Some months ago I wanted to create a new (double) sonic boom wave file for my Orbiter installation, however, I was getting close to pulling my hair out due to the fact that in seemingly every of the STS Shuttle landing videos on youtube which were recorded from Nasa TV had the narrator either speak while the boom could be heard or shortly thereafter while the boom was still echoing. Don't know which missions they were exactly, but some had a really fantastic clear sound of the boom and the reverberant sound, if it just wasn't for the narrator ruining everything... Aaaargh.
You may be interested in ARSS (Analysis and Resynthesis Sound Spectrograph). It converts a *.wav file into a *.bmp, with time on the x axis, frequency on the y axis, and intensity indicated on a grayscale. Speech has a fairly distinct look on a spectrogram, and the boom is likely to be *very* distinct (though I can only say from intuition, not from having worked with sound files containing sonic booms), so you should be able to convert your sound file, edit the speech out of the bmp, convert back, and be left with a fairly pure boom.
And looking at the site, apparently the author has introduced a GUI interface for it, Photosounder, that I haven't tried out yet. I will have to do so.
http://arss.sourceforge.net/
EDIT: Oh, never mind about photosounder. It's not freeware, whereas ARSS is, though a 25 Euro non-commercial license is available.