Disqus - Latest Comments for jeffj

Re: China to Block Out Western Manufacturers?

Jeff Jones — Thu, 19 Apr 2012 17:59:51 -0000

I have seen similar manouevers from the Chinese in the past with broadcast digital television. At that time (early 90's), they changed the polynomial for the Reed-Solomon CODECs (error correction) slightly so that they could control who provided the chips and who could broadcast and receive programming.
In this case my bet is that this is a business tactic driven by the chip manufacturer to give themselves an advantage in the market. Any new standard will ultimately be driven by the Ministry of Public Security because someone there was somehow convinced that this was the right way to go.

Re: Who the Hell Killed the Pelco Brand???

Jeff Jones — Tue, 20 Sep 2011 21:46:33 -0000

John,

I agree with you. Their ASIS booth, whether intended to or not, has clearly made the Pelco brand nothing but a footnote to their overall corporate brand. I've had discussions with others, and it appears to be a widespread perception that this was a marketing travesty and just one more step in the deconstruction of what many used to think of as a major force in the security industry. Emphasis on one more step since there have been other less obvious ones.

- Jeff

Re: Testing Viaas Hosted Video / VSaaS Offering

Jeff Jones — Fri, 30 Apr 2010 15:55:48 -0000

Note that their web interface appears to use Flash, so any mobile device used to access the service must support Flash. Apple mobile devices will not work.

Re: Training: Resolution of IP Cameras

Jeff Jones — Tue, 20 Apr 2010 17:33:42 -0000

A couple of things that were not mentioned...

Lenses - You need a lens that supports the resolution you are trying to capture. Fitting a sandard lens on a megapixel camera will actually reduce the resolution the camera is capable of capturing.

Pixel Size - Assuming the same optical format (i.e.; 1/3 inch), more pixels means smaller pixels. A pixel can be thought of as a bucket that catches light. The smaller the bucket, the less light it can catch. That means less sensitivity, and that means reduced low-light performance.

Re: How Much Storage is Needed for Video Surveillance?

Jeff Jones — Tue, 18 Aug 2009 12:23:51 -0000

John,

If I'm reading your conclusion correctly, you're saying that 600GB in 5 years will be at a similar cost to what 60GB is today. Are you talking about enterprise drives, or commodity/consumer drives? And do you have a source for the cost curve you are using? I'm not challenging the numbers, I just want to see what they're based on.

Most enterprise DVRs use enterprised drives which typically have a cost curve with a different slope than commodity/consumer drives. It could be argued that enterprise drives should have a steeper cost curve as corporate installations move toward centralized storage, but because of RAID and multi-site mirrored storage, a lot of large enterprises are actually using commodity drives rather than enterprise. Google is a prime example, and they have published a paper as to why they have chosen to go that way. Granted, their data isn't as mission critical perhaps, but it shows a trend with a rationale.

Re: Calculating Storage/Bandwidth For IP Video Surveillance

Jeff Jones — Tue, 28 Jul 2009 15:45:15 -0000

Great article. There are a few things that affect compression efficiency that people rarely notice or think of:

1. Noise - To the CODEC, this is just another form of scene complexity. Noise doesn't compress well at all. Noise is the single largest contributor to huge increases in bitrate while capturing low-light scenes.

2. Over-sharpened images - Some image processors tend to over-sharpen the image to give a perception of higher resolution to the viewer. Over-sharpening creates high-frequency transitions, and the CODEC will typically attempt to preserve this "false" detail, and in the process will increase the bitrate.

3. False motion - Interlaced capture results in temporal artifacts between video fields that aren't efficiently handled by CODECs, so you're better off using progressive image capture. Frame based CODECs working on progressively captured images can offer much lower bitrates than either field or frame based CODECs working on interlace captured images. This is because there is almost always better correlation between pixels in the vertical direction when they are captured progressively.

Re: IP Video Industry Weekly - July 17, 2009

Jeff Jones — Fri, 17 Jul 2009 16:22:58 -0000

There is no way to achieve higher than 540 TV lines of resolution in a standard analog video security system. DVRs and display monitors low-pass filter at their inputs, and the ITU standard that governs how the video decoder in the DVR converts the standard analog baseband video into a digital format (so that it can be compressed and stored in the DVR) limits horizontal resolution to 540 HTVL. Sure you can use a studio monitor and possibly measure some of the high-end japanese cameras at 550 HTVL, but the end user derives no benefit in a real-world application. It's all specmanship.

Re: Only 5% market share for IP Surveillance Cameras?

