Soundy

But even the companies selling it as a total solution with the software/hardware dosent tell about the limitations! Because there's still a lot that's out of their control? I could put 1000 cameras on a P3 with a 10Mbit network and it would do just fine if I was using motion-detect to record a glacier in the Antarctic... the list of ifs/ands/buts would be a very thick book.

Probably because performance is more dependent on hardware than software? The hardware part of the equation alone depends on processor speed, processor cores, memory bus speed, I/O bus speed, disk I/O speed, disk configuration... toss in operating system (in)efficiencies, and the matrix becomes unwieldy at best. If you're talking IP cams, there's also the speed and bandwidth capabilities of the network itself... at this point, the matrix becomes impossibly huge. Now factor in resolution, framerate, codec, and such camera-dependent factors, and the matrix is all but useless. The limitations of software itself is pretty much a non-factor at this point.

Could go for any number of covert cameras that record internally to SD memory card - they're typically disguised as PIR motion sensors, smoke detectors, etc. No wiring or wireless receiver necessary - just go in after hours, pop the card out, plug it into your computer to offload the video.

Ya, 50mm is REALLY tight for most CCTV purposes on a 1/4" or 1/3" chip. It should attach to the camera fine, as board cameras typically all have the same mounting threads, but are you sure this lens will fit inside the particular bullet case you have?

Guess I need to repeat myself: Browser based apps suck. Best to install a client. If they want to use crappy software, thats their rights. But ofcourse, everything posted on this forum is just a persons opinion. How do you know their software sucks? Because that it is very bold statement if you never used it. It worked very well when I tested it for 2 months. The only reason I am not using it for our solution is I want one NVR software solution to support and their software doesn't make sense for smaller camera counts. Keep in mind, this is the same guy who said "Slow Shutter speed is USELESS for CCTV." and was given at least one example where it wasn't... Rory seems to deal in absolutes, and if something doesn't work or doesn't have a purpose in HIS world, then it's useless in everyone else's world as well. Wow... anal, eh? Pot, meet kettle. For someone who "doesn't care" you sure spend a lot of time arguing about it. What "sucks" is proprietary client apps that only run on a single platform and require installation anywhere you want to use it. I've seen plenty of client apps that suck on their own, because they were written by software engineers with no sense of aesthetics, who never actually use the stuff for its intended purpose so they have no idea HOW it should actually work. But does that mean ALL client apps suck? No. Does it mean I go around saying "ALL PC-BASED APPS SUCK"? No. That would be ridiculous. And so is your statement.

I doubt they're resized, for the most part: most larger ones will be closer to 640x480 (VGA) to 704x480 (4CIF/D1) , and many will be only 352x240 (CIF). Those are the standard capture resolutions used in analog video. The largest analog-capture image you'll see is 740x480, as that's pretty much the limit of most analog-capture hardware. As for whether they're 420TVL or better, there are many other factors that have a far larger effect on image quality: lens quality; lens aperture; focus accuracy; shutter speed; clarity of housing glass or plastic, if any, in front of the lens; quality of the sensor; quality of the processing circuitry... Once you get the signal out of the camera, final image quality will also depend on capture resolution; quality of analog/digital conversion; compression codec used; compression level used; recompression, if any; post-processing, if any... With variations on all these factors, I could show you pictures from a 320TVL camera that look far better than something from a 520TVL camera...

Yes, there's always an element of luck involved, as well as a bit of voodoo magic - sometimes you can follow all the "rules" to the letter and use only the best components,and still have problems. I have a rough idea of how video baluns are wired internally, but I've never delved into the exact design or WHY they're done that way... but I do know how audio baluns and balanced lines work, and I know they're done differently. In a balanced audio line, you can run right beside high-voltage, high-current AC without a problem - because of the way the balanced line works, any noise (EMI) that's induced in the line, is induced equally in both sides of the line, and gets "canceled out" when it hits the transformer at the receiving end. The transformer also provides physical isolation of the signal source from the receiving equipment, so you avoid ground loops in the signal line. A typical balanced audio also includes a shield, so the two devices are grounded to each other via that, but the signal line itself is isolated.

