On board color reflectance was measured for each core in the visible wavelength band between 400 and 700 nm in increments of 10 nm. Before measuring each core, the spectrophotometer was calibrated for the white color reflectance by attaching a white cap. The reflectance values of the white calibration cap were constant throughout the measurement period. The color data were corrected for "outliers" resulting from small voids. In addition, color data of the first and last measurements of each section were omitted when the resolution was 2 cm because of the proximity to the plastic cap at the ends of each section.
The analysis and interpretation proceeds from the percentage reflectance values as a function of wavelength (reflectance), the ratio between reflectance values of the red (650 nm) and blue (450 nm) wavelengths (r/b ratio) and the lightness expressed in percent of a white surface (total reflectance).
The downcore reflectance records of the dark-colored sediments from the Congo Basin display an overall small range in the total reflectance values and are dominated by high-frequency variability (see "Lithostratigraphy" sections, "Site 1075," "Site 1076," and "Site 1077" chapters, this volume). To consider the analytical variations, one section from Site 1077 was selected and measured four times over a 12-hr shift and was therefore exposed to changing temperature and light conditions (Fig. 1A). The comparison of the 4-run measurements shows that the analytical variation is about 15% of the total variability in the dark-colored section from Site 1077, which defines the reproducibility of the spectrophotometer measurements for this type of sediment (Fig. 1A). This suggests that even subtle color changes often not visible to the eye mostly reflect lithologic changes. A better reproducibility is obtained for light-colored sediments, as shown from a 3-run measurement of a section from Hole 1083A (Fig. 1B).