A black-and-white digital file can be produced two ways: by capturing color through a filtered sensor and then discarding the color information, or by capturing luminance directly on a sensor that was never filtered for color in the first place. The two paths look similar in the final JPEG, but they differ in how much light reaches the silicon and how much spatial detail survives. Understanding why a dedicated monochrome sensor outresolves and outpaces a converted color file requires looking at the color filter array that sits between the lens and the photosites.

The Bayer Array and Its Costs

Almost every color sensor records color through a color filter array (CFA) patented by Bryce Bayer at Eastman Kodak in 1976. The Bayer pattern is a repeating two-by-two tile of one red, one blue, and two green filters, so green occupies half of all photosites because the human eye derives most of its luminance perception from green wavelengths. Each photosite therefore measures only one of the three primaries; the missing two values are estimated from neighboring sites in a step called demosaicing.

Two penalties follow from this design. First, every filter is subtractive: a red filter passes red and absorbs green and blue, so each photosite collects only a fraction of the incident light. Second, no photosite samples the full luminance signal independently. Demosaicing reconstructs a full-color image at every pixel by interpolation, which means fine luminance detail is partly inferred rather than measured.

Why a Filterless Sensor Resolves More

Removing the CFA makes every photosite sensitive to the entire visible spectrum, so each one records a true luminance value. No interpolation is required: one photosite maps to one pixel. Because demosaicing always blends information across neighbors, even the best Bayer interpolation cannot recover the spatial detail of a sensor that measures luminance at every site directly.

Leica quantified this difference for the M Monochrom, an 18-megapixel camera built on the M9 platform with the CFA omitted. The manufacturer states the camera delivers images roughly 100 percent sharper than monochrome images derived from a color sensor of comparable pixel count. The figure reflects the absence of interpolation: each photosite records a luminance value directly rather than having two of its three primaries reconstructed from neighbors.

The Sensitivity Gain

Light not absorbed by a filter reaches the photodiode and contributes to signal. With the CFA gone, all incoming light is collected, raising base sensitivity. The M Monochrom carries a native ISO of 320 against the color M9’s ISO 160 on the same sensor generation, a difference of approximately one stop. The higher signal also improves the signal-to-noise ratio at a given exposure, so monochrome files tend to show cleaner shadows and finer tonal separation than a desaturated color frame.

The trade is absolute: a filterless sensor records no color and cannot be converted back. Contrast control that a color workflow handles in software, such as darkening a blue sky, must instead be performed optically with physical colored filters over the lens, exactly as on black-and-white film. The gain in resolution and sensitivity is bought by surrendering all chrominance at the point of capture.