You can take turbidity measurements of liquids using turbidity sensors such as this (which essentially shine a light through a known amount of the liquid at a light-sensor and measure how much light gets through.)
You can also take turbidity measurements using your eyeballs with a tool like a Secchi disk. With the Secchi disk you drop the disk into your liquid and measure how deep it goes before you can't see it.
Secchi disk measurements will presumably vary based on the amount of ambient light in the area where you are taking your measurement, as well as based on the eyesight of the measurer, so I guess there probably can't be a precise "Secchi depth to NTU" conversion.
But I don't know how significant this variability is -- does anyone know if there is an imprecise but "close-enough" conversion factor, or any other good way to think about comparing data collected through these two different techniques?