots of nocturnal animal activity and so forth are frequently a giant headache with PIRs and other non-intelligent systems and you'll need to step up to more advanced analytics.

You'd have to use an encoder.

A balun is still an analog signal, it just carries that analog signal over a different kind of wiring.

That was hilarious, this post should be framed and signed.

The 78xx series regulators are simple devices, but they're really a poor choice for anything that draws more than a very small amount of current.

78xx regulators essentially work by converting the excess voltage into heat, so the greater the difference between your input voltage and their regulated output, the more heat they put out (which is also impacted by current draw). The recommendation for most 78xx regulators is to keep the input voltage not more than ~3 volts higher than the regulated output.

These days it's better to use a switcher style regulator instead of a linear like the 78xx, even if it requires a couple of additional external components since the switcher will run much cooler.

Not knowing what you wired up it's hard to recommend suggestions...

If you google "24vac to 12vdc converter" you'll find a bunch of products in the $9-$20 range that solve this problems directly. Doesn't seem like a wheel worth re-inventing.

I use Omnigraffle Pro.

A guy that used to work for me was previously a service tech for ADT.

One of their accounts was a "Gentlemans Club". There was a flaky window contact in the girls dressing room.

He told me it took 3 techs a total of 6 visits to "fix" the contact, and then come back to "verify" it was still working OK before they closed the ticket.

I have a small Telephone and Wiring business in Wyoming where I sell, install and support Telephone Systems.

Out of curiosity, do you provide all your customers the same brand of phone system, or do you deal with multiple lines?

If you just paste those part numbers into Google you should find hits for all of them.

Looks like a Geovision fisheye camera, and then a couple of no-name dome and bullet style cameras.

Your going to be hard pressed to find a dome camera that has 100ft of IR. I would add a external IR and call it a day.

Totally, especially if the camera is at the door and you want to see someone within a few feet of it. Most cameras with built-in IR are just going to totally wash out on anything close by if the IR even had a *chance* of reaching half that distance.

Cameras and illuminators are frequently items that need to be chosen, and placed, independently for maximum performance. It rarely works out that you find the perfect + illuminator combo, other than for the most basic scenarios.

The main thing to look out for on the cheaper PoE switches is their *total* rated power. They may supply power on 4 or 8 ports, but sometimes it might have a low limit (30 Watts, etc.), which prevents it from supplying *full* power on all ports. Make sure whatever you intend to use with it won't exceed its power budget.

In all honesty that video does not show much value in auto-tracking.

1) It does not seem predictable in terms of what the camera does or does not key in on.

2) It appears to miss a lot of semi-dominant objects in its decision process

3) The constant motion of the camera creates significant motion blur, obscuring details

4) The relatively slow focus of the camera adds delay to getting a sharp image after it stops at a dwell point

5) The level of zoom does not appear to be enough to give much usable detail, it only provides a tighter relative shot.

In a real life scenario, I think you would find that the auto-tracking would have a greater probability of obliterating or missing detail/info, rather than enhancing the scene in any usable way. Go back through that video and try to find a crisp shot of any of your "suspects"...

Not sure what your background is, and it's not super important, but most of the people I've met who have tried auto-tracking PTZs in professional installs (eg: where they are only getting paid if the system works/continues to work as advertised) have decided the reality of the concept does not live up to the hype.

The only systems that *do* seem to approach a reasonable level of usability are ones where you have a spotter camera that can maintain a view/context of the entire scene, and then drive a secondary PTZ to the key points of activity.

800 ft. is to far for *ethernet*, forget about PoE.

You'll probably need/want to use some sort of distance extender product. There are units that can do data+power over coax or Cat5 at that distance, but even if the camera could run off a battery, you won't get a link over 800ft of Ethernet.

If you can run everything off a 12V bus, you can just use a simple 7Ah battery, and a $15 solar charge controller. The charge controller keeps the battery topped off, and provides a battery-backed 12V power source if the main power supply fails (like when power goes out).

DSL bandwidth in the US is almost universally poor.

It's unlikely your DSL connection will have enough upstream speed, and integrity, to allow you to record 8 cameras at anywhere near a decent resolution and framerate.

Supposing you are one of the lucky few with a usable DSL connection, then all you really need is whatever NVR platform you prefer. You'll setup port-forwarding rules on the router (or, even easier, setup a point to point VPN) so that the NVR can access the cameras. But after that, there are not many technological differences between recording across a LAN, or a WAN. It's all just a matter of getting the packets from point A to point B.

What are the specs of the DSL line that is serving the cameras, particularly the upstream bandwidth speed, and any info about packet loss or latency?

Thanks a lot NotoriousBRK! You seem to know a lot about VideoIQ-cameras!

I hope so. I run Field Engineering for VideoIQ. Almost every technote, application note, or demo video you've seen or read is my handiwork.

