Looking for a Wireless Video Camera

On Tue, 13 Mar 2007 13:29:12 -0500, jim evans wrote in :

or

On Tue, 13 Mar 2007 20:08:41 GMT, John Navas wrote in :

Less expensive but still very good:

No, there are many more. See:

Unfortunatly, I have no idea what are your requirements, so no specific recommendation.

My favorites are the various camera servers, using either video camcorders, NTSC video cameras, or digital cameras. The optics and technology are generally superior to any of the golf ball type of cameras. If you need wireless, just add an "ethernet wireless client bridge". Yeah, it's more expensive but if you want quality images, it going to cost more.

Thanks for the replies. I wonder why I didn't see these in my google search?

-- jim

Thats pretty accurate, I know someone who setup those low end cameras in their building and they were all excited when they could pop a picture up in a web browser not realizing how bad the quality was. They had a break in last week and it was impossible to make out any real details.

What was your Google search? Try something like (wireless OR wi-fi OR "802.11g") network camera

p.s. Please don't switch posting styles (top vs bottom) in mid-thread

-- it's confusing. Thanks.

On Tue, 13 Mar 2007 22:10:49 -0500, jim evans wrote in :

George hath wroth:

One of my customers is a local security outfit, that sells cameras and recorders. The owner is fairly technical and responsible for selecting and purchasing more junk from obscure sources than I've ever suspected existed. In frustration, he once declared the entire industry to be a huge conspiracy to unload useless hardware and exhorbitant prices. In a fit of temporary insanity, I volunteered to find a usable combination, at a bottom of the line price. I dragged in a cheap 4 camera server

and started experimenting with anything I could find that would belch NTSC video. I displayed each camera in 4 quadrants on the screen and invited the employees to vote on their favorites. The winner was various high end digital cameras with automatic iris, auto focus, and a wide focal length lens. 2nd best were various camcorders with roughly the same features, but fewer megapixels. Absolute bottom of the line was a USB CMOS camera which required a seperate camera server for the test, and looked terrible.

However, I got a few suprises during the testing. Lighting was critical. There's no way to get a decent picture of a person entering a store if the lighting through the windows goes right into the camera lens. They look like a black shadow moving through the offices. That's why you want to mount cameras as high as possible near the ceiling.

Another suprise was the effect of software. One of the packages I tried had a nifty feature, where it would detect motion and increase the frame rate to 60fps (not 30fps). It would also bracket the exposure and focus. There were some other enhancements which I don't want to go into. The idea was NOt to give the best compromise picture, but rather to give the best series of frames which could later be decomposed into stills. The results were impressive. Although some of the photos were fuzzy and over/under exposed, there were a few that were absolutely perfect.

There was also a nifty system that would follow a moving target through the office or store. That was fun to play with, but resulted in a series of fuzzy pictures as the image capture wasn't fast enough. I think a bigger lens or faster server might have fixed this problem.

One observation, that I'm not sure is correct, is that the bigger the lens, the better the picture when inspected frame by frame. It might be a possible improved depth of field, but I never bothered to measure or calculate the depth of field.

I'm still learning (by destroying) how all this works. One lesson is clear. One does NOT get a decent image from a $30 CMOS camera, with a junk lens, built into a $60 wireless bridge or server.

On Wed, 14 Mar 2007 10:07:54 -0400, George wrote in :

What? You mean they didn't know about the software Hollywood uses to magically improve these images on crime shows and movies? LOL Humor aside, it actually _can_ be possible to improve images from security cameras when multiple frames are taken of a given static target, algorithmically comparing frames to remove noise.

On Wed, 14 Mar 2007 09:55:13 -0700, Jeff Liebermann wrote in :

When there are multiple frames of the same static target, software can greatly improve an image by over-sampling, algorithmically merging multiple frames into a single image.

What matters most is the sharpness of the lens, not the size. Low end cameras tend to have cheap crap plastic lenses that produce crap images.

The _diameter_ of the lens is related to both sensor size and light gathering power, with larger (wider, faster) lenses doing a better job in low light. The drawback of a faster lens is shallower depth of field, making focus more critical.

