audiowmark

= audiowmark - Audio Watermarking

== Description

`audiowmark` is an Open Source (GPL) solution for audio watermarking.

A sound file is read by the software, and a 128-bit message is stored in a
watermark in the output sound file. For human listeners, the files typically
sound the same.

However, the 128-bit message can be retrieved from the output sound file. Our
tests show, that even if the file is converted to mp3 or ogg (with bitrate 128
kbit/s or higher), the watermark usually can be retrieved without problems. The
process of retrieving the message does not need the original audio file (blind
decoding).

Internally, audiowmark is using the patchwork algorithm to hide the data in the
spectrum of the audio file. The signal is split into 1024 sample frames. For
each frame, some pseoudo-randomly selected amplitudes of the frequency bands of
a 1024-value FFTs are increased or decreased slightly, which can be detected
later. The algorithm used here is inspired by

Martin Steinebach: Digitale Wasserzeichen für Audiodaten.
Darmstadt University of Technology 2004, ISBN 3-8322-2507-2

== Open Source License

`audiowmark` is *open source* software available under the *GPLv3
or later* license.

Copyright (C) 2018-2020 Stefan Westerfeld

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program. If not, see .

== Adding a Watermark

To add a watermark to the soundfile in.wav with a 128-bit message (which is
specified as hex-string):

[subs=+quotes]
....
*$ audiowmark add in.wav out.wav 0123456789abcdef0011223344556677*
Input: in.wav
Output: out.wav
Message: 0123456789abcdef0011223344556677
Strength: 10

Time: 3:59
Sample Rate: 48000
Channels: 2
Data Blocks: 4
....

The most important options for adding a watermark are:

--key ::
Use watermarking key from file (see <>).

--strength ::
Set the watermarking strength (see <>).

== Retrieving a Watermark

To get the 128-bit message from the watermarked file, use:

[subs=+quotes]
....
*$ audiowmark get out.wav*
pattern 0:05 0123456789abcdef0011223344556677 1.324 0.059 A
pattern 0:57 0123456789abcdef0011223344556677 1.413 0.112 B
pattern 0:57 0123456789abcdef0011223344556677 1.368 0.086 AB
pattern 1:49 0123456789abcdef0011223344556677 1.302 0.098 A
pattern 2:40 0123456789abcdef0011223344556677 1.361 0.093 B
pattern 2:40 0123456789abcdef0011223344556677 1.331 0.096 AB
pattern all 0123456789abcdef0011223344556677 1.350 0.054
....
The output of `audiowmark get` is designed to be machine readable. Each line
that starts with `pattern` contains one decoded message. The fields are
seperated by one or more space characters. The first field is a *timestamp*
indicating the position of the data block. The second field is the *decoded
message*. For most purposes this is all you need to know.

The software was designed under the assumption that you - the user - will be
able to decide whether a message is correct or not. To do this, on watermarking
song files, you could list each message you embedded in a database. During
retrieval, you should look up each pattern `audiowmark get` outputs in the
database. If the message is not found, then you should assume that a decoding
error occurred. In our example each pattern was decoded correctly, because
the watermark was not damaged at all, but if you for instance use lossy
compression (with a low bitrate), it may happen that only some of the decoded
patterns are correct. Or none, if the watermark was damaged too much.

The third field is the *sync score* (higher is better). The synchronization
algorithm tries to find valid data blocks in the audio file, that become
candidates for decoding.

The fourth field is the *decoding error* (lower is better). During message
decoding, we use convolutional codes for error correction, to make the
watermarking more robust.

The fifth field is the *block type*. There are two types of data blocks,
A blocks and B blocks. A single data block can be decoded alone, as it
contains a complete message. However, if during watermark detection an
A block followed by a B block was found, these two can be decoded
together (then this field will be AB), resulting in even higher error
correction capacity than one block alone would have.

To improve the error correction capacity even further, the `all` pattern
combines all data blocks that are available. The combined decoded
message will often be the most reliable result (meaning that even if all
other patterns were incorrect, this could still be right).

