Also, setting to a decibel 'value' also isn't entirely accurate and can leave a larger range for operational limits. Example, there might actually be a variation of +/-300 RPM where the measurable difference of a single decibel is seen, but 300 RPM is actually a significant difference in the amount of airflow being moved. 1200 RPM and 1500 RPM actually end up performing quite differently. It becomes difficult to replicate test runs without variables based on decibel levels alone as the same measurement of 35dB for one test can easily fall outside of acceptable deviation. In setting fan speeds, this is much simpler as we can define the PWM control of the fan to be exactly 100% or exactly 50%, which is a precise measurement of the fan RPM with very minimal average fluctuation over time. This is great in concept, except that the reality is that not every cooler manages to reach the same dB level at peak operation or lowest noise levels as every other cooler. Some coolers are not seen to operate at 35 or 40dB while others easily exceed it. These same coolers in comparison also