Audacity is a full-featured, but simple to use application for destructively editing many different types of digital audio, capturing audio from a digital or analog source, converting digital audio from one format to another, and basic (very) multi-track recording tasks.
Digital audio editing is comprised of many things, but common editing tasks include:
There are six major editing tools used to edit the digital audio with Audacity. We will cover a few of them in depth here. Please see the Audacity documentation for help with the advanced tools. Here is what the main editing tools look like on the toolbar.
The main tools are, starting from the top left and working clockwise:
The Select Tool is the most commonly used tool, and it works just like the selection cursor in many word processing applications. The icon is even the same vertical "I" shaped bar, and it is used for selecting linear portions of one or more digital audio files to apply edits to. To use the tool, click on the "I" bar shaped Select tool on the toolbar, click the waveform of the audio where you wish the editing to begin, drag, and release at the point where you would like the edited portion to end. If you want to apply an edit to an entire track of audio, you can either click in the track control box underneath the sliders for panning and track volume, or click the track that you wish to edit, and click Edit->Select All. Here is what a selection of a waveform looks like:
The Envelope tool is useful for changing the amplitude of the waveform at different points. This creates "fades", either fading in or fading out. Many of these volume fades can be applied to a file at different points, and this is what is referred to as applying an envelope. A good example of using the envelope tool would be to remove pops and clicks or tape hiss in between two parts of an audio file. One could simply select the area of space between the program material, click Edit, and choose Silence. This would make the area that was selected completely silent. The problem with this, is that it is very noticeable. Psychoacoustically, we may not even hear the tape hiss if it is quiet enough, but we will perceive the change if it all of a sudden drops out completely. Here is where applying a quick fade out and another quick fade in comes into play. A quick fade out is much smoother from a psychoacoustic standpoint, and will not even be perceived if done correctly.
The envelope tool is node-based. Click on the envelope tool to select it. Now click the waveform at the highest point where you want your first fade start. This adds an envelope node. Decide where either you want this single fade to end, or if you decide to also do a fade in, decide where you want that fade to end. Place another node here. Now place your cursor midway between the two nodes, at the highest point in the waveform, click and drag down. You will see a smooth, curved line appear on both sides of your cursor. This line will ramp down to the left of the cursor and up to the right. This is how the audio will change amplitude. Applying an envelope in this way ensures that the transitions will be smooth. The great thing about the envelope tool is that many nodes can be added, and many fades applied to a specific file, allowing for great edits throughout. Here are a few shots of the envelope tool in action.
One useful tool, especially when working with multiple tracks of audio simultaneously, is the time-shift tool. This tool allows you to move a whole track, or a selection, forwards or backwards in time by any amount, from a few minutes or a few seconds, down to a fraction of a secord, or down to the individual sample. To use the time-shift tool, click the double-pointed arrow icon on the editing toolbar. You are now in time-shift mode. Drag the cursor over the waveform to the left to shift the audio forward in time (making it start sooner), or to the right to shift it back in time (making it start later). When working on a single audio track, this can be useful if there ia a large amount of silence at the start of the program. The time shift tool allows you to simply shift the whole audio program back a few seconds, eliminating a lot of the silence at the beginning of the track. When working with multiple audio tracks, the time shift is most useful for syncing up the tracks. Here are a few before and after shots of a track that has been time shifted.
Sometimes, it is hard to apply the correct edit if the resolution of the displayed waveform is not fine enough. Audacity allows you to zoom in on the waveform, letting you see the audio data down to fractions of a second, or even down to the sample. This can be very, very useful for sensitive edits, like removing a word from a spoken word program (where it would be difficult withough getting right down to fractions of a second). Audacity provides a zoom in and zoom out button on the toolbar to accomplish this, or one can use the Zoom tool on the editing toolbar as a shortcut. When using the zoom tool, left clicking will zoom in on the waveform, and right clicking zooms out. Here are a few shots of an audio file waveform being zoomed in on down to fractions of a second, and then down to individual samples.
