kensplace

Dont know about licensing, but i would hazzard a guess that if the job is not done 100 percent to all rules and regs (such as planning permission, building regs etc) then its likely a scumbag would claim to the court that the evidence is inadmissable as the cameras were unlawfully installed/operated/sited whatever. Then the scumbags client would walk away scot free. So even if licencing is not required, it would not hurt to brush up on the rules and regs to cover your clients backs.

What is "DSP Real Time Analogue output" I can guess at Digital Signal Processing, but what does it do?

Dont forget to test the bandwidth and resolution of the recording device also, you can have a great camera, but a crappy dvr.....

Are they capable of recording 25 fps without missing ANY frames? A good test would be to record the output of a mux to see whether they are dropping frames or not. The naked eye cant tell, but if you record something like a mux then play back frame by frame you quickly spot problems. I need something that can record the ouput from the mux, but the TV card I have misses frames now and again, which means I cant use software to select cameras when playing back. Would love to try geo, but my main pc is mainly amd, so problably not a good move (although I do have a rackmount Intel job, but its not fast, p3 750 mhz)

Never had any decent results with PC cards (although not tried geovision). The pico ones are rubbish, and even the hauupage ones I tried are just reasonable at best (and only then if you dump there rubbish software and use dscaler instead). Deinterlacing problems are also a pain with the pc cards, but dont seem to happen at all with the standalone unit. The standalone gives so much better recording and playback quality compared to a pc unit, it has to be seen to be believed (course thats probably more to do with the fact I use a tft monitor on the pc, and the standalone is connected to a TV.....) As far as the multiplexer goes, I can record ok of it, and view it back frame by frame, but playback into the multiplexer still does not let me select a camera to view, but then again, it does the same if I record to a standard VCR, so maybe its just a crappy (or faulty) tecton mux. There are several modes for recording, ranging from High quality, which looks the same as the source to Super long play (more like old vhs) I stick to standard, which gives about 67 hours recording. You could have up 300 or so hours, but I prefer the quality. The best bit is the built in dvd recorder, so I can burn footage (and tv) to dvd, or even normal cds (vcd or svcd). Probably will end up getting a proper multi cam dvr one day, as I use this one for recording TV far to often

Sorry, no idea, but its a bit worry that a DVR would do something like that after a power loss, would have thought they would be designed to go back to how they were set up before when power comes back on, seems a weird design....

Another problem with wireless wifi type cctv is jammers, someone can jam the signal with a jammer, and the whole network of cameras could be killed. Unlikely to happen unless large sums of money are involved, but thats where you need the most protection anyways......

Hypothetically, say someone wanted to alter a video, that was watermarked, so a person was missing from the image, or whatever. All they have to do surely, is play back the watermarked tape to a pc, that captures the image. Get a pro to edit the video so its altered to suit their needs, then ... Set the time / date on the dvr to when the crime occured, and plug the pc's TV out into the camera input on the dvr, then hit record.... It would then be re-recorded, with a genuine watermark, but with altered footage. Or am I missing something? Watermarking seems pointless to me, other than as tool to stop casual editiing.

What suffices as a watermark? Would a time date stamp on the image itself be enough? Or does it have to be embedded invisibly into the image? Even then though, whats the point? Would not be impossible to digitally edit (even if it had to be frame by frame) the image, to alter a watermark, even if you had to reverse engineer by decompiling the dvr firmware to figure out the watermarking algorithm. If someone wanted to fit someone up and frame them for a crime - it could be done....... Same goes for time/date stamps, its incredibly easy to add a time date stamp to a image that does not have one....

I always thought there were so many cameras because the councils must be making a fortune on grants etc, kickbacks etc. The prices they pay for a few cameras, ptz, etc is beyond belief. Funny thing about england is, we have so many cameras, but so few of them actually produce anything worth using.....

Dont understand what you mean, what are you after exactly?

Cover your back, get the elevator company to install the cabling, or to recommend what cabling to use, then you just need to plug the stuff in....

