jasonkkn

Milestone software has several catergories - XProtect Central, Transact, Business.... Does anyone know what database they are using? Under what circumstances (eg. where there are 300 cameras recording at 4CIF 30fps for 30days??) does the user need to purchase a MS SQL Server 2005? Thanks...

hi there, just joined this forum not too long, and am still browsing around. Just a little background - was employed by Axis, Bosch, DVTel, and represented them in the Asia Pacific region. I'm not sure what to expect from the forum, but guess it's good to be in like-minded community. cheers. jason

hi "thewireguys" i don't have preference. That was a generic question which allows me to better understand the background software architecture that Milestone uses. as the size of an installation grows, it becomes more apparent that a video management software with a sound design architecture will withstand scaling. I remember the first few versions of Milestone software (dated some time back to 2003) has a limit of 600,000 images per camera. They were using proprietary database then. I'm not sure how much is done to correct that, as the database is usually considered the core of the coding. (i do not mean to proclaim that I am a programmer; but these are based on fairly generic assumptions). So when I hear dealers propose Milestone for large projects (eg 300 camera systems with alarms and all), i was just wondering if they had incorporated the MS SQL Server 2005 database into ALL their software offerings? I had this suspicion that only the Milestone Corporate works on MS SQL Express / Server database. (Suffice to say that they have positioned themselves as a low- to mid-end solution)

thanks for taking time out to explain Twilo. i've spoken to a local dealer who handles Milestone, and he does not seem to be able to explain clearly as well. I was looking through www.totalcontrolapp.com - something interesting. Just to clarify - the application communicates directly with the IP camera ya? And not pull video via another application (eg. Milestone, Axis Cam Station...)

thanks for the reply Twilo. I was trying to figure this out - something the manual does not specify. SQL Express 2005 is the free version from Microsoft, but this version will only database file size to grow up to 4GB. (With previous MSDE database, it was 2GB) I'd like to assume Milestone keeps the system configuration data, maybe even metadata and all relevant system information into this database. When the system is large, more information is stored in the database. What i was trying to find out is this - when i size a system, the free MS SQL Express 2005 is good for system up to x size? x cameras? i hope i am making sense here.

it's not easy to pin-point a wireless manufacturer that is 'good' in CCTV application. most wireless transmitters were not manufactured to handle video-type of wireless transmissions, until of late (maybe). Even network switches - they were not capable of doing video transmissions for long, until quite recently. Suffice to say there are several considerations to choose your wireless device: 1. What is datarate the IP Camera / Encoder transmits? 1Mbps? 4Mbps? 2. What mode of transmission - unicast TCP? Unicast UDP? Multicast UDP? 3. what type of transport protocol used by the IP Cam / encoder - RTP/ RTSP? 4. does the wireless device support IGMP - for multicast routing? the above questions will help select a wireless product that is more suitable for video transmission. The question on distance, line of sight, signal strength and all also has to be addressed. hope this gives somewhat of an idea.

yup, Pelco got their initial encoders from VCS (a german manufactuer, which was bought out by Bosch in 2004/5) - and they named it PelcoNet. Then Pelco introduced the Endura system - and they made it so difficult for dealers to sell. A Dealer has to pass a NETWORK exam, before they can sit for a Pelco Endura training. And the software runs on Linux - this is a soft-spot for a lot of debate. Suffice to say, how many engineers can be readily find that can support a surveillance application running on LINUX? Yes, DVTel encoders (some) are from Verint. (Verint bought SmartSight years back). Pelco does not make their cameras as well - they just assembled them, and did a great job making them good. The camera modules in the CCTV market is dominated by Sony & Hitachi anyway. So who's to say they make their own camera module....

i understand DVTel's Latitude v5 and above supports Arecont's and even Axis' H24. just take note that at the viewing end (the client workstation), you might require just a little more computational power compared to viewing MPEG4/MJPEG.

i understand Pelco is re-doing their NVR software. Their first version of Endura was based entirely on Linux platforn, which I understand they did not sell alot. (if at all in most countries). The barrier to entry in their first gen Endura was too great for most dealers. Anyway, guess they should be launching their next generation of NVR software before too long. you might want to be sure you got the correct/relevant NVR information.

CMOS vs CCD : usually CMOS will require more light in order for the image quality to be sharp and decent. CCD will generally perform better in lower light conditions when compared to CMOS. However, there are technology (such as PIXIM) that compensates for the lack of sensitivity in the CMOS sensor. As a rule of thumb, the bigger the CCD, the better it is (and the more expensive it is too). Those found in IP cameras are usually 1/4" or 1/3" CCD (interlaced / progressive) In order of image quality, i'd go PROGRESSIVE SCAN CCD > Interlaced CCD > CMOS. because Interlaced CCD will produce a "motion blurr" image when objects are moving, most cam mfg have incorporated a 'deinterlacing' in their decoding software which generally removes this blurring.

and i read of the 3 mfg who initiated this (Sony, Bosch, Axis), Axis is not even ready to implement the standard as yet.

