Bit Depth

Bit depth (also called pixel depth or color depth) specifies the number of bits allocated to each pixel in a digital image. For a grayscale image with N bits, there are 2^N possible intensity levels. The theoretical maximum signal-to-noise ratio is SNR = 6.02N + 1.76 dB. In practice, camera noise limits the Effective Number of Bits (ENOB) to less than the nominal ADC bit depth.

Scientific cameras commonly use 12-bit (4,096 levels) or 16-bit (65,536 levels) ADCs. The choice should match the camera's dynamic range: if the sensor can distinguish 10,000 intensity levels (based on well depth / read noise), a 14-bit ADC (16,384 levels) captures the full range without bottlenecking.

Insufficient bit depth causes quantization artifacts — visible as posterization (banding) in images and staircase effects in intensity measurements. More critically, low bit depth can make it impossible to distinguish closely spaced intensity populations, directly limiting segmentation accuracy and biomarker classification. For quantitative tissue analysis, 12-bit or higher is essential.

When imaging DAPI-stained nuclei alongside a weakly expressed biomarker, the DAPI signal may be 100× brighter than the biomarker. At 8-bit (256 levels), the weak signal occupies only 2-3 intensity levels — too few for reliable thresholding. At 16-bit, the same signal spans 200+ levels, enabling precise quantification and clean segmentation.