The most important options for getting a watermark are:

--key ::
Use watermarking key from file (see <>).

--strength ::
Set the watermarking strength (see <>).

[[key]]
== Watermark Key

Since the software is Open Source, a watermarking key should be used to ensure
that the message bits cannot be retrieved by somebody else (which would also
allow removing the watermark without loss of quality). The watermark key
controls all pseudo-random parameters of the algorithm. This means that
it determines which frequency bands are increased or decreased to store a
0 bit or a 1 bit. Without the key, it is impossible to decode the message
bits from the audio file alone.

Our watermarking key is a 128-bit AES key. A key can be generated using

audiowmark gen-key test.key

and can be used for the add/get commands as follows:

audiowmark add --key test.key in.wav out.wav 0123456789abcdef0011223344556677
audiowmark get --key test.key out.wav

[[strength]]
== Watermark Strength

The watermark strength parameter affects how much the watermarking algorithm
modifies the input signal. A stronger watermark is more audible, but also more
robust against modifications. The default strength is 10. A watermark with that
strength is recoverable after mp3/ogg encoding with 128kbit/s or higher. In our
informal listening tests, this setting also has a very good subjective quality.

A higher strength (for instance 15) would be helpful for instance if robustness
against multiple conversions or conversions to low bit rates (i.e. 64kbit/s) is
desired.

A lower strength (for instance 6) makes the watermark less audible, but also
less robust. Strengths below 5 are not recommended. To set the strength, the
same value has to be passed during both, generation and retrieving the
watermark. Fractional strengths (like 7.5) are possible.

audiowmark add --strength 15 in.wav out.wav 0123456789abcdef0011223344556677
audiowmark get --strength 15 out.wav

== Short Payload (experimental)

By default, the watermark will store a 128-bit message. In this mode, we
recommend using a 128bit hash (or HMAC) as payload. No error checking is
performed, the user needs to test patterns that the watermarker decodes to
ensure that they really are one of the expected patterns, not a decoding
error.

As an alternative, an experimental short payload option is available, for very
short payloads (12, 16 or 20 bits). It is enabled using the `--short `
command line option, for instance for 16 bits:

audiowmark add --short 16 in.wav out.wav abcd
audiowmark get --short 16 out.wav

Internally, a larger set of bits is sent to ensure that decoded short patterns
are really valid, so in this mode, error checking is performed after decoding,
and only valid patterns are reported.

Besides error checking, the advantage of a short payload is that fewer bits
need to be sent, so decoding will more likely to be successful on shorter
clips.

== Video Files

For video files, `videowmark` can be used to add a watermark to the audio track
of video files. To add a watermark, use

[subs=+quotes]
....
*$ videowmark add in.avi out.avi 0123456789abcdef0011223344556677*
Audio Codec: -c:a mp3 -ab 128000
Input: in.avi
Output: out.avi
Message: 0123456789abcdef0011223344556677
Strength: 10

Time: 3:53
Sample Rate: 44100
Channels: 2
Data Blocks: 4
....

To detect a watermark, use

[subs=+quotes]
....
*$ videowmark get out.avi*
pattern 0:05 0123456789abcdef0011223344556677 1.294 0.142 A
pattern 0:57 0123456789abcdef0011223344556677 1.191 0.144 B
pattern 0:57 0123456789abcdef0011223344556677 1.242 0.145 AB
pattern 1:49 0123456789abcdef0011223344556677 1.215 0.120 A
pattern 2:40 0123456789abcdef0011223344556677 1.079 0.128 B
pattern 2:40 0123456789abcdef0011223344556677 1.147 0.126 AB
pattern all 0123456789abcdef0011223344556677 1.195 0.104
....

The key and strength can be set using the command line options

--key ::
Use watermarking key from file (see <>).

--strength ::
Set the watermarking strength (see <>).

== Output as Stream

Usually, an input file is read, watermarked and an output file is written.
This means that it takes some time before the watermarked file can be used.

