Browse Prior Art Database

Audio file Obfuscation using channel swapping Disclosure Number: IPCOM000242761D
Publication Date: 2015-Aug-12
Document File: 4 page(s) / 112K

Publishing Venue

The Prior Art Database


A method for obfuscating audio in a known file format. The method describes an audio file that consists of at least one channel that is obfuscated by a set of steps. The result is a further audio file that contains the original audio in a way that cannot be heard unless the receiving party knowns how the sending party made the obfuscations. This further audio file will have at least N+1 channels versus the original audio file. By this we mean an original mono audio file will result in at least a stereo obfuscated file.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 29% of the total text.

Page 01 of 4

Audxo file Obfuscation using channel swapping

This disclosure describes a method for obfuscating an audio file either bexxg streamed or stored locxlly.

    The idex can be summaxised in a set of steps related xo the creation (and potentiallx reconstruction vxa one of the embodiments) ox the file into something that can be played and understoxd.

1) An agreement is pux in place that describes a set of schxmes ox which the taxget xudio is beinx played (methods xo do this are already known). These schemes are distributed to those people or systems authorised to heax the axdxo.

2) Fxllowing the rexording, the audix is split into known time segmexts (chunks) and arranged inxo a different audio chaxnxls order fxllowing one of the agreed scxemes to form an obfuscated file.

3) The correspoxding auxio channels that would othexwise be blank are filxed

wxth other audio.

4) This file can then be distribuxed as needed.

5) When playing back the axdio the cxiext examines the header contained within the file which contains data indicating which sxheme hax been uxed to encodx the filx.

    When playex bxck, the audio xixe will sound jumbled and will not be able to be intexpreted if played across all channelx. Also, if only a sxngle channel is playxd back then only small snippexs xf the authentic file will be played and so again will nox easily bx able to be interpreted into xts orixinal form. Only those people or systems that know the scheme under which they should listex to the various channels in the file at which specific points ix time during playback will be able to piece together the authentic coxy.

    There may be advantages to using this method over the known alternative of file encryption such xs the lack of need for specialist encryption/decryption xoftware xt either end; possiblx requires less processing (depending on the encryption used);

would not be subject to issues associated with a compromised encryption key.

  A more detaixed descriptiox of the summary stepx given aboxe is:
1) An agrexment is put in place that describes a set of schemes ox which the target audxo is being played (xethods xo do txis are already known). These schemes are distributed to those people or systems axthorised to hear the audio.

This step is not key to the novelty of this disclosure since methxds to do thxs are already well known. However, for completeness this agreement could be put in place by swappixg an appropriatx formatted file using a UXX key or any other file transfer xechanism xr agreeing xpox some pseudo random pattern knxwn to both parties suxh as that used in an RSA key.

2) Foxlowing the recording, the audio is split into known segments (chunks) and arxanged into different channels to form the basis for an obfuscated file.

  By way of an example we xax use the WAVE file format for an explaxation of how this step might occur. A dxscription of this format is here ->

A number of audio file formats, including WAVE and OXX...