CCTV camera test

[MarioCCTV 0]

Hi all,

 

I'd want to test a few cameras but I'm not so expert on how to perform it well.

 

I've built a sort of test box (4m long, 1m wide, 0.8m high) where I will install my camera and do all the tests.

 

On one side of this box I fixed a test chart that I bought from vidilabs.com. The camera, on the other hand, can be placed anywhere along all the lenght of the box.

 

I illuminate the chart with 4 tungsten lights (that i can move along the box) controlled by a light dimmer. At approx 0.5m from the chart, I fixed 4 fluorescent lights on both side of the box (2 per side). There is no external light.

In this way I can test the cameras' quality at different light conditions (eg. fluorescent/tungsten lights, 1000 lux, 400 lux, low light conditions). Illumination is measured (using a lux meter) in the centre of the chart as close as possible to and parallel to the chart plane.

 

The videosignal can be analyzed using a Video Measurement Set (Tektronix VM700T).

 

Now I know that I can measure the performance of the camera and compare different cameras if I do the measures with the same light conditions and recording the same chart.

In this way I can objectively measure:

Resolution, S/N, color output. Anything else?

 

Furthermore, looking at the details on the chart, I can measure:

reflection, licence plates recognition capability, face recognition capability, intrusion detection capability.

But i guess that for these measures I can just give a subjective value for example using a scale from 1 to 10. Am I right? If not, please suggest me how i can perform an accurate measure of these features.

 

 

Anyway, I have several doubt regarding how to perform these tests:

 

Lights:

Should i care about light temperature?

is there a "recommended" light (and lux level on the chart) to use for the objective measures i mentioned above?

Should i fix the tungsten light in a stable position (e.g. 3m from the chart)?

 

Camera settings:

For varifocal cameras, should i set the lens at minimum, middle or maximum focal lenght?

When possible, should i switch off all the video processing features of the camera (AGC, BLC, etc)?

 

Hope you can help me!

Thanks in advance!

 

Mario

[rory 0]

Only real way to test a camera is to take it outside, test it in the day and night, and with pitch dark - its alot of work.

[MarioCCTV 0]

mmmh.. how can you compare different cameras in that way?

The only way would be recording the same subject at exactly the same time. That's something I can't and wan't do.

 

I would like to have a system where I can test my cameras (and compare them) whenever i want..

[cglaeser 0]

mmmh.. how can you compare different cameras in that way?

The only way would be recording the same subject at exactly the same time. That's something I can't and wan't do.

 

I would like to have a system where I can test my cameras (and compare them) whenever i want..

 

Use a sturdy mount that can support at least two cameras. A sturdy piece of wood should work fine. If you can compare two cameras at a time, that will provide a great deal of objective and subjective information. You can C-clamp the mount to a sturdy A-frame ladder for height perspective. Use cables with sufficient length so you can move the ladder to relevant locations. You should be able to assemble all of this in an hour or so with readily available hardware.

 

Best,

Christopher

[rory 0]

mmmh.. how can you compare different cameras in that way?

The only way would be recording the same subject at exactly the same time. That's something I can't and wan't do.

 

I would like to have a system where I can test my cameras (and compare them) whenever i want..

 

I mount them outside on a wooden siding outside my garage, and also test it in a much darker location on the other side, same exact spots, one at a time - if you can mount more than one then fine, I dont want to leave that many holes in my siding though so I do one at a time, plus I rarely have more than one or two to test at a time. Leave it up 24 hours in one spot and compare them at the same times under same conditions. Biggest problem with Manufacturers and OEMs is they dont properly test their cameras. You must test the camera outdoors no matter what, its totally different from indoors. For one thing the lighting outside will differ at night and it is a much larger area for the camera to cover with much more to process. During the day outdoors there is Infrared from the sun, that is where you test the color and the light handling under glare.

 

Bottom line is unless it is tested outside during the day with full sunlight and at night under artificial lighting, under low light, and under very dark, and in at least 2 different locations, then its no use testing it. If the camera has Infrared then additional testing is required as well.

[Guest]

mmmh.. how can you compare different cameras in that way?

The only way would be recording the same subject at exactly the same time. That's something I can't and wan't do.

 

I would like to have a system where I can test my cameras (and compare them) whenever i want..

 

I agree that testing in the field is the most appropriate way of system testing.

 

The older approach was with Rotakin (still used in high end CCTV in the UK and Europe) and the new test target system set up by the UK Home Office, which looks at image quality at the monitor in three modes (live, recorded and downloaded).

 

This new approach is more suitable to this new digital age and reflects real world requirements.

 

Ilkie

[MarioCCTV 0]

thanks for your reply!

 

Anyway first of all i would like to check if Manufacturers and OEMs tell the truth in their specifics for the cameras they sell...

So, if they say that cameraA has 320 TVL, I want to check if that camera has actually 320TVL.

 

Then i would like to test S/N and other values that are not subjective.

 

Please give a look at www.vidilabs.com to have an idea about what i want to do!

 

Thanks again!

[tomcctv 187]

thanks for your reply!

 

Anyway first of all i would like to check if Manufacturers and OEMs tell the truth in their specifics for the cameras they sell...

So, if they say that cameraA has 320 TVL, I want to check if that camera has actually 320TVL.

 

Then i would like to test S/N and other values that are not subjective.

 

Please give a look at www.vidilabs.com to have an idea about what i want to do!

 

Thanks again!

 

 

mario. i think i agree with the others. and the site you listed is something i would never use. it would just be a waste of money. most installers will have a number of cameras that they trust and know they do the job. then the main test is camera to dvr. you can have a good image on one dvr and bad on an other. you get some people spending $200 on a camera and connecting it to a $60 dvr from ebay and the first thing they will say is the cameras crap ............... i think doing test on big brand name makers just would not work. . its like the test box you have built ............. you can do all the tests and post your findings...... them someone else can make a bigger test box with true daylight fittings ...and the reports will be differant.

[Soundy 1]

Bottom line is unless it is tested outside during the day with full sunlight and at night under artificial lighting, under low light, and under very dark, and in at least 2 different locations, then its no use testing it.

That's a little over-generalized. The best way to test any camera is under the circumstances you plan to use it in. If I'm looking for the best camera to go inside a walk-in cooler, knowing how it performs in bright sunlight really won't be of any use to me. The lighting conditions inside a walk-in cooler don't usually differ much from one to the next.

[rory 0]

Bottom line is unless it is tested outside during the day with full sunlight and at night under artificial lighting, under low light, and under very dark, and in at least 2 different locations, then its no use testing it.

That's a little over-generalized. The best way to test any camera is under the circumstances you plan to use it in. If I'm looking for the best camera to go inside a walk-in cooler, knowing how it performs in bright sunlight really won't be of any use to me. The lighting conditions inside a walk-in cooler don't usually differ much from one to the next.

 

Yes if one only ever needs a camera for one single application then that is true, however a camera can be used for many different applications and that is why when reviewing a camera it is imperative it be tested outside. Im not talking about just testing a camera for a single job, I am talking about really testing a camera, for the benefit of selecting it for other future applications. If one does only test it in a walk in cooler, then they can only tell us how it works in a walk in cooler. When they now need to select a camera for outdoors, they wont know how that camera will work outdoors - though obviously if it is just an indoor camera like a smoke detector hidden camera then an outdoor test would be useless in that case.

 

The whole idea of testing a camera is to make future selections fool proof, or at least less of a gamble. In other words, testing it inside a walk in cooler does not tell us how the camera performs in general, its like saying a certain car is great parked in the lot with the engine running, but never taking it out on the road. We just cannot go by specs alone anymore.