path: root/manual/xml/using_existing_audio.xml

<?xml version="1.0" encoding="UTF-8"?>

<!DOCTYPE chapter PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN"
"http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd">

<chapter id="ch-using-existing-audio">
  <title>Using Existing Audio</title>
  <para>
    There are two primary ways to bring data into Ardour: recording it
    within a session from a live sound source or importing pre-existing
    audio files. This section covers the various ways to import audio into a
    session.
  </para>

  <section id="importing-and-embedding">
    <title>Importing and Embedding</title>
    <para>
      Importing and embedding are two different methods of using existing
      audio files on your computer (or network file system) within a
      session. They differ in one key respect:
    </para>

    <variablelist>
      <varlistentry>
        <term>Importing</term>
        <listitem>
          <para>
            An existing audio file is copied to the session's sounds folder,
            and is converted into the session's native format (WAVE or
            Broadcast WAVE depending on your choice) and sample rate. At
            this time, no control over the conversion process is offered. If
            sample rate conversion is required, it will be carried out at
            the highest quality that Ardour can provide. This means that it
            can be rather slow (many minutes to import an audio file lasting
            a few minutes).
          </para>
        </listitem>
      </varlistentry>

      <varlistentry>
        <term>Embedding</term>
        <listitem>
          <para>
            An existing audio file is used as a the source for a region, but
            is not copied or modified in any way.
          </para>
        </listitem>
      </varlistentry>
    </variablelist>
  </section>

  <section id="supported-external-audio-file-formats">
    <title>Supported External Audio File Formats</title>
    <para>
      The list of audio file formats that Ardour can import/embed is quite
      long. It is based on the functionality offered by libsndfile, an
      excellent and widely used software library by Australian programmer
      Erik de Castro Lopo. As libsndfile's capabilities expand, so will
      Ardour's abilities to import (and export) new formats. Ogg/Vorbis (an
      excellent, unpatented and license free audio compression format
      similar to MP3) is planned for the near future. Currently, supported
      formats include:
    </para>

    <itemizedlist>
      <listitem>
        <para>
          Microsoft WAV
        </para>
      </listitem>

      <listitem>
        <para>
          SGI/Apple AIFF/AIFC
        </para>
      </listitem>

      <listitem>
        <para>
          Sun AU/Snd
        </para>
      </listitem>

      <listitem>
        <para>
          Raw (headerless)
        </para>
      </listitem>

      <listitem>
        <para>
          Paris Audio File (PAF)
        </para>
      </listitem>

      <listitem>
        <para>
          Commodore IFF/SVX
        </para>
      </listitem>

      <listitem>
        <para>
          Sphere/NIST WAV
        </para>
      </listitem>

      <listitem>
        <para>
          IRCAM SF
        </para>
      </listitem>

      <listitem>
        <para>
          Creative VOC
        </para>
      </listitem>

      <listitem>
        <para>
          SoundForge W64
        </para>
      </listitem>

      <listitem>
        <para>
          GNU Octave MAT4.4
        </para>
      </listitem>

      <listitem>
        <para>
          Portable Voice Format
        </para>
      </listitem>

      <listitem>
        <para>
          Fasttracker 2 XI
        </para>
      </listitem>

      <listitem>
        <para>
          HMM Tool Kit HTK
        </para>
      </listitem>
    </itemizedlist>

    <para>
      Sample encodings supported include:
    </para>

    <itemizedlist>
      <listitem>
        <para>
          Unsigned and signed 8, 16, 24 and 32 bit PCM
        </para>
      </listitem>

      <listitem>
        <para>
          IEEE 32 and 64 floating point
        </para>
      </listitem>

      <listitem>
        <para>
          U-LAW
        </para>
      </listitem>

      <listitem>
        <para>
          A-LAW
        </para>
      </listitem>

      <listitem>
        <para>
          IMA ADPCM
        </para>
      </listitem>

      <listitem>
        <para>
          MS ADPCM
        </para>
      </listitem>

      <listitem>
        <para>
          GSM 6.10
        </para>
      </listitem>

      <listitem>
        <para>
          G721/723 ADPCM
        </para>
      </listitem>

      <listitem>
        <para>
          12/16/24 bit DWVW
        </para>
      </listitem>

      <listitem>
        <para>
          OK Dialogic ADPCM
        </para>
      </listitem>

      <listitem>
        <para>
          8/16 DPCM
        </para>
      </listitem>
    </itemizedlist>
  </section>

  <section id="using-audio-files">
    <title>Using audio files as tracks or regions?</title>
    <para>
      When you want to use existing audio files in an Ardour session, the
      first choice you need to make is whether you want to bring the files
      in as tracks or as new regions. Consider the two following scenarios:
    </para>

    <itemizedlist>
      <listitem>
        <para>
          you have an 8 track recording of existing material, with 1 audio
          file per track
        </para>
      </listitem>

      <listitem>
        <para>
          you have a sample library containing 500 small audio files
        </para>
      </listitem>
    </itemizedlist>

