hardwired

The odds are that Samsung could put together something like the other camera, but most users of their products are used to a "Box" style camera. Most products of this type are installed in a way (enclosures, drop-down mounts) that size does not play a large factor, anyway. Heat dissipation may also play a factor, the larger style may end up lasting longer.

2 Megapixel is the highest I've seen, so far.... http://us.sanyo.com/High-Definition/Full-HD-1080p-Day-Night-Pan-Tilt-Zoom-Camera

That could work. The Exacq software, by default, sets the camera NTP to point at the server itself, but there is a setting in there to point the cameras to an alternate NTP source. I've used Mikrotik Routerboards (really powerful... and user UNfriendly), with a serial GPS connected to it, for a NTP server to time sync DVR's at a site that didn't have Internet access.

BTW, I agree it's kind of a hairball for people who want to use camera-provided services at the same time as Exacq software... But I really think in most instances, the Exacq software is all you need. For those of you that disagree, I am working on something that may help. I am in the middle of building an embedded, NON x86 platform to run the Exacq on. Atheros processor (probably), OS on flash, and writing to a SSD. Should be under ten watts or so. I'll update as it progresses.

Problem with those ideas, the recording is a service. If you stop it to do the script, then restart it, on the restart of the service, the settings will be reset. Plus, the script would have to be written to write the correct timezone settings to each CAMERA on each restart of the service... Not the server. You would have to know the proper way to write those setting variables to each camera, and that script would have to be changed anytime you changed a camera (or it's IP).

Three-piece crimp on only, for us. Too many problems have shown up later with the other types to be comfortable with. Installs 10+ years later are just fine with three piece connectors (the connectors, at least, the other hardware is a little iffy at that age!), others (two piece, twist on) have failed much more often. On a service call recently, it took me a while to find my RG59 crimpers at the bottom of my toolbox, it's been RJ45's on CAT5e/6 for us almost exclusively for a while now...

A lot of manufacturers of automotive LCD monitors make them with a flush mount bezel for mounting into a headrest, you could look into those. Look at mp3car.com for some ideas.

Yep, you got it, they are replacement windows for Pelco HS8080 enclosures, regular Plexiglass/Lexan was too wavy, made distortion in the images, the Pelco windows are pretty good. As far as the wattage, I haven't measured it, but on the Core2 boxes, I'd say about 70-80 watts, on the Atom, maybe 50 or so.

Probably around $3K or so.. Surprisingly, that fiberglass box is really expensive ($450+), but it's a lot lighter than steel. Working on a build that looks a little more like a transformer now... And seeing just how light of a platform I can port the Linux version of Exacq onto (NAS on steroids, kind of... ) Stay tuned!

I have some on poles, but they are for a governmental entity, and they have to get a warrant before they install them. As far as power, there are power taps that connect under the photocell on the streetlight and give you a regular outlet.

Not PTZ, but I've been building these for a while now.. 900 Mhz long range Mikrotik/Ubiquiti wireless/router combo, Intel Mini-ITX boards, changed from the Arecont's to Panasonic 502's. This was sort of a prototype, but it has been running through a very hot (105F) summer with no problems. This one's running Milestone, but the newer ones have Exacq.

Not sure about your cam, but most cameras can be supported in MJPEG mode better than H.264 or MPEG4. You might want to try to find out the specific URL string to retrieve a MJPEG image from the cam, and try that in your recording software.

Exacq has a browser interface with NO plugins, no ActiveX, and now playback and live, with more features, as of version 4.0, just released. If you create an account on the Exacq with pass-through login (no authentication), you can request a URL stream with whatever parameters (resolution, etc) you want through it.

It's a pretty nice camera. I have some sample images here viewtopic.php?f=19&t=11322&start=150 Are you talking about the box or dome style 502? if using the box, make sure you use a megapixel rated lens. Panasonic has a great design that pretty much eliminates the chance of moisture getting piped into the dome, a common problem with other cams. The analog spot monitor output makes setup easier, and the auto-backfocus is really nice. It is only MJPEG in the 3mp mode, though, a lot more storage/ bandwidth if you go that way. The pictures are pretty good in the H.264 1.3mp mode, though.

Did you use CCTV RG6 (solid copper core, 95% or better copper braid) or CATV/ Satellite steel core, foil shield cable? CCTV type RG6 should work better than RG59, but if it was foil type, that was likely the problem.