Cheap cameras tend to use small cheap sensors behind fixed focal length lenses with shallow depth of field, resulting is blurry images, and relatively slow speed, resulting in noisy images. Better cameras use larger and more sensitive sensors behind faster auto-focus lenses, albeit with some risk of the camera focusing on the wrong target.

The _length_ of the lens is related to mechanical zoom range, with larger (longer) lenses have a longer zoom range.

Agreed. I'd say the absolute minimum is about $100, with $200 (Panasonic BL-C20A) - $300 (Panasonic BL-C30A) for good results.

Unfortunately a lot of people do believe that anything electronic is "magical" and the only important issue is to find the cheapest thing they can get their hands on assuming because it is "magical" it will work. There are certainly things that can be done with DSP techniques but there isn't a lot you can do with images from the low end "Walmart class" cameras.

George hath wroth:

Yep. I even promote that impression on service calls. I wear my wizards hat, cloak with embroidered stars and crescents, and carry a magic wand (actually a 12" Craftsman #2 Philips screwdriver). The idea is to give the customer the impression that everything I do is incomprehensible, unfathomable, and magical. If they ever figure out that it's mostly RTFM, I'll be out of business. Meanwhile, magic pays.

Yep. Any sufficiently advanced technology is indistinguishable from magic. (Arthur C. Clarke). Anything else can be returned for credit.

Under ideal conditions, a cheap camera works well enough. It's the marginal and often uncontrolled conditions presented by a security camera that offer the major challenges. Security cameras were intended to operate without adjustments during the day. That means auto focus and auto aperture over a wide range of distances and lighting conditions. Methinks expecting a cheap camera to work under these conditions is expecting too much. However, even the high priced cameras fail under some conditions common in security cameras. For example, trying to adjust the aperture to catch a picture of a foreground subject, with overwhelming background lighting. Same with someone shining a flashing into the lens of a security camera. As for quality, most security systems are rated at how many hours of videos can they store. If you want more resolution, you pay for it in hours. Hours can be understood by anyone, while resolution is technobabble. So, the security vendors push hours, not picture quality.

On Thu, 15 Mar 2007 09:41:40 -0700, Jeff Liebermann wrote in :

Price is a pretty reliable guide to quality. Good systems aren't cheap.

Happens all the time...dumb ass people buy cheap cams for $100 a piece and expect to have the same imagery of a $700 (dealer cost) cam you find in a bank.

Thats not the point I was trying to make. Most people wouldn't go to Sears to buy an air compressor for their factory or Home Depot for a truck dock because they know there are different classes of equipment. Yet they don't hesitate going to "barginfinder.com" to buy electronic gear thinking they somehow "you get what you pay for" doesn't apply.

Unless the only intent is to see if "fluffy" is on the couch a cheap camera is worthless (or less, because they actually think they have a security system) for any real security use.

It's the

George hath wroth:

Actually, there was a time when Sears sold excellent power and hand tools. I still have my collection of Craftsman hand and power tools from the early 1970's. However, you're correct about their air compressors and recent offerings. They're not very good or long lasting.

That's because they don't understand (or don't want to understand) the underlying technology. That's not easy and certainly a problem without substantial experience and some inside info on the performance of various acronyms and chipsets. To the average wireless buyer, the price, packaging and aesthetics are about all they are able to use to use as a decision making criteria. The latest sales paradigm is to make the access point look as weird as possible on the assumption that it would impress the customer. For some vendors, simply repackaging the same old circuit board in a new plastic package offers a means of stretching their investment. DLink does this ever so well. Linksys went a step further and simply changed the model number of the WRT54Gv4 to WRT54GL and raise the price. Inside, they're identical.

In some cases, spending more money does give one more features and functions. Then, the problem becomes whether one needs these features. The average home user would probably be satisfied with a fairly crude feature set. The average corporate customer wants the kitchen sink and then some. Spend more money and get more features, but if left unused, they're wasted. How many home users need SNMP monitoring, dual SSID's, virtual LAN's, roaming, and complex access control lists?

