Hi Esa-Matti--
The firmware uses the standard atmospheric model. The following code from the FlySight Viewer might give a better idea of how this is calculated:
We start with an altitude above sea level, y, and first calculate air pressure:
const double airPressure = SL_PRESSURE * pow(1 - LAPSE_RATE * y / SL_TEMP, A_GRAVITY * MM_AIR / GAS_CONST / LAPSE_RATE);
Next we find the temperature at altitude using the standard lapse rate:
const double temperature = SL_TEMP - LAPSE_RATE * y;
And then we can calculate the air density at altitude:
const double airDensity = airPressure / (GAS_CONST / MM_AIR) / temperature;
Finally, we can use the first equation for equivalent airspeed here (relating the correction factor to air density at altitude):
When I originally wrote the speed correction code in the FlySight firmware, I used a source which skipped this derivation and simply gave the final result. However, I can't find that source now. But I've just checked these calculations against the table in the FlySight firmware and they seem to give very nearly the same result (at 11,450 m the original table gives a correction factor of 1024/1944, while the new calculation gives 1024/1945).
Michael