propogation in a gas such as air under normal conditions goes as:
c = sqrt(y P0/p0)
where c is the velocity of propogation, y is ratio of specific heats of the gas (for air, which is essentially a diatomic gas, y = 1.402, and is largely independent of
temperature over the range of such where we'd want to do our listening), P0 is the constant equilibrium pressure of the gas, which at 0C is 1.103*10^5 Pascals, and p0 is the constant equilibrium density of the gas, which at 0C is 1.293 kg/m^3. This leads to a velocity of sound at 0C of 331.6 m/s.
If we then explore the temperature dependency, as it effects P0 and p0 (since we find that y is independent of temperature), the result is that the velocity goes as the square root of the absolute temperature, and this first expression reduces
to a temperature dependent form:
c = sqrt(y r Tk)
To quote from Kinsler:
"For most gases at constant temperature, the ratio of P0/p0 is nearly independent of pressure: a doubling of pressure is accompanied by a doubling of density of
the gas, so that the speed of sound does not change with variation of density"
As can be found in Kinsler, Frey, Beranek and others, one that over moderate distances and the audible frequency range (say, less than 10 meters and at frequencies below 25 kHz), large changes in relative humidity have no significant
effect on the propogation of sound.
One can find, without much effort, other means of determining the dependency on the "characteristics or air" temperature, pressure and humidity to an equal degree of scientific rigor.
The conclusion is fairly straightforward:
Over the range of temperatures, pressures and humidities one is likely to encounter in a home listening situation, with the exception of the propogation velocity, ambient conditions have no significant effect on the propogation
characteristics of sound.
Even is one then considers the propogation velocity, it goes as the square root of absolute temperature. Consider a range of 15C (65F or 288K) to 35C (95C or 308K), the difference in propogation velocity is sqrt(308/288) or about 3%.
Consider the context of what variations are likely to be found: in addition to the small variation due to temperature, pressure does NOT vary over a wide range
except in extraordinary circumstances (usually referred to as "hurricanes"), circumstances under which the critical acoustics properties of air are rather unimportant.
Thus, it would seem that the ACOUSTIC properties of air one might encounter is unlikely to cause any significant difference in sound.
