Soundy

That statement is pretty funny considering you went on to discuss it, then posted another response! I'm still trying to figure out where it's been "twisted in so many different directions". Several people have presented facts, details and knowledge that mostly all agree with each other, someone else has quoted contradicting third-party information... not sure where the "twisted in different directions" part comes in. I guess if you're a consumer, it's fine to be spoon-fed "what you need". Maybe some professionals prefer that as well. Personally, I like to know the "how and why" of equipment I use - makes it easier to know the right stuff to use despite marketing gibberish, and a helluva lot easier to troubleshoot problems. With the non-spoon-feeding concept in mind then... Adjusted that - "tunnel effect" may or may not come into play. It's more likely to occur with very short focal lengths. Not if you maintain the same aperture, as specified by the "f-stop". Lens aperture has no relation to what the lens is mounted in front of. Aye, and there's the rub with generalities... As a general rule, this is true, but you have to trust that the lens manufacturers are being completely truthful as well. For example, it's not uncommon in many fields (and I expect this is true with cameras and sensors as well) that one manufacturing run is done, the lots tested, and those that don't pass QC are rebranded as a model of lower capability. Two lenses could roll off the same production line, with one passing QC and being labeled for 1/2", while the other tests soft around the edges and is therefore designated for 1/3", where the edges of the image won't hit the sensor... the two cost the same to make, but the 1/2" model will naturally cost more. Obviously this isn't true for ALL manufacturers or lens designs, the same as it's not true for ALL cameras, or anything else... but you know, if you want the spoon feeding and choose to believe that there's inherent superiority to 1/2" vs. 1/3" lenses, well... that's up to you.

Unless, of course, you prefer to learn HOW stuff actually works rather that blindly relying on what vendors tell you you're "supposed" to use, especially ones who don't appear to know what they're talking about.

IR *is* light... just very long wavelength on the verge of human vision.

If you actually look at and factor in the rated maximum aperture of a lens, there will be no difference. I get the impression though, a lot of CCTV people don't actually know what an "f-stop" is, or what it relates to, and as such don't often bother to look at that spec on a lens and take it into account. Lenses designed for 1/2" sensors MAY generally be a bit larger in overall construction, which COULD equate to some designs having a larger aperture by default, which I could see easily leading to the impression amongst installers that 1/2" lenses *as a whole* allow more light than 1/3" lenses... but that impression would be mistaken. Alas, such false impressions have a habit of propagating, in ALL professions... and are hard to squash.

Well, these ARE the people who seem to think it's a good idea to auto-install their own NVR along with their camera-management software, AND have it auto-start, without giving the user any option during install to avoid either occurrence... and who actually HAVE a small, lightweight, standalone camera-setup utility but don't make it readily available to end users... so... yeah, forgive me if I don't have a hard time believing they don't understand lens technology. That's your choice. What Arecont is telling you contradicts what my 30 years in photography has taught me about lenses. That doesn't even make sense. Lenses don't have pixels, so where does this "larger pixel size" come from? They are completely out to lunch with that (assuming you're quoting them correctly). The "mm" is the focal length of the lens. That measurement is 110% unrelated to the size of the sensor. A 12mm lens has a 12mm focal length regardless of whether the sensor is 1/4", 1/3", 1/2", 1", 35mm, 4x6", or 8x10", or if there's no sensor at all. Unless you're severely misquoting them... no, I'm sorry, they have NO idea what they're talking about. Unfortunately, that group doesn't appear to include the Arecont trainers.

Not so - at least, not directly. There is one measurement that indicates the amount of light allowed: the f-stop, or aperture size. This is a direct ratio of the opening size to the focal length - f/4 on a 20mm lens will be a larger opening than f/4 on an 8mm lens, but it both cases, the same amount of light is passed. If you have then 8mm lens, rated f/1.4, and one is designed for 1/2" sensors, one for 1/3" and one for 1/4", all will allow the same amount of light. Wouldn't the effective aperture be higher when using a 1/2" lens on a 1/3" imager, due to a portion of the light gathered by the lens being projected beyond the edges of the imager? No... at most you'd just get a cropped view. The light transmitted by the lens is the same. The area that IS imager still gets the same amount of light either way.

Not so - at least, not directly. There is one measurement that indicates the amount of light allowed: the f-stop, or aperture size. This is a direct ratio of the opening size to the focal length - f/4 on a 20mm lens will be a larger opening than f/4 on an 8mm lens, but it both cases, the same amount of light is passed. If you have then 8mm lens, rated f/1.4, and one is designed for 1/2" sensors, one for 1/3" and one for 1/4", all will allow the same amount of light.

Any number of video-capture devices should work, but yeah, a Slingbox is probably the easiest by far. Good idea, tom!

