Why in RF receiver we need impedance matching for maximum power transfer but not maximum voltage? And why there is no Impedance matching or Power matching in audio engineering? Is it depends on the signal frequency to chose impedance matching or impedance bridging?
Audio in fact did used to be done as constant impedance (~600 ohms was typical) as a holdover from Bell Telecomm who developed most of the early tech for long distance phone lines.
Consider that something like an audio amp has an output impedance that is basically a current limited voltage source (near zero output impedance), given an electrically short cable with sufficient copper this condition will exist at the other end of the cable as well, and all is good. Now lets extend that cable into being a transcontinental long distance phone line... Now there is some audio frequency at which that cable is a 1/4 wavelength long, and that low Z source is transformed by the cable to high Z at that frequency (with a corresponding increase in voltage), thus our mismatched line has become a filter, probably not what was wanted. If instead the load matched the cable impedance then you get no filtering effect.
At audio frequencies the lines are (almost) always electrically short (as in less then maybe 1/10 wavelength long) so a low Z source will be seen as such at the other end of the cable irrespective of the cable characteristic impedance.
This is not the case at RF where the cable can easily be a significant fraction of a wavelength long and are often well over a wavelength long.
Note also that the audio band covers approximately 3 Decades of bandwidth, good luck designing a conjugate match for that, bridging is just so much easier for the 99% of cases where it will work!