You did not state if the systems were active or passive baluns, I have found that Interference ground loops are more comon on passive systems running 12v DC,over the ones with 24vAC, even the AC systems can be a victim of the loop or hum bar. With AC cameras, it's more likely the existence of an additional ground path, particularly if the cameras are mounted to something metal that's also grounded. That's not typically related to a ground loop. EMI (electro-magnetic interference) is INDUCED noise from another energy source. One CAN exacerbate the other, but they're not the same cause. That's because the active balun inserts electronics into the path that physically breaks the ground path.

Why are you posting the prices of 400mm lenses? Just for comparison purposes, showing the difference in price for different apertures.

ANY resolution can show them all, they'd just have to be scaled down for a lower resolution.

Agreed. Here's one: http://www.muxlab.com/products/ve_vga_balun.html If you need higher resolution, you can always use an active pair as well.

Thank you very much for sharing your knowledge/experience. From my reading on the Internet there apparently are some RS-485 applications (not neccessarily CCTV) where people were claiming shielding was important - it had me confused. Can't think of a one where it would make a difference at only 40', unless it has to pass through the core of a kilovolt transformer or something.

Like I said, this is where lens "speed" and aperture really comes into play. I'm guessing most people here won't be getting dealer pricing on SLR lenses, so going by retail listings, so using one of our local brick-and-mortar retailers here as a reference: Canon EF 400mm f/5.6L USM: $1489.99 Canon EF 400mm f/4.0L IS USM: $6899.99 Canon EF 400mm f/2.8L IS USM: $8999.99 While the IS does add a bit the price, the big difference is the sheer amount of glass needed for the faster lens... and additional care that must be used in the lens's manufacturing. According the B&H Photo link where I found it, that lens retails for $9,999. The converter is only $300. Of course, if you want to get really crazy, B&H lists a Sigma 200-500mm f/2.8 zoom for a mere $28,999... or if that's too rich for ya, the Canon EF 800mm f/5.6 IS USM goes for a measly $11,499. The Sigma 200-500 lists at 34.6 lbs, BTW...

Just about any single pair will do - Cat5e, station wire, 18/2... doesn't need to be shielded or overly large-gauge. As for readily-available all-in-one cable: you can use Cat3/5/5e/6 UTP: video on one pair (with baluns), RS-485 on one pair, power on the other two pairs.

Sounds more like a basic choke than a "ground loop isolator" - ie. it's not really "isolating" anything. In audio, at least, a ground loop isolator does exactly that: it's a basic matching transformer (1:1 windings) in the audio line that physically breaks the ground path via the audio lines. Of course, "baluns" are a little different design in audio as well, as they too are generally transformers with the input wired to one coil and the output to the other.

Very good observation... What is the key on this is how you describe - as unless someone is spending as much on the lens as for the camera, the outcome will never be optimized. Well.. not necessarily. A Canon EF 50mm f/1.8 can be had for <$200 *new*. But the point is, lens aperture DOES come into play a lot more than it does with C/CS-mount lenses, simply because there IS a much wider range available. I don't think I've seen a C/CS lens under 70mm that wasn't AT LEAST f/1.8 or faster, making it a factor that doesn't often need as much attention. As far as EF-mount lenses for the Avigilon cameras... it's plenty economical, since there are 20 years' worth of Canon SLRs that use them. A number of third-party manufacturers actually have full lines of lenses with a number of different mounts available (for Canon, Nikon, Pentax, Konika/Minolta, etc.), which makes it even more economical for them. However, with the larger sensor, you need substantially longer focal lengths to achieve the same FOV... and for the same aperture, that means a much larger lens structure overall, and a lot more glass. And with the higher resolution, that also means higher manufacturing tolerances than is needed for standard CCTV cameras. All that equates to more expensive lenses... but as with everything, you get what you pay for. Keep in mind too, that a lot of these Avigilon samples are using Canon L-series lenses, which are their high-end professional models that often cost 4-5 times as much as similar "consumer" lenses. Might be part of the reason Avigilon went with the EF mount - there's a HUGE market out there already for the lenses, and several third-party manufacturers, which makes the lenses a lot more economical than they might be if they were proprietary to the cameras.

