survtech

DVTEL is not the only company making encoders and decoders. Pelco has the PelcoNet series, Bosch has the VIPX line and Optelecom-nkf makes a full line. There are others too. Keep in mind that if you convert analog to IP, your video is limited to analog resolution, whereas IP cameras can be megapixel.

Not so amazing. I've seen ground loops through wood telephone poles.

Check for IR interference. Sunlight is notorious for getting into and overloading sensors during the day. Active IR motion detectors and other IR sources can also interfere with IR remote operation. Try covering or blocking the sensor in the DVR except for a small window and aim the remote directly at that window. Worse case, get an IR meter or try an IR Indicator Card to find interfering IR sources.

If you're certain the polarity of the baluns is correct, more than likely the problem is weak sync which usually indicates high resistance or a major impedance mismatch. You can test polarity by reversing the wires at the balun on one end. If the picture becomes viewable, you had the polarity reversed at either end. With twisted pair, it is not important that any specific wire is (+), just that at both ends the same wire is (+).

The voltage is a little low. Most authorities recommend no more than 10% drop. You have nearly 12%. But that still doesn't explain why you are getting a useable picture on a closer monitor. Check your connections. Make sure your baluns' + and - are connected to the same color at each end. Try powering the camera locally. Bad baluns are extremely rare and 280' is well within the distance limitations for passive-passive. If you can, try swapping cables between the two cameras. If the problem follows the cable, that's the source; if it follows the camera, that's the source.

If you mean the CM9760-MDA, it has 4 outputs per input by 16 inputs. Their DA104DT is a only 1x4.

I can't believe Surveillance isn't tied into the system. Who records it, the pits? The table looks like craps or roulette, not BJ.

What kind of table is that - blackjack? And the players can't touch the cards? Not the case in California, so it wouldn't work here.

Kramer Electronics makes DA's with up to 54 composite outputs. On many of them, the inputs can be daisy-chained to provide an unlimited number of outputs. See here: http://www.kramerelectronics.com/products/?viewlistas=table.

We have that. There is a camera located under the rail at each seat. After the bets are placed and before the hand is dealt the dealer presses a button on the table taking a snapshot of the player's bet. If there is a payout discrepency, the pitboss can bring up the bet amount on an LCD screen from the pit. From the overhead table cameras it is nearly impossible to determine the bet amount. You can tell the chip's value (red, green, black) but you cannot tell how many are in the stack. There is another system for determining the chips in a stack that utilizes RFID devices embedded in the chips.

We tried that. The only place that had a chance was our main entrances. Unfortunately, we have wide entrance doors and a fairly dark, narrow vestibule between the inner and outer doors. That makes the backlighting a killer. We've tried every WDR camera on the market and we can't get a useable image during the day. Even if we could solve the backlighting problems, our experiments have shown that we would only get useable images of a small fraction of the customers entering the casino. Elsewhere, the high ceilings make the camera angles too acute to be useable. Yes, masked bandits would stand out like sore thumbs, but they are not the problem - sunglasses, hats and backlighting are. Megapixel cameras show some promise and we plan to experiment with them in the future but we don't have the IP infrastructure and the entrances are too far away from Surveillance to run straight CAT-5 so we will have to install a switch somewhere in between. Also, our DVR system will only transcode Axis and Arecont IP cameras. If we can't record the cameras, they are useless to us. But all of that is moot with the Biometrica system. It uses a Euresys Picolo frame grabber card to capture analog video. The capture must be manually triggered by an operator. It will not analyze streaming video, either analog or digital. Even when an operator is using the system, he/she must watch the subject live until they are facing the camera, then initiate a capture. This is somewhat useful at table games but totally useless at entrances.

Exactly. We use Pelco, Ganz, Ikegami, Dallmeier and, occasionally other brands, depending on the application. There are no camera manufacturers whose products are ideal for every situation. By all means, try Bosch. But don't preclude others.

One of the main issues with drive longevity is platter bearings. Cheap and older drives use(d) ball bearings, which tend to wear out due to metal-to-metal contact friction. Newer drives use fluid dynamic bearings, which eliminate metal-to-metal contact. Another primary cause of drive failures is the "PZT actuators" that move the heads. Still another cause is vibration and physical shock that can make the head hit the spinning platter. Most drive failures are mechanical, not electronic in nature.

