lux rating

[capz 0]

When manufacturers cite a lux rating are they giving you a minimum illumination to get any picture at all or is it min. illumination to get a clear picture? All I'm finding are reports and pictures of noisy images in low light.

 

The more I look at ip cameras for my low light install, the more I'm starting to regret choosing ip over analog. It's frustrating to find that as the technology advances to megapixel, the price of the cameras go up but the output quality gets worse. Is it all hype or am I not looking hard enough?

[Soundy 1]

Well, all technology has to progress in steps. Keep in mind that a megapixel IP cameras' sensor is the same size, or only slightly larger, than most low-light analog cameras' sensors, while containing four or more times the number of pixels... that means each pixel must necessarily be about 1/4 the size, and thus able to collect a lot less light.

 

To find a megapixel camera with the same low-light capability, you need one with a much larger sensor... and a correspondingly larger cost. "I canna change the laws o' physics, Cap'n!"

[capz 0]

Thanks for the reply. Nice explanation. To help me understand it better, the amount of light the sensor can collect is fixed based on sensor size and number of pixels. That answers my next question, which was, can you lower the resolution of a megapixel camera to VGA in order to get a better low light image? Answer: No

It looks like with the current technology we'll stay away from megapixel, unless I can find one with 1/2'' sensor, integrated IR, under $800. If any suppliers read this and want to let us demo a unit that might work for us, feel free to pm.

Thanks again Soundy.

[Soundy 1]

Thanks for the reply. Nice explanation. To help me understand it better, the amount of light the sensor can collect is fixed based on sensor size and number of pixels. That answers my next question, which was, can you lower the resolution of a megapixel camera to VGA in order to get a better low light image? Answer: No

 

Answer: yes. There are plenty of IP-based 4CIF/D1 cameras out there. They're not as "hyped" because adding IP is a fair extra cost that doesn't hold any benefit for the image quality; the main "benefit" of IP isn't the IP medium itself, but the fact that it's needed (at least with current technology) to support resolutions over 720x480, beyond the range of analog video standards.

 

It looks like with the current technology we'll stay away from megapixel, unless I can find one with 1/2'' sensor, integrated IR, under $800. If any suppliers read this and want to let us demo a unit that might work for us, feel free to pm.

Thanks again Soundy.

 

Most (all?) of the IQEye cameras are 1/2", but that's really not much larger than a 1/3" - not enough to significantly improve the light collection capability. To get to that point, you need to get into Avigilon territory (http://www.avigilon.com), whose top 11MP and 16MP cameras have sensors the size of a 35mm film frame.

[cglaeser 0]

When manufacturers cite a lux rating are they giving you a minimum illumination to get any picture at all or is it min. illumination to get a clear picture?

 

The real problem is that few manufacturers include the shutter speed when citing a minimum lux value, and many of the new IP cameras support a shutter speed as low as 1 or 2 seconds. For example, I have an Axis 223M megapixel day/night camera. The data sheet lists a low light value of 0.2 lux. However, my testing indicates that this 0.2 lux requires a 2 second shutter speed, which is useless for anything but stationary objects. My front driveway is illuminated with a 5 lux street light, and even that is too low for the 223M using a 1/30 shutter speed.

 

So, to answer your question, when you compare lux values, you must also compare shutter speed. Since most manufacturers fail to include the shutter speed, it's difficult to do a meaningful comparison.

 

Best,

Christopher

[capz 0]

Good point. And it does make a BIG difference. We have a Vivotek loaner that doesn't handle low light very well. When tweaking the settings, changing the shutter speed was the only thing that cleaned up the noise noticably. Still not to an acceptable level though.

I imagine there's a lot of false advertising going on with all the competition out there. Many companies just sell the product on marketing and hope people won't try and return it when it doesn't stack up.

However I did notice many of the more reputable names do cite the shutter speed, namely Arecont, Mobotix, Axis. Those that do, do so probably because they know they're selling a superior product, and their sales and prices reflect that.

[cglaeser 0]

I imagine there's a lot of false advertising going on with all the competition out there.

 

Before the introduction of IP cameras, I'm guessing most manufacturers were using 1/30 and therefore did not specify the shutter speed with the minimum lux value. Now that many IP cameras support DSS with very low shutter speeds, it's become impossible to compare the lux values in a meaningful way since we can no longer assume a 1/30 shutter speed was used.

 

However I did notice many of the more reputable names do cite the shutter speed, namely Arecont, Mobotix, Axis.

 

While it is true the Axis does specify the shutter speed range supported by each camera, they do NOT specify the shutter speed used to measure the minimum lux value. Based on my measurements of the 223M, I believe they use the slowest supported shutter speed when reporting minimum lux value, which is very misleading. In contrast, Mobotix does report the shutter speed with each reported minimum lux value, and they often cite two lux/shutter speed combinations (one realistic shutter speed and one very slow shutter speed).

 

Kudos to Mobitix. Axis needs to update all their data sheets to include the shutter speed with those lux values.

 

Best,

Christopher

[capz 0]

I double checked the Axis datasheets. You're right, all they give is the shutter speed range. I see what you're saying. It would be more useful to cite Min. Illumination @ ??? shutter speed.

[cglaeser 0]

It would be more useful to cite Min. Illumination @ ??? shutter speed.

 

Precisely. C'mon Axis, get with the program!

 

Best,

Christopher

[Malloot 0]

And to make it even more difficult to compare lux-ratings.

Generally the opening of the Iris isn't mentioned and loss of light through the lens is even not mentioned in the specs of the lens manufacturer.

 

Generally you have to see it, to believe it.

[zmxtech 0]

All the marketing crapola aside, It comes down to experience and the saying "no light=no picture"

 

anything under about 10lux needs lighting as a general rule even if the picture looks good the shutter will be slooooooooooooow

And the guy that just robbed that store will be running !

 

matching the camera to the situation and lighting can be hard work.

 

If you cant use heaps of normal lighting go IR the shutter will speed up

and you will will have a much better image.

 

Yes IP camera's tend to fib !

 

1Axislux=100lux

1Actilux=10lux

1Arecontlux=2lux

 

and so on ........

[cglaeser 0]

Yes IP camera's tend to fib !

 

1Axislux=100lux

 

Yes, that's about right for the Axis. What about Mobotix?

 

Best,

Christopher

[zmxtech 0]

1mobilux=2lux

[robert 0]

Yes, megapixel cameras are crap with low light, but not all. Check CCD megapixel cameras - they are way better than CMOS at night.

[cglaeser 0]

Check CCD megapixel cameras - they are way better than CMOS at night.

 

What are your favorite CCD megapixel cameras?

 

Best,

Christopher

[robert 0]

Check CCD megapixel cameras - they are way better than CMOS at night.

 

What are your favorite CCD megapixel cameras?

 

Best,

Christopher

 

Many manufacturers have them now - Sony, Samsung, Messoa, Lumenera, Geovision