rkolb86

I don't intend on hijacking, but I have a question since you've already hooked up the PIR to your 8332. I plan on hooking up a PIR as well and I'm curious if you had a diagram or something. I want to use the camera's power to power the PIR. i noticed that 2 ports on the connecting block are 24VAC. I have the camera powered by PoE. Will these provide 24VAC? Most of the PIRs I find operate on <18V DC. My guess is that DI and GND is a normally open loop and it triggers the sensor input when closed? Any information would be useful. Anything you can offer.

It should work, right? I'll be using 1 Vivotek IP7631 (2MP) and 3-4 IP8332 (1MP) on a 10/100 switch? I'll be using the at max resolution if it helps. My guess is that it should work fine, but I feel I should ask the experts here. I doubt it matters, but I plan on using 2 NICs inside my PC, one for internet access, the other for the IP cams. Switching Capacity (Gbps): 3.2 Gbps Switching Forwarding Rate (Mpps): 2.4 Thanks

This is true, but I'm the kind of person who likes to be prepared for the worst case scenario. If it glitches while the next door neighbor is throwing broken glass in the driveway again, I'm screwed

If anybody cares, this is the switch I'll be using. http://www.newegg.com/Product/Product.aspx?Item=N82E16833181164 It's specs are very good for what you pay for...a little over $100 for 10/100 and 8 PoE ports @ 16.25w per port. It was actually cheaper than buying a standard switch and 8 PoE injectors. I've used ZyXEL before for a VPN firewall as well as some ethernet-ADSL bridges. They are very reliable and their customer service was pretty good.

This sounds like the issue is a function of a slow or bad switch (either CPU, Bus, cache speed,or processing algorithm), not the buffer size. Sure a larger buffer would help, but only until the buffer becomes full again and it starts dropping packets. If you were to theoretically enlarge the buffer of that D-Link, I bet the same issue would occur. (It was probably a low-end D-link...which is almost all of them) Well that's easy: 3.2Gbps = 3200Mbps/16 ports = 200Mbps bidirectional = 100 Mbps in a single direction per port (I originally did this calculation using the pps, a 64 byte frame size with a 20 byte interframe spacing, and a 8 byte preamble originally...but I won't subject you to a page of math) I'll only be using up to 8 ports though because only 8 are PoE....with a 130W budget across the 8 ports.

lol...you should tell the hardware manufacturers the opposite! Back in the day, 100Mb was huge...who gives a **** about them these days? The switches should be rated in MBps! 12.5Mbps and 125MBps for 100mbps and gigabit respectively. That's good to know that each 1MP cam will prob be on the order of 1 Mbps. If that is the case, I could probably hook up about 85 or so 1MP cameras to a 100Mb/s switch (assuming a **** ton of overhead)

Not me. Your 5MP cam would use nearly 1.5 MBps then. Say your 1MP cam was 1Mbps @ 5fps, I'd guess that it would probably use about 6 Mbps [megaBITS] at 30fps (using a constant bitrate)...or 0.75MBps [MegaBYTES] I think he stated that the 1MP Vivotek with h.264 would be a little less than 1 Mbps at 1200*800, 30fps (or whatever max resolution is) Mb to MB...divide by 8. 8 bits in a byte....unless you're a HDD manufacturer who uses a decimal multiple system in order to legally mislead consumers Am I missing something here? Did I do something wrong?

So on a 100Mbps switch, i can achieve 12.5MBps on the link to the NVR. 5 cameras using MPEG4 will give me <5MBps...enough room for up to 7 (theoretically) more cams before I saturate that link (probably 5 cams additional with overhead)?

Just to be clear, your numbers are in Megabytes per second, not Megabits per second?

IDK if I'm expecting too much out of my analog system or maybe my camera is a POS. The camera is a varifocal (4mm-9mm)...zoomed all the way to back to 4mm for a bigger picture. At the closest place in the video, the person is about 16 feet away. I tried focusing, this is the best I could do. I notice a lot of picture blur as well on motion. The live video looks the same. Recording and live is 640x480, 5 or 6 FPS. Camera: http://www.supercircuits.com/Security-Cameras/Infrared-Security-Cameras/PC509 or A42YI9MgaEs The problems is that under any type/quality of lighting, I can hardly even make out her face. I have another manufacturers camera on the system (3.6mm) which looks better, but the distance isn't that far.

Yeah. I decided to scrap that project and ordered some cheap 2MP/1MP cams (Vivotek IP7631, IP8332). On that note, I'm pretty sure you use the horizontal pixels for that calculation, which would be 640/60 which is approximately equal to 10ppf

Did you determine that electrical interference was the issue? If the electrical interference is the issue, I doubt PVC will do anything to fix that. The metal (steel?) electrical conduit might do it, but that is a shot in the dark as well. Also, using shielded twisted pair (instead of UTP) may also help resolve the issue.

