jhonovich

I recently released a video surveillance book. It's free and it's open meaning that you can copy and reuse parts of the book (it's a creative commons license btw). I see the book as a work in progress that will continue to improve. There are clearly many topics not covered appropriately in the book. If you are interested in contributing an essay, that would be great. You'd be helping the community and also getting your name/company's name out there. Please let me know if you have any questions or feedback. Best, John

Carl, All, A few points of information: - We (IPVM) bought at least $20,000 worth of cameras for testing last year. We prefer to buy our own. - As for the one manufacturer you claim we 'constantly' bash, you are referring to the system you bought for your site (IndigoVision). We have tested IndigoVision's products but we have not tested them in the last few years because their market position has shrank significantly. Recognizing their challenges, IndigoVision has shifted their strategy in the past year https://ipvm.com/reports/indigovision-vms?code=cf - Regarding your accusation that I did not refund your membership, we actually gave you extra $23.45 credit (nearly 3 more months) for commenting after you cancelled, and you replied, "LOL. I cancelled my membership weeks ago." We take our independence seriously and strive to be fair and accurate. I hope this helps clarify to CCTV Forum members our position and actions. [Mod - removed personal information]

Zakmak, This is John from IPVM. The pass rate averages ~40%. To pass, the most important thing is to (1) read all the readings and (2) do all the quizzes multiple times if needed. The most common cause of failing is people who do not make the time to read the readings and people who assume their own field knowledge will be enough to pass. Unlike the typical manufacturer course test, we want to make sure people know the technology well and don't aim to pass people just for showing up. If there are specific test topics you want feedback on, I am happy to provide it.

Are you positive they have the exact same MAC address? They are supposed to be unique. Dahua could have given multiple cameras the same MAC address, you could have gotten ones that were the same but sounds crazy if Dahua would allow that.

30' away is not far for even cheap illuminators or those integrated into cameras. The tri-illuminator that eeproject1 mentions is being sold on Amazon for $15, is self-rated for 50-100 feet, so even if the spec is liberal, you probably should be ok. Alternatively that same company that does the tr-illuminator has quad and six IR illuminators for just slightly more ($20 or $30). There's certainly a risk that they will fail but they are so relatively inexpensive probably worth going for those.

Trauts14, How far away from the camera do you want to illuminate and how wide of a FoV? IR LED count is not a good metric for tracking performance as there is great variation in IR LED performance and application.

Look under Zhe Jiang Dahua on the ONVIF list. That is Dahua's official company name. Many Chinese company names start with their province. That said, ONVIF could / should make it easier by listing Dahua under 'Dahua' instead.

VMD = video motion detection

I agree with mkkoskin about checking ONVIF's site / official directory. One thing that is not always true is this: "If a camera supports ONVIF Profile S and recorder supports ONVIF Profile S, they work together." Usually, if both devices are Profile S conformant and listed on the official ONVIF site, at least basic streaming and configuration will work. However, it is not guaranteed. There are still some instances of combinations not working. The other major issue even with official Profile S conformance, is that more advanced features, like PTZ controls, camera side VMD, analytics, etc., may not work. Camera side VMD is the most commonly used feature that frequently does not work with Profile S. If you need that, verify that your particular preferred combination of IP camera / VMS supports it.

Check this directory of integrators that includes a dozen from Utah at http://ipvm.com/integrators Also, if you are an integrator (anywhere) and do not find yourself in the directory, just use the submission link on that page and we will add you.

On the card type, Axis provided this statement to us: "Axis strongly recommends using SanDisk Extreme cards as these were extensively tested without timeout failure," adding that "Enhanced memory cards like SanDisk Extreme are much better suited for demanding applications, such as professional 24/7 surveillance applications. Lower graded consumer SD cards – mainly used for multimedia applications – will in many cases work fine, but the bigger variation in quality and durability of these cards makes them less suitable."

For each 16:9 resolution, there is typically a 4:3 resolution alternative that has the same horizontal AoV but a greater vertical / 'tall' AoV. Like so: IPVM has a free full test report on 4:3 vs 16:9 aspect ratio coverage areas.

VancouverBC, I am the founder of IPVM. We have a free 141 page IP camera ebook that teaches camera fundamentals, lens and camera selection. Here is the direct link to that ebook. On our paid member side, we have a huge array of training resources covering pretty much everything in video surveillance and access control. You can scan through this directory to see what we cover.

40th floor, you have provided no evidence to support any of you claims. To the contrary, we have a dozen plus independent reports from Axis end users with problems using SD cards. Indeed, as I have earlier cited, Axis has admitted the problem.

40th Floor, your opinion aside, I am relaying numerous significant field events of SD cards being used in production IP cameras.

Sunny, No, I have not seen any ONVIF supplied source code. They do have a test tool that one can test their own ONVIF implementation against. That said, to your broader point, there would certainly be value of them providing reference code to simplify and better standardize implementation.

