C100: native color temperature of the cinema series' sensors?

D

drummondb

Well-known member
Hi there. In the heady days of the DSLR revolution, I remember reading from shooters who were experimenting with leaving the camera's white balance set to daylight, and would convert to tungsten situations with a color filter on the lens, the way one shooting film would do. The argument was the camera's processor would jack up the blue channel to compensate for tungsten situations, thus introducing more noise, at least more blue noise.

I was just tinkering around with my C100 mkII and my Ninja Blade monitor. There's a setting for "Blue only exposure" which the manual says "Displays a black-and-white image that is based on the blue channel of the input signal. Used to observe the noise content of a video image and the judge the overall exposure quality."

Truth be told, this black and white image did look quite different based on which white balance setting was used. Certainly seemed quite a lot noisier the lower the kelvin number got. I also jacked up the ISO, put a lens cap on, and moved through the kelvin setting. No real surprise, the lower the kelvin number got, the blue noise became quite pronounced, while red noise became more and more strong as I moved to higher kelvin values.

So is there a "native" color of the sensor in, say, a C100? A setting where the noise levels are equal between red, green, and blue? I found an old 80A filter laying around. So I did a comparison of the camera set to tungsten white balance indoors at night. And then I switched to the daylight preset and held the 80A filter in front of the camera. I thought I was on to something until I realized the 80A cuts quite a lot of light out. Maybe as much as 2 stops? So then I guess the question is, does the C100 look better with higher ISO on daylight than it does a lower ISO on tungsten?

Thoughts?
 
Last edited:
drummondb said:
Hi there. In the heady days of the DSLR revolution, I remember reading from shooters who were experimenting with leaving the camera's white balance set to daylight, and would convert to tungsten situations with a color filter on the lens, the way one shooting film would do. The argument was the camera's processor would jack up the blue channel to compensate for tungsten situations, thus introducing more noise, at least more blue noise.
Click to expand...

Yup. That is how it works. White balance is a processing of gaining up the channels that need gaining up. Even if they get noisy.

drummondb said:
So is there a "native" color of the sensor in, say, a C100?
Click to expand...

There is. But you'll probably have to find it by experimentation. Not many manufacturers will publish that kind of information.
 
The native color temp of silicon sensors , The black body temperature at which red, green, blue wavelengths produce equal voltage output, is between 5000K and 5500K. Theoretically the camera will produce equal noise in all three color channels under light filtered to this range. The color temp setting of the camera, which controls encoded video color balance, only modifies the gain balance applied afterwards.
 
Is there any way to filter the light upon entry to the lens without drastically affecting how much light the sensor reads?
 
drummondb said:
Is there any way to filter the light upon entry to the lens without drastically affecting how much light the sensor reads?
Click to expand...

Depends on how much of a change you want to make, and how you define "drastic". Filtering is by definition a subtractive process. You'll always have fewer photons after filtering.
 
Bruce Watson said:
Depends on how much of a change you want to make, and how you define "drastic". Filtering is by definition a subtractive process. You'll always have fewer photons after filtering.
Click to expand...

Well I meant it looks like the strongest color conversion filter to convert daylight to tungsten results in a 2-stop loss. If you're in a controlled shooting environment, like a feature, that wouldn't matter. But I'm talking avoiding ugly high-ISO noise in low light situations, like a wedding reception. Then the question would be, is a lower ISO with tungsten WB better or worse than a higher ISO with a daylight WB. I suppose I must test.
 
drummondb said:
Well I meant it looks like the strongest color conversion filter to convert daylight to tungsten results in a 2-stop loss. If you're in a controlled shooting environment, like a feature, that wouldn't matter. But I'm talking avoiding ugly high-ISO noise in low light situations, like a wedding reception. Then the question would be, is a lower ISO with tungsten WB better or worse than a higher ISO with a daylight WB. I suppose I must test.
Click to expand...

Higher ISO is worse. You're blocking light with the filter. Maybe you will have more even distribution of noise in each channel but it will be noisier overall.

If you want a cleaner image, get an A7SII or A7S or just get faster lenses. Really the only camera that's meaningfully more sensitive than the CX00 (excepting Canon's new $20k FF camera).

It's worth boosting the noise reduction setting in-camera. The codec really starts to fail when you underexpose noisy images at high ISOs so in-camera is better than noise reduction unless you have an external recorder, plus who has the budget to spend all that time rendering. The C100 seem usable to about 10,000 ISO–really, really good.
 
Policar said:
... The C100 seem usable to about 10,000 ISO–really, really good.
Click to expand...

Yes and no, I think. People often suggest what a usable ISO on the C100 is as if it is an absolute. The picture profile in use, the gamma, the noise reduction setting and the white balance setting all have a big impact. In my experience, 10,000 ISO on a cine1 gamma, with noise reduction at 3 on a daylight white balance will look substantially better than the same thing using Wide DR and tungsten white balance.
 
drummondb said:
Yes and no, I think. People often suggest what a usable ISO on the C100 is as if it is an absolute. The picture profile in use, the gamma, the noise reduction setting and the white balance setting all have a big impact. In my experience, 10,000 ISO on a cine1 gamma, with noise reduction at 3 on a daylight white balance will look substantially better than the same thing using Wide DR and tungsten white balance.
Click to expand...

WideDR is rather noisy.
 
You must log in or register to reply here.
Back
Top