This is a visual from an experiment we started aiming in evaluating visual improvement objectively using what we refer to in image analysis as structure similarity metrics. We will talk about those metrics in a future post. Here we simulated very low visibility and measured water turbidity in nephelometric units (NTU). We made the water more turbid than at a coastal port in the Gulf of Mexico. For instance, at Galveston, we average turbidity at 15 NTU. Here we simulated shallow lake conditions. We start with clear water images for comparison. We did not use light other than daylight. The right side of our tank shades the water.