40th Floor, Axis has acknowledged problems with SD cards used in cameras. In response, Axis tightened their recommendations. Here's a copy from the statement they shared with IPVM: "Axis strongly recommends using SanDisk Extreme cards as these were extensively tested without timeout failure," adding that "Enhanced memory cards like SanDisk Extreme are much better suited for demanding applications, such as professional 24/7 surveillance applications. Lower graded consumer SD cards – mainly used for multimedia applications – will in many cases work fine, but the bigger variation in quality and durability of these cards makes them less suitable."

I'd recommend that you get your software officially ONVIF G conformant. That way, it will make it easier for the camera vendors to have software to test against. From talking to manufacturers, it's a little bit of a chicken and egg dilemma to get G conformant. Each side needs to test against other conformant products, but those are in short supply. One thing that will help is that Genetec plans to be officially G conformant soon (they already have the code in production and they are on the technical committee). Once, Genetec is official, it will spur many of the camera manufacturers who want to get Genetec to support them. That should help you too. Btw, you are probably already aware of this, but the other big issue with edge storage is that a lot of (low to mid tier) SD cards have problems reliably recording video surveillance.

The bigger issue is getting VMSes to support ONVIF Profile G. Even if you had a cheap ONVIF Profile G camera, what do you plan to connect it to? While I am sure uptake of Profile G will be slower than Profile S (for streaming), it is just 3 months since G was officially released. I am sure there will be dozens, if not more than a hundred by early next year.

If you have quality problems with VBR, then you should look at what quality / compression / quantization level the camera is set to and adjust it. VBR does not guarantee 'good video'. It guarantees that whatever quality / compression / quantization level is selected is maintained. If you or the manufacturer defaults to low 'quality' then that is what you will get. Manufacturers vary on what they call this setting. Check yours. If it is called quality, increase it. If it is called compression, decrease it. Here is a summary of different manufacturer names and options:

"Basicly CBR (Constant Bitrate) will guarantee you better quality at all times, but at a higher bandwidth cost. All the systems we install, we install with CBR." Our testing disagrees with this. Also, your description is technically incorrect. CBR does not guarantee better quality. It guarantees the bit rate to be maintained. The 'quality' of CBR only comes from setting a very high bit rate to cover the worst case / most complex scenario. If you set the CBR bit rate too low, you will have poor quality. If you set the CBR bir rate too high, you will waste bandwidth / storage. VBR, by definition, does guarantee quality. It locks the quality / compression / quantization level, allowing the bitrate to fluctuate depending on the complexity of the scene. VBR is the way to go. In our testing of dozens of leading cameras, it does not have issues maintaining quality when motion increases, because the quality level is locked, it simply increases bandwidth of the stream.

We have released a Camera Calculator, it's a free online tool that lets you try out different lenses and resolution, dynamically showing what they look like and what type of pixel density those combinations provide. We looked at existing lens calculators and tried to make one that was easier to use plus shows realistic images. I'd be curious to hear your feedback and things you'd like to see or alternative tools you like better.

There's not an easy answer but there are a few rules of thumb to keep in mind: (1) The big difference with ONVIF support is "Profile" vs "Archived". Everything 2.x is Profile, typically S which is for cameras / streaming video. Archived is everything 1.x. In our testing 1.x products have a far greater probability of failing to connect / work than 2.x products. That does not guarantee that all 2.x products will work but it is much more likely. (2) Make sure to verify that the product is officially conformant with ONVIF. You can check online - here's for Profile products and here's for Archived ones. Lots of companies say they are ONVIF but have never officially conformed / verified performance. In our testing, products that are officially listed much more frequently work. (3) Even with Profile (2.x) products, the 2 most common features that most likely will not work are motion detection and PTZ controls. In our testing, we found some motion detection integration working but a number of them were assisted by custom enhancements from the VMS provider. Hope this helps. Bottom line - always verify ONVIF integrations yourself.

"To say one cow is insignificant is short sighted, they are all part of the greater good." This whole analogy is silly. First of all, even using it, this is the equivalent of Canon already having 500 cows and buying 1 more. It's literally a trivial move from a cash / milk generating perspective. More importantly, Canon, like all other major manufacturers selling surveillance, generates far more profits and revenues by cross-selling cameras and recorders. Returning to your original assertion, "I doubt Canon will kill a cash cow to sell more mediocre, yet expensive Canon cameras." Milestone does not generate a lot of cash, relative to Canon, the industry, or IP camera manufacturers. The most profit-maximizing move is to use Milestone products, sales channels and install base to help sell more Canon cameras.

Milestone is a tiny part of Canon, however you cut it. Even for their division, Milestone revenue is just 2%. And if Milestone is such a 'cash cow' as you claim, how is it that their 2013 net income (i.e., profits) was just ~$6 million? And what is $6 million in profits to a company that did $2.2 billion in profits in 2013? I return to your original proclamation: "I doubt Canon will kill a cash cow to sell more mediocre, yet expensive Canon cameras." Network cameras generate far more revenue and profits than VMS software. If Canon can use Milestone to sell more Canon cameras, and they can, just like Tyco is using Exacq to sell more AD cameras, it will generate far greater returns than simply selling VMS licenses.