Panasonic FZ28 Images of Noise - Part 2

I was looking at the last post I did on this subject and thinking “What the heck? Why did I post the inverted images?” But there’s a good reason for that, the eye sees slight variations of white a lot more keenly than slight variations of black. Still, I thought it best to show the noise non-inverted, but this time instead of adjusting brightness and contrast, I went straight to Levels and clipped everything above a value of 7 on a scale from 0 to 255. The “black” images have some values up to 16, but they are very few and there’s a clear cutoff at 7, so it seems like a good place to clip.

FZ28 noise, clipped to peak of 7, reduced to 12.5%

FZ28 noise, clipped to peak of 7, cropped to 12.5%

And then I did the gif movies too. I was surprised at the blotchiness of the noise. It isn’t just random bits over the whole sensor, it’s random bits and random ranges, big blotches of dozens of pixels with near the same value.

Aaaagghh! Looking at the cropped gif, if I stare long enough it looks like the noise is rotating counter clockwise, but then I can make it reverse rotation, or make it go up or down or sideways. Weird.

FZ28 noise, clipped to peak of 7, reduced to 12.5%

FZ28 noise, clipped to peak of 7, cropped to 12.5%

You can see in all of these the blue and purple cast to the noise. That’s because green has the least noise and blue the most. Most of the noise has a value of 1, 2, or 3. In the red it’s mostly 1 and 2; in the green it’s mostly 1, but actually green was mostly 0 which is not noise; in the blue there’s hardly any black, there are a lot of values at 2, 3, 4, and 5. I’ll show the channels separated in my next post.