You can start from the assumption that the natural C14 to C12 ratio is a constant. And from there you can measure the actual ratio in the present time for a sample. Using math you can estimate how long ago the material was part of a living creature, since in the meantime the ratio would not have kept pace with the natural environmental ratio and instead would change depending on the half-life of C14. However, the assumption of a constant natural abundance of C14 is incorrect, there are lots of complex effects (like solar activity) which impact the ratio, which in turn affects the estimated age of an object. The more data you have on ancient C14 ratios the better able you are to calibrate radiocarbon dating.
In the past ancient trees and lake/seafloor samples have been used because they have many individual samples from separate years, each of which can be independently dated (separate from radiocarbon analysis).