Because maximum resolution of 99% of the video-encoding chipsets on the market IS 604-480 to 704x480, so the DVRs CAN'T be set any higher. To go beyond that would require a new chipset, which would mean redesign and retooling of existing cards and software. That would equate to a very expensive hardware upgrade to support MAYBE 10% higher usable pixel count. Meanwhile, megapixel camera prices are still dropping fast - 1.2MP is four times the resolution of 4CIF/D1 for very little extra money and no additional hardware required... it's just not cost-effective to substantially redesign video-capture hardware for such a minimal improvement.

bpzle, I know what you mean! I used to be in IT as well; back in 2000, I interviewed and was hired for one job, and they told me that if I had my MSCE, I would be "worth" at least $5k more per year... well... I ended up working in a room full of MSCEs and could tech most of them under the table any day. One couldn't figure out how to get an internet connection working on a Mac, because he didn't actually understand TCP/IP. Another didn't know how to change out a dead floppy drive... and this was in an IBM computer that didn't even have screws holding it in, just flip a couple levers and pop it right out. Some of the techs were really good, to be fair, but most of the ones who waved their MCSEs as some sort of badge of honor were useless if what they faced didn't match their Microsoft-specific training. And yet, someone had determined that those letter meant they were "worth" more than me.

That's probably the difference there: there's probably no issue with DIY installs, and hey, there's nothing wrong with you helping your buddy out if he's doing his own install

Glad you got it sorted out You'll have to excuse bpzle, he sounds like he's recently had a run-in with some "trunk-slammers", guys with no knowledge of the industry who figured it was just a good line of work to get into and come swooping in underbidding everyone in sight. Actually, your story is pretty common around here recently: IT guys who, since they're "technically oriented", are being asked by their companies and customers to start doing their cameras as well. You have an advantage as well, since more and more cameras are going to IP rather than analog video, and your network and systems knowledge will come in handy for that.

Yeah, that's the main limitation, "current" PoE spec doesn't provide for a lot of current. New spec does, and is slowly making its way into the market, but expect to pay more for it.

I've very rarely used the baluns with the power pass-thru... mostly because they're substantially more expensive and other than convenience there's no advantage to them. I'll typically just use the blue pair for video with a standard GVI or GEM balun (Muxlabs are WAY more expensive, and I've never noticed a difference in video quality), orange and green for power (either split out of the jacket, or extended with 18/2 or station wire if necessary), and save the brown for future use, although it could be separated and added to the orange and green runs for a little more current capacity as well. Nothing wrong with the power supply - I believe that's a 24VAC, 16-channel, UL-listed model with circuit breakers on each channel? You could possibly go a little cheaper with something that uses glass fuses, but they're functionally the same. Local building codes may dictate whether Class 2 is sufficient or UL-listing is required for this purpose.

+1. +2.... Look for the Phihong PSA16U-480 for single channel 802.11af compliant PoE injection, you can find it cheap enough to not need to bother with what you are describing. I found those on eBay for $100... this by you is "cheap"??? I can get a switch with eight 10/100 PoE ports and two gigabit ports for $300!

Nothing wrong with those stats, especially for a four-input card. "Decent" may be pushing it... from the description they sound like pretty generic mediocre offshore stuff. That may be part of your problem. Higher resolution. Yours only does up to 352x288, according to the specs you listed. Maximum for analog video is 740x480, so you're only getting about 1/4 the resolution you should be able to. No, that should be fine. Your existing PC should be fine. You can go to a standalone, or a better card - just make sure it's spec'd for 640x480 or higher resolution, or 4CIF or D1. I suspect you'll find the cameras are half the problem as well, though.

Welcome back! You are correct, it would depend to some degree on the DVR - some allow you to divide the total FPS among the cameras as you desire (for example, you could have two cameras at 20fps, two at 5fps), while others will divide the total evenly across all channels and that's your maximum per-channel. In your case though, either way, you should be able to get 25fps each with only two cameras.

+1.

Yikes! That looks like the camera is failing...??

Attachments are broken - you need to upload the picture(s) to an image hosting site like TinyPic, Imageshack or Photobucket, and then link to that.

Could you provide photos of how this "noise" looks onscreen?

Thanks for reply. A combo modem/ router is present at the client's/ video server site. Have you found that using a separate modem would allow connection where it couldn't occur via the combo router? Not as such, but it would allow you (or the client) to plug the encoder directly into the modem so it has an "outside" IP and gives you direct access to it, thus possibly eliminating the router and port forwarding as a problem.

There's lots of other bandwidth and storage calculators out there, but this one seems to be the most complete, and popular, combining just about all the necessary functions for system design in one place. But again, I go mostly by experience, in figuring storage needs as well as lens selection, so as nifty as this software is, I don't really use it much.

Sounds like you're doing everything right so far. Try setting the device's IP in the router as "DMZ" (Demilitarized Zone) - essentially that completely unblocks that IP, and any ports that aren't forwarded somewhere else will go there instead. That should remove any firewall issues as well. Offhand, I can't think of a problem, unless the ISP on the server's end doesn't allow incoming connections *at all*. Is the router in this case a combination broadband modem/router? If not, try bypassing the router and plugging the server directly into the modem.

I generally don't bother - I can usually give a rough estimate based on experience, but the reality of recording varies so widely, that's probably about as likely to be accurate as any actual calculations. If you want a really good handy calculator, check out the offerings from http://www.jvsg.com/ - it will let you plug in most of the related factors (scene complexity, compression type, framerate, resolution, expected percentage of scene change and motion, etc.) to give you a fair idea.