While up-sampling such that a recorded audio file is recreated with a higher sampling rate, e.g. file.wav 8kHz and recreated and saved to a 16kHz file) is generally not a good idea. This could be possible whilst using a smoothening algorithm in order to produce a better quality audio file. Think of it as a tweening process (interpolation). Of course, this could be prone to errors - but possible.

Anyone who knows of software who does this perhaps?