survtech

Brent, The image on the right is taken by a 5MP IP camera, not an analog camera. The one on the left is about the best any analog camera can do, no matter the TVL claims of the manufacturer. Regarding your sharper images, likely your first DVR was limited to CIF, which is 320x240 (320x256 PAL). Also keep in mind that the 480 line spec would be vertical resolution (essentially of horizontal lines), rather than horizontal resolution, which is measured with vertical lines. I know that can be confusing but think of it as individual dots, rather than vertical lines. Vertical lines generate individual horizontal dots on an imaginary horizontal line and that's what is being counted. It's just easier to see vertical lines at some horizontal point than to just show one row of horizontal dots. Perhaps a better explanation is that you should consider the vertical lines on the test pattern as a series of horizontal rows of dots, where each row lower has the dots slightly closer together.

Although I often disagree with thetoss, in this case, I wholeheartedly agree. There is no validity to analog CCTV claims of higher resolution than ~600TVL. So-called 1000TVL and 1200TVL cameras are manufacturers' hype based on the number of pixels in the sensor, not the actual image resolution. They cannot be duplicated in real life. TVL is, by definition, the maximum number of distinct vertical lines that can be counted on a monitor in a square area where the height and width of the area used is equal to the total screen height just before they blur together. In practice, the camera under test is aimed at a test chart of increasing vertical lines per inch like below: The number next to the place where individual lines just start to blur together is the TVL resolution of the camera expressed in hundreds of lines. The chart above only shows part of the total chart, which must cover the entire screen as seen below: In the top picture, Camera 1 is a typical analog while Camera 2 is a 5MP IP camera.

Probably not. Maybe the factory can help but that's why the firmware and even Operating System updates warn "Do not turn off ."

You also may be able to put a power cube on the top of the elevator. Many have outlets there and it would save you a pair in the stringer cable. You should be able to use one pair for video but be aware it might be noisy. UTP transmission requires twisted pairs, preferably with a high number of twists per foot. Elevator stringer cables have little-to-no twists in their pairs and adjacent cables can carry relatively high voltages and electrical noise which will get into the signal. The key advantage to twisted-pair cables is that each wire receives an approximately equal amount of interference which is cancelled out by adding the signals out of phase. Untwisted pairs will tend to have different amounts of noise on each wire so the result won't be totally cancelled out by the balun or active receiver. One possibility might be to use an IP camera and send the stream through an EoUTP adapter.

Nine. Plus three Monitor Wall clients. I've run at least 25 instances of the same camera on multiple clients with no problems. We've also run the same camera simultaneously on all 12 clients with no problems.

That is a major difference, IMO. All IV encoders and cameras can deliver multiple (up to 8, I believe) 30fps streams in any combination of TCP/IP, RDP or UDP Unicast and Multicast. Their servers don't normally have to redistribute the streams to clients due to this, except when they have to accommodate other manufacturers' single-stream-only capabilities.

I believe the maximum is 400Mbps but we chose the conservative route. At <200Mbps, server resource utilization is quite low: as I recall CPU load is in the 20% range and memory and network utilization are also on the low side. We are using their client. IV cannot re-distribute streams without using another set of servers named "Proxy Servers" so recording is via Unicast and the clients see the "live" Multicast streams from the encoders or IP cameras. Control of viewing stream distribution is via multicast groups.

We're connecting approximately 63 cameras per server with a mix of 2.5Mbps, 3.0Mbps and 3.5/5Mbps depending on camera application (3.5/5 is PTZs) so that works out to about 200Mbps per server. We also had to consider retention time versus storage so for the cameras we retain for 60 days we only have 13 cameras per server. MP cameras are set to stream at 4.0Mbps and their servers are handling around 47 cameras each. As we add cameras, we will determine which server to record them on by available storage. Some servers have more than enough bandwidth but not enough storage to guarantee our desired retention time. I also haven't tweaked the encoders and IP cameras yet. I will be adjusting GOP values, bitrates and other settings once retention has stabilized and I have a better handle on what we need from each camera. I don't see any need to use high bitrates to cover a drink station but may want to increase bitrates elsewhere. I'm also experimenting with GOP settings / I-frame intervals and their affect on fast playback smoothness and quantization noise. Not exactly. IndigoVision uses a range of hardware depending on the application. For casino applications, they're currently deploying Dell PowerEdge servers and PowerVault storage.

Right now, we have 67 IndigoVision 11000 HD cameras in operation and another 33 in stock. Performance is pretty good but not excellent. Tests of Axis P3354, P3364 and P3384 fixed dome cameras yielded somewhat better pictures but with the caveat that they were unable to provide two 30fps streams. During system evaluations, we also tested a Ganz and a Bosch fixed dome. We received the Bosch dome too late to test with IndigoVision but did test it briefly with Geutebruck. The Ganz's picture was not very good - it was noisy and had AGC issues. The Bosch locked up the VMS so that the server kept rebooting. That was not specifically a Bosch problem but an ONVIF driver issue with Geutebruck. We haven't tested any PTZs in situ yet except the one IndigoVision 9000 SD PTZ they provided. We are not terribly impressed with its performance, hence the desire to test others. We do have a trade in offer from IndigoVision which expires at the end of the month. They're offering to give us a trade in on encoders for IP PTZs. Not one-for-one but tempting if we can find an IndigoVision PTZ we like. The gist of it is that we chose IndigoVision fixed IP cameras due to their guaranteed compatibility and the limited time we had to test IP MP cameras. While the 11000 HD cameras didn't provide the best pictures during our system evaluations, they did provide more than acceptable video. Despite a few lens problems, we're pretty happy with the results.

