Jump to content

Daylight Balanaced cmos


Jed Shepherd

Recommended Posts

I would think that all CMOS/CCD chips are ideally balanced for either daylight or tungsten light and, as-such, every camera is natively cleaner at either a daylight, tungsten, or intermediate Kelvin temperature setting.

 

Don't some HD cameras actually use optical filters for color correction, or is that the viewfinder only that I'm thinking of?

Link to comment
Share on other sites

  • Premium Member

Don't some HD cameras actually use optical filters for color correction, or is that the viewfinder only that I'm thinking of?

 

 

Karl are you being serious?

Because you do know that film camera's also use optical filters for colour correction right?

Link to comment
Share on other sites

  • Premium Member

Thanks A. Makes sense since they probably have equal amounts of RGB sites, and tungsten would be disproportionately red light (so more blue and green sites needed natively).

 

Actually it has to do with the native sensitivity -- or lack thereof -- of the silicon under the RGB filter mask. It's not as sensitive to blue as to the rest of the spectrum, so giving it more blue by going for a daylight-ish 5000 Kelvin yields more image photons and therefore a higher signal to noise ratio in the blue.

 

There are two widely used mask patterns. Bayer uses twice as many greens as red or blue (the four squares to the top left of this post), the Sony and Genesis use a stripe pattern with equal numbers of each.

 

 

 

 

-- J.S.

Link to comment
Share on other sites

  • Premium Member

Actually it has to do with the native sensitivity -- or lack thereof -- of the silicon under the RGB filter mask. It's not as sensitive to blue as to the rest of the spectrum, so giving it more blue by going for a daylight-ish 5000 Kelvin yields more image photons and therefore a higher signal to noise ratio in the blue.

 

There are two widely used mask patterns. Bayer uses twice as many greens as red or blue (the four squares to the top left of this post), the Sony and Genesis use a stripe pattern with equal numbers of each.

 

 

 

 

-- J.S.

Strictly speaking, it's more to do with the fact that red light penetrates further into silicon than blue light, which is the basis of the colour separation system used in Foven sensors.

 

In the past the convention was to have cameras natively white balanced for tungsten, because camera tubes are more sensitive to blue and green than red. That worked out nicely, because tungsten lights are deficient in green and blue, and 3200K turned out to be pretty close to optimum white balance for Saticon tubes and the like. In daylight you usually had light to throw away, and so you could put on an orange filter to knock out some of the green and blue without introducing noise

 

With silicon sensors, the situation is reversed; the silicon sensors are more sensitive to blue than red, however for a while the convention was to retain a "native" white balance of 3200K, simply for compatibility with existing tube cameras.

 

The first generation of CCD broadcast cameras had to have peltier solid state coolers fitted to the blue CCDs (the things they have in compact 12V fridges), because the blue channels were so noisy in tungsten-lit studios!

 

The situation with single-chip sensors is more complex, because you can't individually adjust the ADC scaling for the red green and blue channels like you can with a 3-chip camera. If the scaling is set for full output from the red-filtered pixels, it's going to be way down for the blue.

 

With a 12-bit ADC, if the maximum red output is 4096 levels (ie 212), on 3200K balanced white, the blue-filtered pixels might only be putting out 512 (28levels or even less. One of the many drawbacks of single-sensor cameras, and one of the many reasons 3-chip designs are alive and well and even appearing in high-eng consumer cameras.

 

As far as the best white colour to use for the RED, surely the answer is whatever RED recommend!

Link to comment
Share on other sites

  • Premium Member

In the past the convention was to have cameras natively white balanced for tungsten, because camera tubes are more sensitive to blue and green than red. ....

 

With silicon sensors, the situation is reversed; the silicon sensors are more sensitive to blue than red, ...

 

Did you mean "red than blue" in that last sentence?

 

 

 

 

 

-- J.S.

Link to comment
Share on other sites

  • Premium Member

Did you mean "red than blue" in that last sentence?

-- J.S.

More or less :rolleyes: \

It was actually supposed to be:

"the silicon sensors are less sensitive to blue than red"

but yours is an acceptable workaround :P

Link to comment
Share on other sites

  • 3 months later...

The surprising development is that manufacturers of single chip cmos cameras do not demand the use of cc filters?

But once a photo site is saturated thats it and it makes sense, in theory, to at least roughly balance the optical path to match the sensor to avoid massive amounts of gain in one or two channels being added within camera or in post to rebalance the image.

 

Tests by end users tend to not explore this aspect, perhaps because it is a convenient truth not to have to use cc filters on location:)

 

 

 

 

 

Mike Brennan

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...