A camera sensor does not directly measure colors in XYZ or sRGB. Its RGB filters have their own spectral sensitivities, and in general they do not satisfy the Luther–Ives condition, so there is no perfect reversible mapping to human color matching functions.

In practice, the camera or RAW converter uses a camera profile to map the sensor’s native RGB values into a standard color space such as sRGB. A common approach is a color transform matrix fitted from measurements of known reference colors, often using targets like a ColorChecker/GretagMacbeth chart. The matrix is chosen to minimize color error, often in a perceptual space such as Lab.

So when a camera “uses sRGB,” it usually means it can convert its raw sensor data into an sRGB output image internally. RAW converters such as Lightroom or Darktable also do this using camera-specific profiles or simpler color matrices.

Manufacturers may tune these transforms not just for accuracy, but also for pleasing color rendering.