It already is. That said, IV recommends TCP/IP Unicast for recording due to its inherent stability and easier troubleshooting of the stream. While we could both record and "live view" via Multicast, we prefer (and are basically encouraged) to limit it to live viewing.

We're currently testing IndigoVision's 9000 and 11000 series and have been offered Pelco Spectra HDs by their rep. We've requested an Axis Q6044, whatever Bosch's and Sony's reps decide to provide and, I think, one from HIKVision (whose rep approached me). We're still open to suggestions but to be considered, PTZs will have to be able to provide at least two 30fps ONVIF streams (one TCP/IP Unicast and one Multicast). We'll be evaluating overall picture quality, frame rate, system compatibility, low light performance, ease of installation, etc. We will also inspect, as much as possible, physical construction and repairability. To give you an example of how thoroughly we approach testing, one sample (manufacturer shall remain unnamed) failed our evaluation before we even plugged the PTZ into our test setup. A physical inspection of the unit revealed many shortcomings, including an overly complex installation method that required attachment by many screws and open slip rings. Who wants to be standing at the top of a ladder trying to feed screws through the back box into a plenum box? That requires access to both sides of a ceiling tile or hard lid and juggling screws that are destined to fall and disappear on the floor. And my years of experience make me averse to non-sealed slip rings. I've had to clean or replace them too often.

No, but they can usually answer technical questions and/or point your tech to the correct parts to solve a problem. My PTZ Tech learned a bit from me (Bench Tech background), some from people at TLS, some from examining invoices from Moore's and the rest OJT. He does get stumped on rare occasions but usually figures it out with or without help by now (~5 years). It will be interesting to see where we can take PTZ repairs as we start deploying IP PTZs. TLS doesn't have any parts, except for maybe Spectra IV Mini IP models so it may be difficult obtaining them for many brands. That is among the questions I will have when evaluating IP PTZs - are exploded views, parts lists and the parts themselves readily obtainable? We rarely buy extended warranties for even expensive cameras (and IP PTZs range from $2,000 and up) and standard warranties are typically quite short (typically 2 years) so we don't want to get bombarded with high repair costs when they fail out of warranty.

You should consider repairing them. We used Pelco Repair for a while but found their pricing exorbitant so we switched to Moore's Electronics for a few years http://www.cctv-repair.com/. One of my techs volunteered to learn to repair PTZs and TLS Electronics was very helpful. They worked with him to troubleshoot most problems (most repairs are either simple mechanical or replacement of entire sub-assemblies). He has become an expert and by doing our own repairs we have saved many thousands of $$$ over either Pelco or Moore's. We set up a test bench with a pendant back box and IWM-GY mounted on a 2x4 frame, made a long male-to-female extension cable for the multipin connector and use a spare 9760 joystick set to directly control the PTZ and a monitor for testing. The entire test jig cost a few hundred $$$ and has paid for itself many times over. That won't help you with the Spectra II's because many of their parts have been discontinued. That is the primary reason we chose the same route - replacing them with Spectra IV's. We started that last year and have replaced around 30 so far and have budgeted the remaining 40 for 2014. We'll still repair Spectra III's and IVs. In fact, that is a consideration we'll have if we decide to switch to IP PTZs. Many of them are impossible to find repair parts. Considering the relatively short typical 2-year warranties on PTZs, we would prefer to buy brands where we can obtain replacement parts easily.

Good luck finding a buyer for the Spectra II's. We've tried to find buyers for our surplus equipment for years with no luck. Since we also repair our own PTZs, we have resorted to cannibalizing what parts we could from Spectra II dome drives that fail and replacing them all in the long term. We are down to about 40 in production and have budgeted to replace those in 2014. One issue is that analog PTZs look like crap when fed through encoders. This is exacerbated by AGC in low light areas. To that end, we are debating whether to purchase Spectra IV's at around $1,100 each (SD423-F1 for most indoor applications) or bite the bullet and pay $2,000 and up for IP PTZs. On that note, anyone legitimately interested in buying a complete 1088x64 Pelco 9780 matrix system including CPUs, hot switch, patch panels, ribbon cables, keyboards and accessories should PM me.

TLS Electronics (Time Lapse Supply) sells "remanufactured" PTZs, including Pelco, AD, Kalatel and Ultrak. http://www.tlselectronics.com/ptzdomecamera.html Prices start at $545 for the dome drives. Keep in mind that Spectra III and IV dome drives don't work well in Spectra II back boxes due to the differences in cooling systems - Spectra II's had the cooling fans in the back box while Spectra III and Spectra IV have either fans or passive cooling located within the dome drive. The Spectra II cooling holes don't "line up" with Spectra III and IV holes. Because of this, air circulation is restricted between models. Spectra III and IV are almost totally inter-compatible. I strongly suggest replacing at least the back box mount and door assembly or the entire back box. TLS has been very helpful to our repair technician with questions about parts and service. Pelco used to have exploded views and parts lists on their website but it appears the documents have been taken down. You can, however, buy complete back boxes from Pelco. Spectra IV Spec Sheet lists major components like bubbles and back boxes on the last page.