Yes, the iP camera has an RJ45 connection. This can either be connected to your network switch or directly to your router which has additional network ports. if you want to configure the camera to be able to transmit video over the INTERNET (which includes recording outside of your LAN), then these are the things you will need to do: 1. assign a local IP address to your IP camera (probably in the 192.168.x.x or whatever the IP addressing schema is being used by your router). 2. then configure your router's NAT to port forward the public ports to the private IP address of your IP Camera. There's a post somewhere ( pardon me, i Can't recall) which details how to configure such a connection. If you have a PoE enabled switch (eg. NetGear's 8port), and the camera is PoE ready, then the power will be provided by the switch to the camera. If your switch is not PoE enabled, then you can opt to purchase a power injector.

yup, you surely do have the options of configuring the motion detection parameters, as well as configuring how and where to send the motion-detected image to. Most brands of network cameras have these features built into it.

hi Jhonovich, Maybe it might be good to also highlight the networking aspect. (ie. multicast, bandwidth consumption, subnetting..) It's an arguable issue - some say it's the IT's side of the responsibility. However, most video surveillance systems are running on their own IT infrastructure. A good understanding of the network field will make the 'video system designer' more complete i think.

Firstly, the video signal from your CCD-based camera is 1Vp-p composite video signal, which is compatible with those of your TV. Now, your camera is connected to the DVR directly. What you need to do is this: 1. remove the existing connection to the DVR 2. the camera is connected to the Coax-splitter (see image). 3. connect the coax-splitter to the DVR 4. have another coax cable from the coax-splitter connect to the BNC-RCA converter (see image) 5. and connect the RCA end into the TV. this should solve your problem

I guess a good place to start is asking yourself questions like these: 1. how many IP cameras will I install in total? (this will give you an idea of the amount of bandwidth that will be consumed by the cameras) 2. what is the expected viewing frame rate? (this will let you know the bandwidth consumption per camera) 3. Is there gonna be recording? (if there is, it only means that there is constant traffic from camera to recording server) 4. How many clients will be viewing the same cameras simultaneously? (if camera is transmitting in unicast, number of clients multiply by number of cameras will be the total bandwidth) 5. Where are the clients located logically (with respect to the LAN)? (this will help you determine if video traffic will be routed across subnets) 6. the video stream from the camera - are they sent in multicast? 7. if its in multicast, can my existing network equipment handle multicast traffic? (most existing network equipement are not multicast ready, unless your corporate LAN is being used for lots of multimedia / video applications) Typically, if you have small number of cameras (say below 15 units) and does not require any form of recording, then I'd assume that it'll be safe to have it in your existing LAN. Other than these, it'd be better to opt for a seaparate infrastructure for your IP cameras. hope this helps somewhat.

with data over power lines, as described in earlier posts, the data is limited to the same circuitry in the same power distribution unit. For consumer / entry-level usage, the amount of bandwidth stands at 200Mbps as of today - but the issue of "routing" data across different power DB is still present. I've witnessed a data-over-powerline trial, where data and internet access are routable across different power sub-stations across the country. This was achieved in 2003, with internet speeds at excess of 45Mbps (at that time, it was blazingly fast). The trial used Mitsubishi's products, but the cost of making data "routable" across different power DB is very high, thus making it a major barrier for such a technology to be deployed on a large scale.

might i add - usually MJPEG will require more bandwidth to transmit the same amount of video (same FPS, same resolution) as compared to MPEG4. the ration can range from 1:2 to 1:4 (MPEG4: MJPEG). Another (not so strong) indicator could be the processing power used when viewing the video streams. MPEG4 will require slightly more PC processing power as compared to MJPEG.

when the size of the project is large / complex, and the number of camera channels is large, careful selection of software is very important. As you have tested, there are many IP camera manufacturers who will throw in the software for free. Those software are probably worth only that much - free. ACTi, Axis, Etrovision, Vivotek, just to name a few, are known IP cam manufacturers, whose expertise is in the hardware, thus the software that is bundled is probably for very basic use - record, store, and some playbacks - for a limited number of cameras. For a project your size, you might want to consider the enterprise range of video management software, such as DVTel, Verint, Indigo Vision, Bosch, Genetec. The underlying database design & structure is thoughtfully designed and implemented - which means the software will remain stable and perform even if the camera count is huge. Hope this gives some insight.

DVTel has launched new encoders with their own encoding boards. The single-channel version is smaller than that of Verint's.

Problems with the Vidos-NVR - kindly elaborate. This is my 2 cents worth: - each NVR can have a maximum of 70 channels (license wise). - the NVR stores video in proprietary format, and without meta-data - the only way to search / export / playback is via their interface. Even if you use their SDK, it is not easy to get the video from the Vidos-NVR. You might want to consider their BVMS. It seems pretty interesting, only thing is that the BVMS only supports Bosch branded products (and maybe Axis), and is priced at the top tier.

I'm not sure if you're trying to stream 16 Arecont megapixel cameras onto a display monitor. If you are, then you might want to take these factors into consideration as well: 1. the total amount of bandwidth coming into the display PC 2. the pc specifications - the higher the better, since you're gonna run two dual monitor display card 3. how many other recepients will be displaying the same similar 16 channels I understand a 1.3MP camera at 20fps in MJPEG can consume up to 45Mbps per camera. [/b]

Hi, You can start with a site maintained by Axis. It makes good reading material, and should give some background information to IP video systems. If I could summarize, an IP Video system will consist of the front end (IP Camera), with the transmission medium carrying video over the TCP/IP protocol, and being recorded into an industry server (eg. HP, Dell) running a video software. Of course, this is a 1-liner, but different manufacturers will define "IP Video" differently, in an attempt to differentiate their offerings. (which invariably leads to confusion)