I guess that, as we are only going to use one pair to signal TO the camera from our central alarm unit when our home security system is armed, and then another pair of cables FROM the camera to tell our central alarm control system if someone has triggered an alarm event, that will leave us with TWO pairs that could be used for powering the camera? Is that correctly understood?

 

Could you explain why 24VAC is better than 12VDC?

 

As the main power source will be a PoE-switch in our house, we will, for back-up, attach the PoE-switch to an UPS unit in case of power failure. As this unit has several 230VAC outlets, could one of them be used in order to back-up the 24VAC tranformer as well so that we will have double redundancy?

 

Thanks!

24VAC will be a little less prone to voltage drop than 12VDC.

A second Cat5 cable for the aux power and alarm I/O will be fine.

I'd recommend 24VAC though, and double up on a pair to minimize voltage drop over the run.

You can use our iCVR Encoder as a single channel DVR.

Options up to 500GB storage. PoE, 24VAC, 12VDC.

http://www.videoiq.com/products/iCVR-Encoders

Depending on what you want to use it for, it might be more expensive than a small 4 channel DVR-only device, but it is a fairly small form-factor, and you can utilize the on-board analytics in a number of ways that may be of benefit to you.

EDIT: nevermind in this case, I see that you're looking for a completely stand-alone unit.

You can accuse me of bias, but Honeywell does not have an analytics option suitable for outdoor use. Add-on software or not, that is not their area of expertise.

Your results are *mostly* going to be related to the lens/sensor combo of that camera.

That being said, you're in the range of what many megapixel cameras require for a good image. It's been a while since I tested a Mobotix camera, they were known for doing very slow shutters in low light conditions, resulting in any motion being very blurred. So, I'm not sure if you tested while watching a moving object in the scene.

Also, for whatever rating you get, it's more than just "2 lux". 2 lux might be OK 50 feet from the camera, but at 500 feet, you'd need more light at the target (light disperses/scatters over distance).

For our cameras I spec 2lux on target per 100' of distance from the camera as my guideline.

One of my dealers did a large home a year ago (I think around 32K sqft.) and mounted our cameras inside Pelco recess mount housings. They basically look like recessed lights with a smoked glass panel.

The images are very good, and you don't notice the cameras at all. There are around 40 or so cameras in the house.

You can use the Live Cams Pro app on an iPhone and view live video from the mjpeg stream, and pull the clips associated with alarm events.

Are you sure? I was told If you have an HDCCTV SDI camera hooked up to an HDCCTV SDI DVR, it is supposed to work, atleast that is the theory, regardless of where it was made. This assumes you have the quality cable and connectors that it requires to run this equipment. Correct me if I am wrong.

For starters, I think you're mixing terms. That's understandable given all the confusion and poor marketing around this.

HD-SDI is a standard (or, set of standards) in the broadcast industry for working with hidef signals from TV camera equipment. It is *mostly* interoperable across various devices. Some manufacturers have started releasing HD-SDI cameras and recorders that should allow you to mix and match brands and equipment readily.

HDCCTV is a quasi-standard created by a group of mostly two people to try and make a proprietary subset of the HD-SDI standard that mostly consists of some set parameters and interoperability testing. In theory all HDCCTV equipment is supposed to be guaranteed to work together. But then I think that HDcctv is a different standard and not guaranteed to work with HDCCTV certified equipment. Any or all of this may or may not work with certain HD-SDI gear.

The biggest issue is that HD-SDI (the "mother" of all this) is a broadcast standard, and doesn't have much concern for things like PTZ control, higher resolutions (3MP, 5MP, etc.), 180degree cams, alternate aspect ratios, and all the other things that are served by IP cams today. You're basically limited to 720p and 1080i/p cameras and recorders.

ALL of the current recorders (and anything predicted for the future) record video in h.264 compression. So, you can talk all you want about the live view quality, framerate, whatever, but the recorded video (what most people watch, and the evidence in any case) will look the same as IP video.

Personally (and I may be biased) I view HDCCTV mostly as a very low end product (once the prices come down) and what we will likely see as the Costo DIY systems. It is going to be forever hindered by the limitations of the HD-SDI standard and the fact that CCTV just doesn't factor heavily into their concerns. We've been hearing for 3 years that HDCCTV was going to be the Next Thing and an IP Killer, but it barely has a pulse and by the time it actually gets any traction it's max resolution (1080p) may already be passe and considered entry-level.

You can use the Axis Bandwidth Calculator - http://www.axis.com/products/video/design_tool/calculator.htm

I'd almost recommend NOT using that. It's highly likely to give incorrect results for almost any other application.

In order to answer this question it requires knowledge of the camera brand/model being used, encoding type, and a rough guess of the scene lighting/activity.