So you're saying that IP cameras encode/compress at the camera instead of using some sort of streaming protocol and let the NVR do the encoding/compression?

Isn't it only limited because you are getting what you pay for? Maybe all you need is a few cameras, so they build a piece of hardware that can do what you need for a good price? I'm sure every NVR or NVR Software has it's limits as well, or even the hardware it is built on. I highly doubt you can go and hook up 50k 16MP cameras to some "real" hybrid nvr or software. It just isn't physically possibly unless your using some sort of clustered infrastructure (computer scientist by profession). If this is true, it you prove your claim of a "true" or "not true" NVR invalid. I'm not a CCTV expert (i'm actually on the complete opposite end of the spectrum), but this makes a ton of sense to me. If you can show me one NVR or NVR software that doesn't have a limit, either written in black and white, or and actual physical limit to the amount or cameras/resolution you can use, I will stick my foot in my mouth. I think all they are saying is that 8-ch 4CIF/4-ch 720P Real Time is just the physical limit of that machine, and maybe they enforce the limit (or the limit is just so you know its abilities) in order to keep complaints saying that "my $1500 NVR can't record my 800 5MP cameras at 30FPS" ...or maybe I don't have a clear understanding of the definition: "true" NVR

I am curious as to what a true hybrid is. What is the difference between a "true" hybrid and something that is not a true hybrid.

I need a cheap monitor. Doesn't have to be wrist mount, but def portable. B/w is ok.

I'm thinking about getting a Hikvision DS-4208HFVI, and I want the ability to add some cheap vivotek (or other) megapixel cams in the future. I'd like to know how comparable their hybrid solution is with other manufacturers IP cameras as well.

PC running Blue Iris, software compression

So i've come to a conclusion that i'm not gonna the quality I want with with analog given the the amount of space I want to view. This is solved. I need to fix the blocking issue when there are lights and such. Check out the video with the jeep. There are a bunch of setting inside the camera...I bet its one of those. Any idea what I should mess with? Is it a sensitivity or shutter issue or something? I'm a computer guy...camera's are relatively new to me, although i catch on quickly. Thanks guys

Seriously...how the heck do/did people get decent pictures of a 16 ft driveway without needing 2 (or more) cameras? I fiddled with a new camera...I look way better at night, but day it only moderately better. If I turn of the auto gain control (picture goes really dark until a plate drives past), I can almost get a plate at night, but i'm not zoomed in all the way. I am not too concerned about this, but it would be nice. My biggest concern is that when I get a car (at night) or even something reflective/motion I get these block on the screen. The camera's OSD has a ton of settings, but my video monitor won't arrive until a few days from not, and I'm not gonna run back and forth. Any idea what setting I should screw with? And advice is welcome.

I could just be settings (video format or something else), could be the baluns, could be the cat5 to the camera. I'm no camera expert, but I don know structured wiring. You shouldn't run that cat5 parallel to a/c within 12 inches for any more than a few feet. This could be the problem.

Surprisingly, the added compression from youtube and such doesn't change it much...if at all. When side-by-side, it looks identical to the AVI. I hooked up a new camera this evening facing the driveway in the other direction (http://agicctv.com/AGI_product.aspx?item=VC-CA-TIR7-742), plugged it in, and actually said "Wow!". I hooked it up in the dark, but the picture quality was already 150x better. I could actually make out faces, the zoom and focus actually worked really well. I only pair $125 for the camera, but I could tell the quality was much better than that supercircuits POS. The new one has adjustable IR intensity and an OSD to adjust all kinds of settings like shutter speed, contract, and a whole bunch of other stuff. The other camera was as clean and focused the best it could have possibly been when I recorded that video...which means I'll be replacing it with an identical AGI camera and I'll post the video here so you can compare. The only thing I can't get from that angle is license plates at night with the new camera (I still haven't checked during the day though), but I think this is going to be a low light/IR issue. Plates just look like bright reflections.

Is it run parallel to a/c for any long distance?

Im gonna focus the **** out of it when it isnt raining and get back to you. Its very touchy, but I looked at the recorded video of me playing with it and I saw a brief moment where it looked decent. Maybe ill have to suck it up and get a wrist monitor.

I know the area is huge. What I'm concerned about is the picture quality of the person's face. She is standing about 16 feet from the camera with the camera zoomed out to 4mm and I still can't see her well. The funny thing is is the $60 3.6mm camera I have appears to give a much better picture (although I haven't tried it @ 16 feet). Both cameras have a 1/3" Sony CCD. Are my expectations too high to get a person's face with 4mm at 16 feet? What would you suggest for a better picture?