    <para>
      In the first case, your goal is probably to have 8 tracks (at least),
      with each track containing a single audio file. In the second case,
      its a lot more likely that you simply want to be able to use any of
      the samples easily, but do not want any tracks created as a direct
      result of the import/embed. It is very important that you understand
      this distinction: many new users think there should be a "simple" way
      to import existing audio without understanding that the goal of
      importing/embedding is not always the same.
    </para>

    <para>
      Ardour provides two different options when importing. You can
      import/embed audio files as new tracks, or you can import/embed them
      into the region list, where they will be available as regions to put
      into new or existing tracks. You can also insert import/embed audio
      files directly into an existing track.
    </para>
  </section>

  <section id="how-to-import-embed">
    <title>How to import/embed</title>
    <para>
      There are three pathways for importing/embedding an audio file into a
      session. One is initiated from the File menu, one is initiated from
      the track context menu, and one is initiated from the region list.
      These methods are all equivalent: they open the file import dialog.
      Once the dialog is open, you can choose to add new audio as new
      tracks, as regions in the region list, or as audio in the selected
      track. You can change this behavior once the dialog is opened.
    </para>

    <para>
      <caution>
        <para>
          If you want to import a file directly into an existing track, you
          must have the track selected in the editor before beginning the
          import.
        </para>
      </caution>
      If you use Nautilus as your file manager, you can easily import files
      into your project by dragging them onto the desired track, then
      releasing the mouse button. The file will then be
      <link
    linkend="gt-embed">embedded</link> into your session.
    </para>
  </section>

  <section id="how-to-embed-a-file">
    <title>Working with Tags</title>
    <para>
      A "tag" is bit of information, or <emphasis>metadata</emphasis>, that
      is associated with a data file. Specifically, tags are keywords or
      terms that you feel have some relevance to a particular soundfile.
      Ardour can store these tags in a seachable database so that you can
      quickly search for sounds based on the tags that you have assigned to
      them. For example you can assign the term "120bpm" to a sound, and
      then when you search for this tag, the file will appear in the search
      list. Tags are independent of the filename or anything else about the
      file. Tags, and the file paths that they are associated with, are
      stored in a file called "sfdb" in your Ardour user folder.
    </para>

    <section id="embedding-via-the-region-list">
      <title>Adding tags to a file</title>
      <orderedlist>
        <listitem>
          <para>
            Select a file in the import dialog.
          </para>
        </listitem>
        <listitem>
          <para>
            Enter the tags in the box to the right labeled "Tags".
          </para>
        </listitem>
      </orderedlist>
      <para>
        Tags are stored when the input box loses focus, there is no reason
        to explicitly "save" them.
      </para>
    </section>

    <section id="embedding-from-a-track">
      <title>Searching for files using tags</title>
      <orderedlist>
        <listitem>
          <para>
            Select the "Search Tags" tab on the import dialog.
          </para>
        </listitem>
        <listitem>
          <para>
            Enter the tag(s) to search for and press "Search".
          </para>
        </listitem>
      </orderedlist>
      <para>
        Files which have been "tagged" with the input terms will appear in
        the results window. You can audition these files and apply tags to
        them from this window.
      </para>
    </section>
  </section>

  <section id="searching-freesound">
    <title>Searching Freesound for soundfiles (optional)</title>
    <para>
      Freesound (http://freesound.iua.upf.edu/) is an online database with
      thousands of soundfiles which can be freely used in many projects
      (visit Freesound for the exact licensing terms). Ardour has an option
      to download files directly from freesound directly from the import
      dialog. Freesound files are tagged with metadata which you can use to
      search for relevant sounds.
    </para>

    <para>
      To enable the Freesound search engine inside Ardour, you must build
      Ardour with the option FREESOUND=yes, and you must have the "curl"
      library installed. This is currently optional but may later become a
      standard feature of Ardour.
    </para>

    <para>
      To download files from Freesound, you must first register for a
      username and password on the freesound website. Enter this information
      into the text entry boxes, enter your tags to search for, and then
      click "Start Downloading". Ardour will begin downloading files, and
      the "Start Downloading" button will change to say "Cancel". Clicking
      the "Cancel" button will wait until the currently downloading file is
      finished, then the button will switch back to "Start Downloading". The
      button will also say "Start Downloading" once Ardour has downloaded
      all of the files that match the given tag. In the future there may be
      better indication of the file progress and better filtering on the
      filetypes that are downloaded.
    </para>

    <para>
      If you later search for the same terms, you will see that the files
      appear in the list more quickly, because any files in the search set
      that are already downloaded won't have to be downloaded again.
    </para>

    <para>
      Sounds that are downloaded from Freesound will automatically be given
      tags in Ardour that match their tags in the Freesound database. This
      means that once they are downloaded, Ardour can search for the local
      files very quickly using the "Search Tags" tab.
    </para>
  </section>
<!--
	<xi:include xmlns:xi="http://www.w3.org/2001/XInclude" 
		href="Some_Subsection.xml" />
	-->
</chapter>