The largest shortcoming for using HD-SDI repeaters is the need to place them mid-span. For longer runs, it gets even worse. First, your DVR must be capable of sending power down the coax, and that can limit the number of repeaters you can use. Second, you would have to find point(s) along the cable to cut it and insert repeaters every 80m or less (EqcoLogic specs out 83m max distance at 3.0Gbps = 1080p on RG59 (Model SL-EQCO-3G)). Ethernet-over-Coax (EoC) products can be placed at the ends of a cable and can send IP signals up to a mile. For instance, the Vigitron Vi2401 specs out at 1,800 feet at 100Mbps and up to 5,000 feet at 10Mbps. There are also extenders for UTP, something HD-SDI cannot do. The Vigitron Vi2301 can send 100Mbps data and power 2,100 feet over a single 4-pair CATx cable. Cutting cable and inserting devices along it is never a good thing. Each added device creates three additional "single points of failure" (two connectors plus the device itself). While both HD-SDI Repeaters and EoC devices add multiple failure points, at least EoC's points are located at the accessible ends of cable runs, not somewhere in the middle and likely hidden. Add the difficulty finding the 80m point of a single cable possibly buried among many and you have an instant recipe for frustration.

Who says? A number of IP cameras can do even multiple 1080p streams at 30fps. Some can even stream 1080p at 60fps.

I agree with Ilker. As long as you don't choose a lens designed for a smaller imager, which would cause a tunnel effect, you're good to go. For instance, on a 1/2" camera, a 16mm 1/2" lens would provide the same field of view as a 16mm 2/3" lens or a 16mm 1" lens.

Agreed, wireguys. Anyone who thinks they don't need tech support is like the defendant in a criminal case who thinks they don't need an attorney.

Are you certain it's not a dog giving you "presents"? I had a cat once that used to leave birds and mice on my doorstep.

I'm not saying IP should necessarily be your choice, either. You should define your project and its objectives first. Distance between camera and whomever would "buzz" the person in. Existing infrastructure, in case it can be utilized instead of having to run all new cabling. Existing system, if any. For just simple access control, you could buy something like an Aiphone Video Intercom. They have options for cameras and intercom types. You could also just buy a microphone or intercom and camera, monitor and maybe a DVR to provide video and audio. There are IP cameras that have microphones and/or microphone inputs and/or two-way audio capability. The choices are endless.

Hah!!! You do realize that video was produced by WorldEyeCam, a "distributor" of mostly cheap, no-name CCTV products? These companies are a dime a dozen on the internet and any claims made by them are typically suspect. That video is almost hilarious in its amateurism. It's pretty obvious that the IP system was deliberately throttled back to around 4-5fps and that latency was deliberately added or emphasized. That video reminds me of the "infamous" train video produced by the HDcctv Alliance a few years ago. It compared the video of a toy train going around a short track from IP cameras to the same train via HDcctv cameras. It was obvious that the footage from the IP cameras was deliberately made to look bad and the Alliance's chairman was called out on that. The video has since disappeared.

Latency would be a non-issue in your case unless the system was so poor that its latency was seconds, not milliseconds. Pick a technology that fits your needs and your site. For identifying who is at a door, pretty much any properly designed system type will work, even analog. IP has some distinct advantages over HD-SDI/HDcctv in that there are far more manufacturers and equipment choices and the video can be transported farther over various cable types with appropriate converters and extenders. HD-SDI/HDcctv does have the advantages of lower latency and better "live" image quality but those "advantages" are often over-emphasized by proponents.

Hire an owl.

Add zoom in/out and focus near/far - 4 more functions if the DVR supports them. LEFT: FF,01,00,04,3F,00,44 RIGHT: FF,01,00,02,3F,00,42 UP: FF,01,00,08,00,3F,48 DOWN: FF,01,00,10,00,3F,50 LEFT-UP: FF,01,00,0C,3F,3F,8B RIGHT-UP: FF,01,00,0A,3F,3F,89 RIGHT-DOWN: FF,01,00,12,3F,3F,91 LEFT-DOWN: FF,01,00,14,3F,3F,93 AUTOSCAN: FF,01,10,00,00,00,11 STOP: FF,01,00,00,00,00,01 ZOOM+: FF,01,00,20,00,00,21 ZOOM-: FF,01,00,40,00,00,41 FOCUS+: FF,01,00,08,00,00,09 FOCUS-: FF,01,01,00,00,00,02 If course, calling presets is the easiest way to add commands. CALL PRESET: FF,01,00,07,00,, All commands start with the word 'FF'. The second word is the receiver's address. For the above examples, I used Address #1. With preset calls, the 6th word is the preset number. The checksum is the sum of words 2-6.