Member Discussion

Super High MP Cameras Like Avigilon Or Multiple Mp Cameras?

I'm working on a project for a rather large parking lot. No way to put cameras out in the lot that the customer will accept. He wants to use super high resolution cameras from the roof of the building. He is suggesting Avigilon 16 Megapixel. Is this approach appropriate?

I say high megapixel cameras have high costs. Typically, the larger the pixel count, the slower the frame rate and the poorer the nighttime quality. I want to stay in the 7-10 frame per second area, and have minimal blur at low light levels.

I'm thinking of using multiple 5 megapixel cameras with a more narrow viewing angle to keep the pixels per foot count higher at longer distances. I can put many of those in for the cost of one 16MP camera. That also gives me a more redundant system and less cost of a single failure. It also keeps my frame rate up.

I don't see any tests of super high megapixel cameras, and I really doubt some of the hype the manufacturers are claiming.

Chuck I tend to agree with your strategy of putting in multiple lower resolution cameras to cover the lot. Even at 5 megapixel you're going to see relatively poor light performance compared to 720p and 1080p, but that's starting to improve (Bosch starlight 5MP for example).

We have a test of 10MP vs. 5MP vs. 1080p vs. 720p you may want to check out. My main concern would be the light level in the parking lot. Is it lit at night?

I would call light level standard parking lot levels - about 5 lux.

I have not seen any recent tests of the newer super high megapixel cameras. I checked the Avigilon 16MP camera. It has an image sensor of 1.12" with a minimum illumination of .005 lux at F1.4. Using one of their recommended lens - Sigma 18-35mm, f/1.8 - can it produce parking lot images - way beyond the 100' range? I'm talking about 500' - 1000' and beyond.

Customers want to see a parking lot crash or breakin after the fact. Not necessaily a face, but watch the person leave and get in another car, or where they walk.

Many of the security directors don't have the "CSI effect" request when asking for this stuff. They do have realistic requests. Unattended PTZs are quickly going away when they can spend the $5K on a fixed - always recording solution - and at least see the complaint.

For what you're talking about, the multiple individual cameras will probably give you a better result.

I've seen the high-res (10MP+) cameras used in similar scenarios where you can legitimately replace 4 or 5 lower-res cameras that would have usually been covering specif spots, and then you also get all the detail/views in between.

Even though a 16MP camera might have 8x the resolution of a 2MP camera, it wouldn't neccessarily replace 8 2MP cameras, it might replace 3 or 4, and then also add additional scene context and info that would have ordinarily been missed.

I would caution you not to assume the "super high res" cameras are high cost, low frame rate, or anything else. You can determine those things directly and do a practical side-by-side comparison and then use the right approach.

We have not tested Avigilon's PRO series, so I cannot speak specifically to their performance.

As you and others have noted, the big structural things are (1) price, (2) frame rate and (3) low light performance. The other valuable element of multiple cameras is that you can adjust their angle to specific areas of interest to match sure you get direct head on shots (whether it is license plates on vehicle entrance, head shots of people walking down a sidewalk, etc.).

More angles will be far better than a single high res location.

even if you can't put cam's in the lot, can you get a few angles around the lot?

I'd even go 10 x of the $80USD 2CD-2332's, they seem pretty awesome for the dollar, if you must, you can run all of them on a 12mm lens.

I guess I'd first ask for clarification on one point. WHY do you feel the customer will not accept cameras in the lot? That's a game changer in terms of price and capability. It's not that expensive to put multiple wireless cameras up in the lot with multiple viewing angles then add some very high res "overall view" cameras. Would be less expensive and more effective.

First - I stay away from wireless anything in a metropolitan area. I've been burned way too many times with wireless devices going in and a year later they quit when other devices light up. We once lost a customer over it.

Second - cameras will still need power. When the poles are 480 volts switched, battery powered, charging when on or using solar power is just not worth the trouble.

Third - I'm selling against the 16MP camera that they insist is the way to go. "Four of those cover the lot the way I want.... "

Third - I'm selling against the 16MP camera that they insist is the way to go. "Four of those cover the lot the way I want.... "

Have they deployed even one, even temporarily in their parking lot? That would go a long way to proving their conviction? Put up a 16MP and a 5MP side by side, let them see, etc.

Have they deployed even one, even temporarily in their parking lot? That would go a long way to proving their conviction? Put up a 16MP and a 5MP side by side, let them see, etc.

They had a demo - they recorded it. It was done during the day, nice even sunlight - perfect conditions. I pushed to see it, to make sure they didn't want the CSI effect of looking at a license plate at 800 feet after the fact. They have a good practical expectation for that distance. To me, the moving cars moved major distances between frames. They insisted it was smoother - more frames per second than what they recorded.... hmmmmm...... I then asked about night time view. Again he insisted it would be good coverage, it's what he wants. I went through the calculations based on the part numbers he provided. They have not given me the opportunity to do any demos yet. I have demonstrated other projects we have done, but none of them had this big of a lot.

It seems that the buyer has made up his mind and is just trying to justify his decision.

Did they really not demo / try it at night at all? Do you know the f number of the lenses they are proposing? Did they try a lower resolution camera side by side to see what the difference / gain would be?


I am in favor of multiple cameras to cover a given area. I am trying to find camera mounts that would allow me to mount them properly. Any idea of somehting that would allow mounting multiple cameras on a pole or a wall to align them properly while covering the desired angle?


NOTICE: This comment has been moved to its own discussion: How To Mount Multiple Cameras On A Single Pole?

Btw, here is a good counter example of an Avigilon dealer using (3) 3MP cameras instead of a 'super high resolution one' to replace a PTZ:

I think that is a solid, sensible approach.

Is the 'super high resolution' option even viable in this case?

Even assuming you had 29 megapixels, what kinda of lens would you need to get that FOV, which looks like close to 180? Canon fisheye? Do people ever do that? If they have, it would be an interesting image....