Thomas

First, MPEG-4 isn't defined in an RFC. It's an ISO standard and the definitions of it are in the ISO guidelines. There are RFCs for streaming it. About a half dozen depending on how you want to do it and what you're optimizing for. Second there are currently 20 standards with 3 more in working groups. For instance h.264 is MPEG-4 part 10. This doesn't begin to start counting variations of MPEG-4 like WMV or Dvix which share a lot of base technologies. And this doesn't begin to take into account the use of external plug ins like Active-X controls, all of which require some way to extract the video into a usable format. But hey, Axis uses RTP, a poor and ****ty spec that creates nightmares for firewalls, so that must mean the whole IP camera industry is using open standards, right? Full of ****. There is no open standard used by the IP camera industry that allows the end user to make plug and play use of the camera. You even admit that it would take a software developer time to code for your specific product. That doesn't indicate an a consistent standard among IP cameras. Don't something on top of another widely used standard doesn't make what you do a consistently industry standard. If it did, there wouldn't be a requirement to have development time.

There are already well defined standards for IP camera streaming: http/rtsp, H.264/mpeg4/vc-1/mjpeg. But there are companies using variance. Unless many people use their cameras, these are the last we want to integrate. Standardizing the stream type is not the same thing as standardized connections given that almost all of the SDK's have a different API for actually getting to the camera's stream. Simply building on existing standards does not mean that you have standardization. And even then MPEG-4 is a standard that is all over the place and different decoders work differently with different encoders.

Pfft....Trash-80. Too poor for a nice C64?

Hardware compression isn't necessarily going to give you better quality of image. All it's going to do is the compression on the board itself, putting less strain on the system's CPU. And it's a solution that arrived after it's time.

Funny AOL The snide comments aren't helpful. If you don't like someone's choice of computing platform or ISP, keep it to yourself.

I would assume any camera with an Auto-iris lens would make that almost impossible.

You'd configure the prolink for port forwarding for the camera and it should have a set of ports for wiring your physical connection. I haven't played with a prolink so I have no idea what the configuration is for it. Generally for a set up like this you want to use a single router.

No.

Axis is an exception rather then a rule. They make it painfully easy to pull the basic video stream from their software. When you start talking about more painful things like integrating proprietary software from the camera....development time goes up.

Pretty much all modern PoE gear uses a chip to determine if the other end is looking for PoE or not. Some older equipment doesn't have that but it should be fairly rare.

IP is so far from open it's scary. Interoperability between manufacturers is non-existent and all it takes is one firmware change and your camera no longer works with the software. There are no standards for transmission of IP video. And you're going to ask "Well we're all using MPEG-4/MJPEG/Whatever..." and that means nothing. There are so many MPEG-4 variants that it doesn't begin to allow one decoder to handle all of them. There is a right time and place for IP but I'd be so much happier with the companies if they would just let me shoot all of their marketing departments.

Kalpesh, Thanks for your help, I tried but it failed, cause the NUUO software need choose camera brand name and model number. That mean, we should pay for NUUO, then they just can add our IP camera list on it! So disappointed with that! That just can say business just business. My clients in Europe told me that they always use NUUO for Hybird system, so our IP camera lost the market. Nest step, we should talk with NUUO for this issue. Have no ideal how much we should pay for them. I don't know of any company that would charge to add support for a camera. At the same time that doesn't mean they are going to jump for joy at you approaching them with an SDK. Development time is not cheap and throwing more developers at a project does not make it go faster. As such development time is one thing they won't be willing to waste. There are a couple of factors in play: 1. Do their customers want to buy your cameras? If the answer is no, then don't bother. If it won't make them any money they won't be interested. 2. How hard does your SDK make it to integrate? If you have a ****ty activeX control that needs a wrapper written for it then it's going to take longer. 3. Your country of origin plays a factor. Dealing with manufacturers in China always adds a layer of risks from poor quality control to the fear of stolen code in the SDK.

Wait, you're running alternating current on unshielded pairs next to the video pairs? And this works without lots of interference?

Good point, I did not know that. There are tons of lens calculator sites out there. Do you have a favorite? Not really, the math is pretty straight forward.

That should be fairly standard on many DVRs.

Milestone Video Insight Luxriot All of those might meet your needs.

I've never seen a system do that. Most of the analytics systems are after the fact. It might be possible when you start talking at the high end like IBMs systems but it's not something you'll find on most systems.

How many cameras and what kind of price range are you aiming at?

Great. Given that 90% of the shell stuff should be trival to move over generic XP you'll be fine going that way. Window's Media Center comes with all kinds of lovely codecs and directX modifications. While it is entirely possible that you'll find someone it's not an issue with, you'll spend far more time doing the R&D then it's worth.

How much are you willing to spend?

Please don't use Windows Media Center with any DVR/NVR software. You're buying nothing but grief and over paying for it.

There are some companies like Object Video that do some analytics work but most of the useful Analytics stuff is done separately from the DVR with the exception of motion detection and some basic triggers.

Keep something in mind. Photo camera lens size and CCTV lens will give differing fields of view for the same size lens. Consult a lens calculator.

Ahh... good point about the hub vs. switch issue. Thomas, regarding low end PoE equipment, what do I look for in the specs. that would indicate what is low vs. high end? Brand is the fastest way. I generally perfer to work with the quality gear like Cisco's or Juniper's stuff. In the middle range 3com and D-Link have some solid gear. But all of that stuff is fairly pricey. Linksys on the low end has some okay PoE stuff. But a PoE switch for $70 would cause me to run in terror unless I knew the company name. Specs are harder to compare. You need to start looking at OEM's and trying to figure out who makes what.

If you can extract the video files themselves, check the manufacturers website. They should have an application to handle viewing/decrypting the files.