An alternative is to output the watermarked file as stream to stdout. One use
case is sending the watermarked file to a user via network while the
watermarker is still working on the rest of the file. Here is an example how to
watermark a wav file to stdout:

audiowmark add in.wav - 0123456789abcdef0011223344556677 | play -

In this case the file in.wav is read, watermarked, and the output is sent
to stdout. The "play -" can start playing the watermarked stream while the
rest of the file is being watermarked.

If - is used as output, the output is a valid .wav file, so the programs
running after `audiowmark` will be able to determine sample rate, number of
channels, bit depth, encoding and so on from the wav header.

Note that all input formats supported by audiowmark can be used in this way,
for instance flac/mp3:

audiowmark add in.flac - 0123456789abcdef0011223344556677 | play -
audiowmark add in.mp3 - 0123456789abcdef0011223344556677 | play -

== Input from Stream

Similar to the output, the `audiowmark` input can be a stream. In this case,
the input must be a valid .wav file. The watermarker will be able to
start watermarking the input stream before all data is available. An
example would be:

cat in.wav | audiowmark add - out.wav 0123456789abcdef0011223344556677

It is possible to do both, input from stream and output as stream.

cat in.wav | audiowmark add - - 0123456789abcdef0011223344556677 | play -

Streaming input is also supported for watermark detection.

cat in.wav | audiowmark get -

== Raw Streams

So far, all streams described here are essentially wav streams, which means
that the wav header allows `audiowmark` to determine sample rate, number of
channels, bit depth, encoding and so forth from the stream itself, and the a
wav header is written for the program after `audiowmark`, so that this can
figure out the parameters of the stream.

There are two cases where this is problematic. The first case is if the full
length of the stream is not known at the time processing starts. Then a wav
header cannot be used, as the wav file contains the length of the stream. The
second case is that the program before or after `audiowmark` doesn't support wav
headers.

For these two cases, raw streams are available. The idea is to set all
information that is needed like sample rate, number of channels,... manually.
Then, headerless data can be processed from stdin and/or sent to stdout.

--input-format raw::
--output-format raw::
--format raw::

These can be used to set the input format or output format to raw. The
last version sets both, input and output format to raw.

--raw-rate ::

This should be used to set the sample rate. The input sample rate and
the output sample rate will always be the same (no resampling is
done by the watermarker). There is no default for the sampling rate,
so this parameter must always be specified for raw streams.

--raw-input-bits ::
--raw-output-bits ::
--raw-bits ::

The options can be used to set the input number of bits, the output number
of bits or both. The number of bits can either be `16` or `24`. The default
number of bits is `16`.

--raw-input-endian ::
--raw-output-endian ::
--raw-endian ::

These options can be used to set the input/output endianness or both.
The parameter can either be `little` or `big`. The default
endianness is `little`.

--raw-input-encoding ::
--raw-output-encoding ::
--raw-encoding ::

These options can be used to set the input/output encoding or both.
The parameter can either be `signed` or `unsigned`. The
default encoding is `signed`.

--raw-channels ::

This can be used to set the number of channels. Note that the number
of input channels and the number of output channels must always be the
same. The watermarker has been designed and tested for stereo files,
so the number of channels should really be `2`. This is also the
default.

== Dependencies

If you compile from source, `audiowmark` needs the following libraries:

* libfftw3
* libsndfile
* libgcrypt
* libzita-resampler
* libmpg123

== Building fftw

`audiowmark` needs the single prevision variant of fftw3.

If you are building fftw3 from source, use the `--enable-float`
configure parameter to build it, e.g.::

cd ${FFTW3_SOURCE}
./configure --enable-float --enable-sse && \
make && \
sudo make install

or, when building from git

cd ${FFTW3_GIT}
./bootstrap.sh --enable-shared --enable-sse --enable-float && \
make && \
sudo make install

== Docker Build

You should be able to execute `audiowmark` via Docker.
Example that outputs the usage message:

docker build -t audiowmark .
docker run -v :/data -it audiowmark -h