Jeff Jones — Wed, 15 Jul 2009 18:50:55 -0000

Trade show exhibitors are always looking for some sort of new "sizzle" to get people to come in to their booth. In 2008 that sizzle was from IP, and thus far in 2009 it has been from the terms "megapixel" and "HD". So, if you sprint through the show floor without stopping to talk in detail with exhibitors and take a look at their entire range, it's easy to get the impression that everyone is doing XYZ and has abandoned all else. In a way, that can be dangerous for an exhibitor unless they're very strongly positioned within the "sizzling" market.

I think that Bosch did a great job with their positioning at ISC West this year where their tag line was, "Think IP. Maximize Analog." They clearly let you know they hadn't abandoned their mainstay.

Re: Only 5% market share for IP Surveillance Cameras?

Jeff Jones — Tue, 14 Jul 2009 20:38:27 -0000

Good point, although I was talking about 2009. Then one or more of the following is true:

1. I'm way off on ASP
2. Their revenue contribution from streamers and software is significant
3. They don't have a 33.5% market share (They've been pretty consistent in claiming 33% market share)
4. IP cameras are much less than 10% of the market, or the overall security camera market is much smaller than 40M units
5. Their volumes increased significantly since the beginning of 2008 (they were talking about a 40% CAGR back then, I believe)

I'm not sure that $1000 cameras are big sellers. That's why I picked the 216FD. But I'd be interested in the opinions of others in this group on that as well as the rest of the above.

Re: Only 5% market share for IP Surveillance Cameras?

Jeff Jones — Tue, 14 Jul 2009 19:34:42 -0000

I know that simple math is sometimes flawed, but it's generally good for a sanity check.

Researching parts of this discussion further, I stumbled on a June 25th article that says "Axis Communications has continued its No.1 position in network video cameras market with 33.5% market share, according to IMS Research." (http://www.nationmultimedia...

I then took a look at the Axis 2Q09 interim report citing flat sales of 542M kronas (~$69M USD) of which 98% is from video sales. Not accounting for streamers, software, etc., and assuming an ASP of around $200 (I may be wrong, but their 216FD looks like a workhorse and sells at retail for just under $600), I calculate that they sold around 338K cameras in Q2 or a projected 1.35Mu in 2009. If the 33.5% number is correct, then the network camera market is around 4M units in size. If it's really just shy of 10% of the overall surveillance camera market, then we're talking about an overall market that's around 40M units in size. Of course, I could be wrong about their ASP, but in the absence of per-model or categorized sales data, you need to make some assumptions and I think this is a safe one.

This is far less scientific than the way I usually go about this sort of thing, and it only includes the numbers from one publicly traded company and one analyst, but it gives a quick "back of the envelope" sanity check for the numbers that are being tossed around on this discussion board.

Having said all of this, I believe that the 40Mu number is fairly accurate. I believe that the 10% number is somewhat high. I'd guess more like just south of 8% based on 2008 "derived actuals" and 2009 growth projections, even though Axis has indicated that demand has softened. Please note that my company is agnostic to IP versus analog, but we do find it useful to segment the overall market by camera type and geography so that we have data to guide us in prioritization and decision making throughout the year.

Comments are expected, welcomed, and appreciated. :-)

P.S. On a different note, I'm wondering who in their corporate communications group picked a photo of someone staring over a precipice (as if they had already reached the proverbial peak) for inclusion in their slide deck given their reported outlook on business and the economy (see slide 3). Bad choice.

Re: Axis Market Weakness Continues in Q2 2009

Jeff Jones — Mon, 13 Jul 2009 17:31:18 -0000

A lot of us in this business estimate the market size by looking at a variety of information sources including published figures from public companies, published retail pricing information, published data various industry analysts and trade publications, and verbal communication with people we know in the business. You collect all of the information and then judge it to see what makes sense and what does not. It's more of an art than a science. While Craig's 42M unit number does not match the one I'm working with exactly, it's within the margin of error and is quite credible as is his share estimate for IP cameras.

Re: Axis Market Weakness Continues in Q2 2009

Jeff Jones — Wed, 08 Jul 2009 17:17:12 -0000

Does anyone care to take a guess at what the growth rate for IP will really be for 2009? Axis is talking about 12% and Mobotix talked about 30% in their May announcement. I've heard other numbers in that range, but it would be interesting to know what others are forecasting for this year and 2010. It would also be interesting to know if there are any predictions as to if and when the days of 40% growth are expected to return.