Come to think of it, wireless (especially unencyrpted) is probably illegal in the UK for firms where their footage falls under the data protection act, as you have to ensure only authorised personel can see the footage. With wireless, pretty much anyone that wants to can view it if they put their minds to it, after all it is being broadcast..... Dont think I have heard of a case involving the DPA and wireless, but its probably bound to happen one day.

Regular maintainence, cleaning - its suprising how fast the cooling system on a pc gets clogged up with dust, reducing its effectiveness dramatically. Sort out a policy on security updates, are you going to let it auto-update, and hope microsoft gets it right? Are you going to install them manually when proven safe? What about anti-virus, if the machine has a drive, or usb ports, and people have access to it then they may run software on it, plus if its connected to the internet then av and firewall need to be spot on. Stop any un-needed services on XP, such as fast disk indexing, etc, quick google search will reveal lots of services that you can probably stop (obviously each user has different needs), and any you stop will help to free ram and processor time. Get rid of any desktop wallpaper, use classic start menu, install a free-memory utility to clean up the memory when it leaks, schedule scandisk to run at startup. Fire up task manager, research what every running process is doing, and if any dont need to be running, then stop them from starting in the first place. Make sure all devices drivers are the latest (stable and proven) version. Turn off anything that you dont use in the BIOS, such as spare ports, and on board hardware that is not used. Check ALL bios settings, especially for memory and processor to make sure they are set to best (stable) performance, and check bios is latest version. Make sure the end user is backing stuff up, and the master software is kept safe, so when it does fail, which it will, they have backups, and a easy way to get back up and running. Delete temporary files on a regular basis (they can build up very fast), if need be write a batch file to delete them at startup.

If the capture has captured them so the frames are never skipped (dropped frames are a pain...) then you can use avisynth, check out a thread on where I asked a similar question, the solution works well (if the frames are in sequence) and you can even view them quad view, or even all 16 on screen if you feel like it.... http://forum.doom9.org/showthread.php?t=112758 If your capture and plexer saves the frames 1,2,3,4..16,1,2,3,4..16,1,2..16 etc then it should work fine, mine plays up as it seems to skip frames sometimes. Worth a shot anyways, saves buying very expensive software...

Dont suppose anyone has a circuit diagram/schematic for anything that decodes up the coax telemetry? I now have a old bbv tx1000 unit, that does up the coax, and for once I know loads of info on it (lovely site they have, lots of info) but cant get my head around the analog part of grabbing the signal fro the coax. If I can get the signal then I can do the rest with a pic chip, and save buying expensive telemetry receivers, which are a waste of money for home use really.... The telemetry is encoded in the vertical blanking, as a FSK signal (222khz logic 0, 250khz logic 1 at 1200 baud) but analog stuff does my head in, digital programming stuff is more my cup of tea (but even thats harder these days, cant think straight half the time) Anyone done or know anyone that has done diy telemetry stuff? Its not really important, just a curiousity, a itch waiting to be scratched sort of thing, a nice project to take on.

He has a wireless camera from the original post, so just power needed.

As hinges are metal, could you not just use the hinge itself to pass the signal? One hinge for ground, another for +V? You would still need to get the power cables down the door frame to the hinges, and then from the hinge on the door to the camera though, which could be tricky, unless your handy at diy (assuming low voltage 12v dc camera, obviously dont do this with a higher voltage camera in case someone touches the hinges)

As you are in the UK, you may need planning permission depending on how you plan on mounting the six cameras, if two or more cameras are located on a single building, any less than 10 metres apart or more than four cameras are fixed on any single side of a building then you will need planning permission, or the council could get nasty. Its also possible that extremes of temperature in the loft (it will probably get very hot up there...) may shorten the life of any electrical equipment you install up in it, specially delicate stuff like a dvr.

Whats the colour depth set to on the video card? Could be set to low, which could give a look like that.

100 foot of cable in a loop will also act as a antenna and may be liable to radio frequency interference. The loop acts as a inductor and this may show up as a quality issue, or it may not, all depends. Probably just the cams though......