The timezone on the server/client setup stays intact, what Chris is talking about is that Exacq sets the individual camera timezones to GMT 0, and the Exacq server as the NTP source, when it initializes. Not a problem if you are only using the Exacq, but when you try something else with the cam, you will have incorrect timestamping (if you have a cam that has OSD time and date stamping options, you can turn it on and see that)

I agree, that's a headache. The Exacq seems to rewrite the configuration to each camera on a restart of the service, not sure if there's a good way to hack a patch for that, it would involve a script to reset the NTP server location on each camera, you'd have to know a lot about the cams involved to do that. Incidentally, you seem to have some experience with with low-power boxes, seems like you could run the Exacq server on something almost as light as your NAS, do the full res storage there and eliminate that problem... Was there some kind of bandwidth or other problem that prevented that?

Sorry, I thought that's what you were recommending.. Anyway, he definitely has the wrong lens for that application, and that's why he's seeing this amount of edge-field distortion and lack of resolution. For really detailed LPR type shots, I've found the best results with fixed-focus Fujinon machine vision lenses.... About 4+ times the price of these lenses, though. He should be a lot better off, though, with any of the true megapixel lenses we've mentioned..

That one's not megapixel, either, and it's 1/3". In the Computar, the H3Z4518CS-MPIR 4.5-13.2mm is the one to use. Also, Fujinon makes the DV3.4X3.8SA-1 3.8-13mm lens that's F1.4, gives a little better light gathering than the Computar at F1.8, also the Tamron M12VM412 4-12mm is F1.4. In the 1/3" megapixel format, Tamron has them down to F1.0 now, handy considering the native handicap CMOS megapixel imagers have compared to CCD.

I agree, why not try it. I think the note about using 6 or more pairs has to do with the distance from the shield to an individual pair, which affects capacitance, and the high-frequency attenuation. So, CAT5, at four pairs, is at least partway there in that regard.

At least two manufacturers say no... here's from Nitek's FAQ page. Q: Can shielded twisted pair be used? A: NITEK does not recommend the use of individually shielded twisted pair cable. Individually shielded pairs have a capacitance much greater than the recommended nominal value of 17pF/ft, which severely degrades the video signal, drastically reducing operating distance.In some cases use of an amplified receiver can compensate for some of the signal loss, however use of individually shielded pairs should be avoided. Multi-pair wire (six pairs or more) with an overall shield can be used without signal degradation. And from NVT: Can shielded twisted pair be used? Some customers have successfully used shielded wire up to a few hundred feet. However NVT does not recommend it. Its high-frequency roll-off will severely degrade the distance performance. Use of an amplified receiver can compensate for some signal loss. Multi-pair wire (six pairs or more) can have an overall shield without degradation.

I'm fairly sure you can do something like that in VLC, with playlists and optional syntax arguments/ scripting. I say fairly, because I looked through the documentation, and found so many options, I'm pretty sure it's there... Some heavy reading, though.

I do wonder about the opportunity for cheating there. One player dropping out early as a shill for another player, with some type of wireless communication equipment, could really screw things up pretty easily. If you could build in a little delay (30 seconds or so), it would go a long way in preventing that. No really easy way to do that comes to mind right now, though.

Although I'm not really familiar with that piece of hardware/software, a couple of thoughts come to mind: How do you know that you don't have that latency in the JVC cameras, if they are fixed and not PTZ? Latency tends to be most noticeable in PTZ's, the only way you would notice latency in fixed cams is if you had a analog cam, or direct view, of the same area. I'd check and see if there are separate viewing and recording rates selectable in the unit, a number of the NVR softwares I use support this, not sure about the Pelco, though. What type of compression are you using, MPEG4 or MJPEG? That can have an effect, as well. you might compare the stream output in a browser, at the same time as your client software, to see if it stems from the camera itself, or the software. Also, I'd give Pelco's tech support a call about it, see if they have seen anything similar.

Haven't seen anything like that, but when focusing megapixel cams, I often use system menus that show the file size of frames, it usually correlates pretty well with focus, as file size increases with detail in the image. That comparison falls apart when you start closing the iris, though, as the noise in the image has the same effect. I would think you could build a plugin for VLC or something like that fairly easily. Let me state very clearly I'm not saying anthing about lines of resolution here....