I once had a conversation with a former bank network admin who is also somewhat involved with the banks electronic security. The problem is that a really good security system costs more than the potential loss from bank robbery and theft. There's really no incentive for a bank to install a really high quality video system. A fairly simplistic minimal system offers considerable return on investment. Incremental increases in resolution, performance, retention, sensitivity, and such, generally only adds complexity and cost, with a proportionately smaller improvement in overall security per dollar expended. Better cameras, mean more data to be stored, which require bigger hard disks, etc. Complex integrated systems with a large number of sensors are difficult to keep running. For example, it would be trivial to setup a 3D system using 2 cameras to give bipolar vision (kinda like the viewpoint rotating special effects seen in the movies). However, doubling the cost of the camera system would not yield a proportional increase in security and would probably result in increased operating expenses.

Obviously, a home system isn't going to be as expensive or as complex as that used by a bank, but one wouldn't know that looking at the advertising literature. There's plenty that can be done with relatively crude and cheap hardware. If that's inadequate, there's always the never ending upgrade path.

If you need something to do, figure out how to use a wireless access point as a motion detector. Most of the parts of the puzzle are there.

On Fri, 16 Mar 2007 09:34:37 -0700, Jeff Liebermann wrote in :

I think it's more because they assume technical products are a commodity. That's because no technical product vendor has done a decent job of product differentiation. People are willing to pay more (much more) when they perceive a product to be better, even when they don't understand why, and even when the products aren't actually better. Cases in point: Bose Wave Radio. Monster cable.

Actually not -- a crap lens gives crap results. I can easily demonstrate the real difference between a cheap camera and a good camera, even under ideal conditions. For example, I recently took a portrait with my Leica lens, and the subject was literally blown away when he saw how sharp it was. "My God! I can actually see each and every hair in my beard!"

Agreed on all points. Radio is magic and magic sells. Electronics is art, and making the package artsy also sells. Acronyms and standards are great. Every company should have one (or more). Model numbers are like money, the bigger the better, and large or long model numbers sells. Add a few hyphens and it's even longer.

I really like the technobabble behind Monster Cable products. I don't think I understand any of it. However, that's small potatoes compared to the vendors that cater to the audiophiles. Think $400 power cables.

My crystal ball and Ouigi Board both predict a grand future in selling technology using technobabble, hype, and methods commonly found among used vehicle salesmen. The current muddle over HDTV standards is the current example. Some day, the IEEE will approve the multitude of standards that they've been working on and we have even more muddle.

Product differention? You must mean do you want it in black, gray, beige, or fire engine red. Yeah, I can see the difference between products easily. This one, with the antenna that's twice as long, must be twice as good. The one with 4 antennas must be 4 times better.

Ok, I'll concede the point. I used to own a Leica IIIc with a fairly complete set of screw on lenses. I would take a roll of 35mm film, shoot the first few frames in the Leica, rewind, move the film cart to a commodity SLR camera, skip forward, and shoot the rest of the roll of roughly the same subject matter, with the same lighting. When developed, the image quality was very easy to compare without any subtle differences created by processing. Yeah, if you know what to look for, quality is obvious.

However, if you ignore the very bottom of the market in cameras, it becomes more difficult to differentiate the quality of some of the more expensive imagers. Each tends to have one or two areas in which they are superior, at the expense of others. For example, I've seen low light cameras that have impressive sensitivity, but look awful in normal lighting. Go even higher in quality, and the products tend to become special purpose. I guess the moral is to know the technology before buying.

I must be slipping. Think $1,000 power cord:

That gives me an idea. The market for replacement antennas is already saturated, so I'll sell aftermarket wall warts. Something like a "MegaBoost Super-Regulated Ultra Clean Power Supply". Voltage regulated to 0.001% of perfection. Filtered to insure the 60 Hz hummmmm is down at least -50dB and 120Hz hummmmmm at least -3dB more. Fully shielded and guaranteed not to emit any stray magnetic, electric, radio, acoustic, or ethereal fields. Using my power supply, you'll have the cleanest Wi-Fi signal in the neighborhood, when used in conjunction with Wi-Fi speed spray:

Quality and buzzwords sell so I guess I'll price it slightly above outrageous.

On Sat, 17 Mar 2007 00:40:19 GMT, Jeff Liebermann wrote in :

Agreed. As always, get the >> right tool for the job