[OT] rather advanced video editing question

[Petrus de Calguarium]

I have 2 versions of the same video from you tube.

Version 1: in 2 separate segments, great video, bad audio with long silences
Version 2: in 1 long file, poor grainy video, great flawless audio

Concatenated, if that is possible, both versions would have the same length.

Question: How can I join the 2 files from Version 1, erase the audio and 
supplement it with the audio from version 2 (using programs available in fedora 
& rpmfusion repos, of course)?