Audacity comes with a core group of effects that can be applied to digital audio data. The application can also be expanded with the use of effects plugins. On the GNU/Linux platform, these plugins will be in the form of LADSPA effects plugins and Nyquist plugins. LADSPA stands for Linux Audio Developers Simple Plugin API, and it is currently the standard API to write effects plugins for audio applications on GNU/Linux. Nyquist is an electronic music and signal processing programming language, and it is possible to write plugins for Audacity in Nyquist. A great, and very large selection of LADSPA and Nyquist plugins are available in most GNU/Linux distributions. The two packages to look for for the LADSPA plugins are the swh-plugins package (may be named differently in different distros) and the tap-plugins package. The swh-plugins were written by Stephen H. Harris, who is an expert in the field. The tap-plugins package is Tom's Audio Plugins, and provides some really great time-based effects, like reverbs and delays, and some good dynamics effects, like compressors.
Since Audacity is cross-platform software, on Windows and Macintosh, it is also possible to use VST plugins with an add-on available on the Audacity project page called the VST enabler. VST is a plugin architecture by Steinberg Co. that has been, along with DirectX, the standard plugin architecture for Windows audio applications. While an API and developer tools are available to write both VST and DirectX effects, both VST and DirectX (from Microsoft) are closed standards. LADSPA and Nyquist are free and open. LADSPA plugins only work on GNU/Linux, although there has been some success to port it in part to other platforms.
Audio effects can be grouped into several major groups. These are:
Most of the plugins available are mostly useful for pro-audio work, but there are a few that are very useful for the casual user. One really good one is the Amplify effect that is distributed with Audacity. Have you ever burned audio tracks to a CD, only to discover that you have to crank the stereo volume way up on certain tracks? Before burning these low-volume tracks, import them into Audacity, select the entire audio program (Select->Select All, or click the track control box underneath the volume and panning sliders), click on the Effect menu, and choose Amplify. One really great thing that Audacity does is analyze the audio program and suggest the maximum level of amplification to apply without allowing clipping (distortion due to levels being too high). This is very useful and user-friendly. If you need to amplify a track and don't want to monkey with the levels until you find the distortion point, simply follow the procedure above, make sure that the "Don't allow clipping" option is checked, and accept the level of amplification that Audacity provides for you.
Another effect that is great for the casual user is the Normalize effect. What this effect does is normalize the amplitudes of the audio program, so that there are no very, very quiet sections and no very, very loud sections. Basically, it analyzes the program, and adjusts the amplitudes of various parts of the program and averages them to a degree. This effect does have parameters, so it can be used conservatively, but if used to liberally, it can make musical program content sound unnatural. This effect is great for spoken-word program content.
Audacity is capable of importing and working with lots of different kinds of digital audio formats. These include raw, headerless digital audio, RIFF WAV files, Apple AIFF files, Sun AU files, MP3 encoded audio, and Ogg Vorbis encoded audio natively. Audacity is also capable of exporting edited audio data to any of these formats. It will do so natively (without external libraries) for all formats except for MP3. This is due to software patent issues. Audacity can export to MP3 with the LAME library installed, but does not ship with it's own MP3 encoder. This is also a limitation of commercial products (SoundForge, the most common sound editor for Windows, needs a commercial MP3 encoder plugin, I am assuming to cover patent issues and to pay license fees.) LAME is available for most all GNU/Linux distros, and can always be obtained as source, so one way or another, Audacity will always be able to export to MP3. On Windows, other MP3 encoders can be used, also, such as BladeENC.
To export audio data to another format (or the same format), several techniques can be used. If you are working with a single track of audio data (a single mono or stereo audio file), you can either choose "File->Eport to WAV", or any of the equivalent format choices. This will export the entire file, applying all editing changes and effects. If working with multiple audio files at once, this option will mix all of the tracks down to a single file. By default, this will be a stereo file, although mono is an option. Once can also export a selection of the audio file. Select the portion that you would like to export with the Selection tool, and choose "File->Export Selected to WAV" or any of the other format options. This will export only the selected portion, ignoring all other parts of the file, and ignoring all other audio tracks if working with multiple files at once. A third option, and this applies only when working with multiple files simultaneously, is the File->Export Multiple option. This command allows one to export each audio track to a separate file. This is very useful if you work with multiple files, apply edits to each file, and wish to keep them all separate after editing. This is very common in pro-audio and broadcast editing work.
Audacity is capable of working with, importing, and exporting digital audio in a variety of sample rates and bit depths. These settings are accessed by clicking "File->Preferences", and customizing the settings on the Quality and File Formats tabs. The sample rate and bit depth used are application specific, so change according to your need or preference.