Gee Rory, wasn't it you that was just telling me in another thread, it's the installer's fault if he has trouble mounting a bullet cam??

and yet it still cant see at night (0.1 lux is NOT good enough for most apps and yes I have seen their couple so called night shots which are loaded with outdoor lighting) Ah, but then the question of the lens really comes into play. Remember, where your average CCTV lens has a maximum aperture anywhere from f/1.2 to f/1.8, f/2 is considered VERY fast for most EF-mount lenses, especially with varifocals (talking in the upper-mid range, four-digit-pricetag range), and f/2.8 to f/4 is far more common. The Sigma 300-600mm beast seen in the "Megapixel images" thread is f/5.6 *at best*, and with the 2x converter on the rig shown there, it won't open beyond f/11. In a nutshell, unless you're spending as much on the lens as you are the camera, most of the lenses you're putting on the Avigilon are at least two stops slower - or one-quarter the light transmission - than typical "low-end" megapixel C/CS-mount lenses. At f/11, that's a full SIX stops slower than f/2... or 1/64 the amount of light.

That's not noise... you have ghosts! OOooOooOoOOOOOooOoooo... There are tons of noise-reduction utilities and plugins in the photographic world... I've used NeatImage with some pretty good success. Noise Ninja is another popular one. Most better photo-editing programs (Photoshop, Paintshop Pro, etc.) have decent noise-reduction tools built-in. There's a rather extensive listing and reviews of different packages here: http://www.photographyreview.com/cat/digital-photography-software/noise-reduction-software/pls_5647crx.aspx

I don't have any grand expectation that it's a magic fix, but like I say, it was something I had an idea to experiment with. If it DOES work, a single diode will be just as effective as a bridge rectifier - all the bridge effectively does is add another diode to the power line.

Sunlight on its own should not be causing motion detection. It may be something else moving through the field causing changes in brightness, or reflections off moving objects. Changing to a manual- or fixed-iris lens isn't likely to solve your problem - in fact, if you have wide variations in lighting, the lack of ability to adjust to that with an auto-iris will probably make your situation worse - either it will be completely washed out in the bright sun, or if you adjust for the bright sun, it will be too dark once the light is gone.

In the absence of a model number, photos of one of these devices might help.

That would really depend on that camera's specific design. We could say "probably" but there are just too many variations on the theme out there for anyone to say with certainty without at least a make and model (and even that may not help). You're pretty much on your own to experiment with that... That's a definite "maybe". Chances are, the wires attach to the camera board with a simple three-pin header, as the power and video likely share a ground. The trick is, if you do get it apart, you have to find another board with the same connector using the same pinouts... and depending on how things are packed into the case, the pins may have to be in the exact same position. There is no "standard" for these things. I wouldn't count on the wiring for a cheap camera being "high quality", or even "medium quality".

if you use an adjustable regulator, you can just insert your diode or rectifier, put a meter at the camera and tweak your output voltage for the desired results

The limited DOF in this cameras is not because of the lens mount, and not DIRECTLY because of the sensor size. The problem is, with the smaller sensor, you need to use shorter lenses, and THAT is why you get greater DOF. Typically CCTV cameras have 1/4" or 1/3" sensors, compared to 16mm frame sizes in film, or 24-35mm sensors in DSLR cameras. For example, my 50mm f/1.8 lens on my Canon 40D ("1.6x crop" APS-C sized sensor), I get about a 25x17-degree field of view and very shallow DOF at f/1.8. To get that same FOV on a 1/4" CCTV sensor, you'd use an 8mm lens... and 8mm will give you pretty substantial DOF no matter what you do (well, unless you can get your aperture down to f/0.2 or so - good luck with that). None that I'm aware of. Generally, the camera and lens for this purpose are a single unit. You'd need some way to power the lens and feed it control signal separate from the camera. Frankly, if you do find something like that, it's likely to be fairly expensive.