Facial recognition is a hit-or-miss (mostly miss) proposition. This is especially true in a casino where it is difficult to get a face shot from directly in front of the subject. Usually the angles are poor (cameras in high ceilings, etc.) and there are often lighting problems (backlighting or very dark areas). Accurate facial recognition requires that the subject be at the same angle to and approximate distance from the camera, with similar lighting, etc. as the original image was taken and looking directly at the camera. This is nearly impossible unless you can totally control the subject's location, like in a waiting line or a narrow doorway. Also, the wearing of sunglasses or anything else that prevents the camera from measuring the distance between the pupils of the eyes will prevent facial recognition software from working. We have been looking at this field for years and have come to the conclusion that it is not a viable technology at this time. That said, the Biometrica databases and SIN are worthwhile for casinos.

The video would be no problem. A simple method of terminating the cable would be to use "F" connectors and an F-to-RCA adaptor. Not the best method but you won't need special crimpers. The power is another thing. There would probably be too much voltage drop with 9VDC in 18 gauge at 300 ft. I'd recommend 16 gauge stranded power cable. You can also obtain that in direct burial. West penn part number AQ225: http://www.westpenn-wpw.com/index.php?option=com_cable_finder&func=PDF_Info&download=AQ225

You will need to run "direct burial" cable. 300' is well within the capabilities of RG-59 but you will have to determine the requirements for power cable by knowing the power requirements of the camera(s). Pelco has a wire gauge calculator for power wire here: http://www.pelco.com/support/tools/wiregacalc.aspx. For direct burial cable it is still advisable, but not absolutely necessary, to run it in PVC conduit. West Penn and others make direct burial cable. Here is a link to West Penn's coax cable finder: http://www.westpenn-wpw.com/component/option,com_cable_finder/func,SearchSec/section,Coaxial_Products/.

If they are recording 7/24/365, they should probably be replaced even more often. Sanyo recommends every 2-3 years for their DVR's. It also depends on the type of drive. Many DVR's use lower-quality desktop computer hard drives that aren't made for continuous use and will fail earlier than SCSI or SAS or even RE drives. The other problem is that most DVR's can't use either of these more robust types.

RCA = RCA = RCA. There is no such thing as a 75 ohm RCA connector. They are all just RCA connectors. Although impedance is somewhat critical for video, it is not so critical that connector types have a major effect. An example - many inexpensive outdoor cameras use short 18/4 cable to bring the power and video out of the enclosure, then attach a BNC female to one pair. That is perfectly acceptable if the cable is short. The same thing goes for 50 ohm vs. 75 ohm BNC connectors.

Here are some links to both online and hard copy CCTV training: http://www.atvideo.com/support/training.asp http://www.naaa.org/index.php?s=18 (also available at http://www.ltctrainingcntr.com/00_Default.htm) http://www.stamweb.com/home.html http://www.securityceu.com/shop/home.php?cat=262

You're welcome.

TDR = Time-domain reflectometer, a device for checking cable impedance. Not normally necessary if you use high-quality cable and follow good cabling practices. Useful for solving tough problems. Also, not very cheap - typically $200 to several thousand $$$.

Anything to make a sale. Still AFAIK, there has never been an instance where a streaming compression codec has been successfully been challenged in court. That includes all of the MPEGs, Wavelet, h.261/263/264 and MJPEG.

So, you are saying that you are sending the signal (750' + 850' = 1600' total)? That's way too far for passive-passive. You will have to use active receivers or transmitters. Also, it is not advised to use beans to splice cables. Use either punchdown blocks or patch panels.

Who says MPEG-4 (or any MPEG for that matter) is not admissable in court? I've never heard that. Most casinos use MPEG in one form or another to record their cameras. We would not use anything that would not hold up in court. When you are recording 1000+ cameras at 4CIF or D1, MJPEG is not an option. The cost for the 250TB of storage we needed for MPEG-2 was close to $1M. MJPEG would require at least 3-4 times that.

Why would you want PTZ's? You realize that someone will have to be there to aim and zoom them? Also, PTZ cameras tend to be very expensive compared to fixed cameras (often 10x or more).