Re-added tutorial changes that mysteriously dissapeared...
This commit is contained in:
@@ -17,7 +17,7 @@ To view all available options simply run:
|
||||
|
||||
A list of all commands available is then presented:
|
||||
|
||||
.. code:: bash
|
||||
::
|
||||
|
||||
usage: concatenator [-h] [--src-db] [--tar_db]
|
||||
[--analyse [ANALYSE [ANALYSE ...]]] [--analysis_dict]
|
||||
@@ -30,31 +30,12 @@ A list of all commands available is then presented:
|
||||
Concatenator is a tool for synthesizing interpretations of a sound, through
|
||||
the analysis and synthesis of audio grains from a corpus database. The program
|
||||
works by analysing overlapping segments of audio (known as grains) from both
|
||||
the target sound and the source database, then searching for the closest
|
||||
matching grain in the source database to the target sound. Finally, the output
|
||||
is generated by overlap-adding the best matches.
|
||||
|
||||
positional arguments:
|
||||
source Directory of source files/database to take grains from
|
||||
when synthesizing output
|
||||
target Directory of target files/database to match source
|
||||
grains to.
|
||||
output Directory to use as database for outputing results and
|
||||
match information. Output audio will be stored in the
|
||||
/audio sub-directory and match data will be stored in
|
||||
the /data directory.
|
||||
|
||||
optional arguments:
|
||||
-h, --help show this help message and exit
|
||||
--src-db Specifies the directory to create the source database
|
||||
and store analyses in. If not specified then the
|
||||
|
||||
...
|
||||
|
||||
For this demonstration, the following file structure will be used:
|
||||
|
||||
.. code:: bash
|
||||
|
||||
::
|
||||
/Users/samuelperry/concatenator_test/
|
||||
|-- source_db
|
||||
| |-- Trumpet.novib.ff.A3.stereo.aif
|
||||
@@ -90,18 +71,16 @@ explained in the :ref:`overview` section. Output is stored in the audio
|
||||
directory of the output database that has been created.
|
||||
This produces this directory structure:
|
||||
|
||||
.. code:: bash
|
||||
::
|
||||
|
||||
/Users/samuelperry/concatenator_test/
|
||||
|-- analysed_source_db
|
||||
| |-- audio
|
||||
| | |-- Trumpet.novib.ff.A3.stereo.aif -> (Symlink)
|
||||
| | |-- Trumpet.novib.ff.A4.stereo.aif -> (Symlink)
|
||||
| | |-- Trumpet.novib.ff.A5.stereo.aif -> (Symlink)
|
||||
|
||||
...
|
||||
|
||||
| | |-- Trumpet.novib.ff.F5.stereo.aif -> (Symlink)
|
||||
| | |-- Trumpet.novib.ff.G3.stereo.aif -> (Symlink)
|
||||
| | `-- Trumpet.novib.ff.G4.stereo.aif -> (Symlink)
|
||||
| `-- data
|
||||
@@ -125,8 +104,6 @@ This produces this directory structure:
|
||||
|-- source_db
|
||||
| |-- Trumpet.novib.ff.A3.stereo.aif
|
||||
| |-- Trumpet.novib.ff.A4.stereo.aif
|
||||
| |-- Trumpet.novib.ff.A5.stereo.aif
|
||||
| |-- Trumpet.novib.ff.F5.stereo.aif
|
||||
|
||||
...
|
||||
|
||||
@@ -153,6 +130,27 @@ creation of partable databases that can moved to other machines without
|
||||
breaking links to the original files. (Any pre-existing symbolic links will be
|
||||
overwritten with hard copies when using this option.)
|
||||
|
||||
Configuration Flags
|
||||
-------------------
|
||||
For quick modification of analysis parameters, parameter flags can be specified
|
||||
directly when calling the script. For example:
|
||||
|
||||
.. code:: bash
|
||||
|
||||
concatenator ./source_db ./target_db ./output_db --src_db \
|
||||
./analysed_source_db --tar_db ./analysed_tar_db --reanalyse --fft \
|
||||
'--window_size 2048'
|
||||
|
||||
This overwrites the value specified for window_size in the config file with the
|
||||
value provided.
|
||||
|
||||
When databases have already been created, previous data is used when re-running
|
||||
the script over them. This allows for different databases to be used without
|
||||
continuous reanalysis. However, if analysis or matching parameters are changed,
|
||||
the "--reanalyse" and "--rematch" flags can be used to force the overwriting of
|
||||
old data, using the new parameters.
|
||||
|
||||
|
||||
config.py
|
||||
---------
|
||||
The config.py file is used for specifying all user defined options and can be
|
||||
@@ -282,23 +280,3 @@ each parameter. The default config.py file looks like this:
|
||||
"channels": 1
|
||||
}
|
||||
|
||||
Configuration Flags
|
||||
-------------------
|
||||
For quick modification of analysis parameters, parameter flags can be specified
|
||||
directly when calling the script. For example:
|
||||
|
||||
.. code:: bash
|
||||
|
||||
concatenator ./source_db ./target_db ./output_db --src_db \
|
||||
./analysed_source_db --tar_db ./analysed_tar_db --reanalyse --fft \
|
||||
'--window_size 2048'
|
||||
|
||||
This overwrites the value specified for window_size in the config file with the
|
||||
value provided.
|
||||
|
||||
When databases have already been created, previous data is used when re-running
|
||||
the script over them. This allows for different databases to be used without
|
||||
continuous reanalysis. However, if analysis or matching parameters are changed,
|
||||
the "--reanalyse" and "--rematch" flags can be used to force the overwriting of
|
||||
old data, using the new parameters.
|
||||
|
||||
|
||||
Reference in New Issue
Block a user