Can converting a RAW file to black and white reduce image noise?
Asked 11/21/2018
1 views
2 answers
0
I understand that low-light images are noisy because there are fewer photons, so the signal is weaker. That made me wonder whether a RAW converter could deliberately trade color information for lower noise—perhaps even before or during demosaicing—by using neighboring red/green/blue-filtered pixels to smooth away errors. In practice, does converting to black and white reduce noise, and do noise-reduction algorithms already work this way?
Originally by user67208. Source · Licensed CC BY-SA 4.0
user67208
7y ago
2 Answers
2
Is there a way to use B&W conversion to reduce noise?
That depends on what you mean by 'noise'.
The conversion to B&W will effectively eliminate all chrominance noise.
It won't do much for luminance noise.
You must keep in mind that even though the values reported by each photosite (a/k/a pixel well, sensel, etc.) on a digital sensor are monochromatic, they're all filtered by one of three differently colored filters. If most "green" filtered pixels have a lower luminance value than adjacent "blue" filtered pixels, it most likely means that the light falling in that area has more "blue" than "green" in it. Noise reduction algorithms are more likely to interpret "green" filtered pixels that are brighter than other nearby "green" filtered pixels as noise.
The only real way to do what you suggest is to eliminate the Bayer mask altogether so that each photosite can be purely monochrome when the light is collected. There are a few monochrome digital cameras available that do just that.
Originally by user15871. Source · Licensed CC BY-SA 4.0
user15871
7y ago
0
Generated from our catalog & community — verify before relying on it.
Yes, but only partly. Converting an image to black and white can reduce or eliminate chroma (color) noise, because color differences are no longer being shown. It does not do much for luminance noise, which appears as grain in brightness.
The idea of using neighboring pixels and sacrificing color detail is broadly how many noise-reduction methods already work. They analyze surrounding pixels and try to distinguish real detail from random variation, sometimes with edge-aware processing to avoid smearing detail. More advanced processing may happen around demosaicing as well.
However, you can’t simply assume a blue-filtered pixel is wrong because nearby green pixels are darker; that may reflect the actual color of the scene, not noise. Each sensel records monochrome intensity through a colored filter, so differences between red, green, and blue sites often contain real image information.
Also, not all noise is just photon-count variation; electronic read noise and amplification contribute significantly, especially at higher ISO.
So: black-and-white conversion can help hide color noise, but it won’t magically remove all noise. Dedicated RAW noise reduction is still the main tool.
Recommended products
UniqueBot
AI7y ago
Your Answer
Related Questions
Is there any image-quality benefit to converting RAW directly to black and white instead of demosaicing first?
Does converting a camera RAW file to DNG perform irreversible demosaicing?
Is there a best demosaicing algorithm for converting Bayer DNG files to black and white?
Can noise reduction be applied to RAW data before demosaicing?
How can I avoid sky banding/posterization when denoising RAW files in Darktable?