Back in the 90's a lot of its camera's were dummy cameras (all the car park ones, which caused a stir when the multi-million pound robbery occured, and it turned out no footage was on cctv as they were all dummy cams) Not sure what exactly they use now, but http://www.airport-technology.com/contractors/security/novar/ may be a good starting point.

funny thing is, I find the cabling aspect the hardest (home cctv, but done pro computer systems for many firms in the past) part to do. The actual connections are easy, crimp or solder, but i am no good with diy parts like hiding cables etc. Personally I am used to rs232 etc, so find them easy, but I do think the cctv world in general is for want of a better word, immature, as they have in general not latched on to the fact that if they open their systems up, they will sell more. At present they are relying mostly on the fact that most cctv people dont understand their protocols (and those that do probably cant get data for them, apart from pelco which seems well documented) so they can get away with charging a fortune for stuff that should be cheap. There is also a lack of 'repairability' in modern systems, and this does not apply to just cctv, it applies to pretty much all modern kit - try repairing a board populated with surface mount devices, its impossible unless you have very expensive kit, and expensive people...... Which makes the cost of repair more than the cost of replacement in most cases these days, so everything is just binned when it plays up these days.... Why cant manufacturers show pinouts for their equipment, and make the info available to people - many (but not all) manu's either charge a fortune, or refuse entirely to just give simple pinout data for equipment to end users - I cant see the point in that, after all they cant even claim its for security purposes as a 'dodgy' person could easily disable or tamper with a camera in much easier ways than via its connections. Telemetry kit seems VERY expensive, when its pretty simple to design, its just a interface that transmits serial data over wires, then decodes it at the other end to control a set of relays - nothing hard in that, hardest part is probably the transmitting over the wires bit - the rest is a doddle. I shudder at the thought of how much people pay for 'pro' kit, when in just a few years time, even the manufacturer of the item is likely to turn around and say thats obsolete and we dont support it any more. The thing is though, cctv kit is not like a pc, its a camera, recorder, display, and stuff to light it and move it. Non of that stuff should become obsolete in just a few years (apart from major improvements like digital versus timelapse recorders, or ccd rather than tube cameras etc) if the kit is high end, and still does the job, then they should still support it. But the manu's in general are not in it for the end user, or even security, they are in it for money - thats why things are the way they are. Security systems could easily be more integrated, but the cctv world rarely talks to the alarm world (who also seem to be big on privacy, like certain alarm forums refusing to give out engineers manuals to non trade, when even a non trade person like me has every scantronic manual in existance (for older models)) - its just a we know it, we charge you for it scenario. That mindset does not bring innovation into play - it restricts it. Alarms should link to cctv systems, cctv systems should link to home tv systems, easy interfacing is the way forward - not propriatery formats which hinder that.

Yes Rory, It was about 50 US dollars I guess, about 20 or 25 Uk pounds (including postage) for a 16 input multiplexer / multiview (4,8,16), zoom, freeze etc tecton kramplex, which supports loads of timelapse vcrs (not actually got a timelapse vcr, just record either single camera or quad, as needed, or record the plexed output on the pc). Very solid well built unit, tad on the large side (looks like 2 or 3 U rack mount case, but not rackmount more VCR style. They probably cost a fortune years ago, but go for next to nothing on ebay these days..... Plus its built out of hundreds of old ttl chips, so when its of no use to me I can strip it down and get a gold mine of spare parts

Seriously, IGNORE SPECS... This may sound insane, but specs are just that, specs written by people in a perfect world, with perfect components, with X amount of perfect whatever interference thrown in. In the real world, experience will tell you what works and what does not, I remember when people said a parallel printer cable could not be more than X meters long (cant recall the exact length, but say 10 or 15 meters) and I went to a job where a 50m+ lead was installed (whatever the exact lenghts were, lets just say it was at least 3+ times the max ever length it SHOULD work at) and it worked perfect, and had worked perfect and continued to work perfectly. Cable is cheap, experiment, even specs can fail, as the camera you are trying to hook up may not be up to spec, and may need a shorter cable, or amp, only way is to test. If you go by specs your just guessing that all will be perfect.