Adding noise overlay to a recording

[LangDoc]

I use Audacity 1.3.9 to process .wav recordings (14.1 kHz/16 bit) of conversation in an endangered indigenous language. The recordings will be archived. As part of preparing these recordings for archiving, I sometimes need to obscure short sections of the recording for reasons of sensitivity or privacy. Is it possible to use Audacity to add a white noise overlay to the sections I wish to obscure? (And ideally, is there a way to do this so that the noise-overlay is reversible 50 years hence, allowing obscured sections to be recoverable in cases where such sensitivity eventually lapses?) Thank you!

[kozikowski]

That's a good one. We can probably think of about a hundred different ways to obscure the track, but none that let you unobscure it later. Koz

[Robert J. H.]

I would probably archive both versions, the uncensored and the censored one just with other security levels.
The problem of the obscuring is, how will a person in 50 years know what procedure is required to recover the signal?
You could for instance modulate the signal with a certain carrier frequency. The exact value can be stored, along with the start/end time. Normal playback would then either be silent or audible as a sine tone.
If you prefer noise, the noise must be stored too, in order to recover the signal. However, the noise level has to be very high to cover the signal.
Test it yourself, generate white noise and listen, you'll probably can still make out the underlying speech.

[Trebor]

Encrypt the uncensored copies , TrueCrypt is free mature encryption software ... http://en.wikipedia.org/wiki/TrueCrypt
which is good enough to baffle the FBI ... http://www.theregister.co.uk/2010/06/28 ... _lock_out/