Device Toolbar

Gale 27Feb13: ToDo-2 WASAPI text added.

Device Toolbar  provides an easy way to select your required combination of interface host, playback device, recording device and input channels, without having to open Devices Preferences.

Using the drag handle at right, you can expand Device Toolbar rightwards to be wider than the default width shown, or drag it leftwards to be less wide. You can hover over any of the four boxes at any time to see the full text for the selected item.

Device Toolbar grays out when playing, recording or paused. You cannot change device choices while the device is in use.

Audio Host

Selects the particular interface with which Audacity communicates with your chosen playback and recording devices. On Windows, the choice is between MME (the Audacity default, more compatible), Windows DirectSound (a slightly modern replacement), Windows WASAPI (Audacity currently provides only "loopback" inputs for this host - these inputs can be used to record computer playback on Windows Vista and later) and Windows WDM-KS a (host which bypasses the kernel mixer that is normally used to mix outputs from multiple applications and can thus provide very low latencies). In Linux there is often only one option: ALSA, other options could be OSS and/or Jack Audio Connection Kit (also known as "Jack" or "Jackd"). On Mac OS X the only choice is Core Audio. Generally you will want to use the default host.

• On Windows XP or earlier (given a recent computer), WDM-KS's and DirectSound's much shorter path to the hardware should produce lower latency than MME. Under WDM-KS, only one application is allowed to access the audio device at a time. So for example if you are already playing audio in the web browser or another application, you will not be able to play audio in Audacity using the same device that the web browser is using.
• On Windows Vista, Windows 7 and Windows 8:
  • Windows DirectSound and WDM-KS may have only slightly lower latency than MME because both interfaces are emulated. However if you reduce "Audio to Buffer" in Recording Preferences you can achieve extremely low latencies. Just reduce "Audio to buffer" until the sound begins to break up.
  • Windows WASAPI  host currently provides only "loopback" inputs. These inputs can only be used to record computer playback.

Selecting Windows DirectSound and enabling both "Exclusive Mode" boxes in Windows "Sound" allows Audacity to request audio direct from the device without resampling or kernel mixing. If you use WDM-KS on Windows Vista or later, the currently playing application has exclusive access to the audio device irrespective of Windows "Exclusive Mode" settings. See the Wiki page for Windows 7 for more explanation.

Output Device

Choose the built-in or attached sound device that you want to use for playback.

In the image above, "Microsoft Sound Mapper - Output" appears when the Windows "MME" Host is selected. The "Mapper" output is not a separate device, but is the device from the output list that is currently chosen as the default system playback device in the Windows Control Panel. When "Windows DirectSound" Host is selected, the device that maps to the current Windows playback device is called "Primary Sound Driver".

In most other cases (for example, the inbuilt computer sound device), each entry for output device consists of the output type (such as speakers), followed by the name of the audio device the output belongs to.

Input Device

Choose the built-in or attached sound device that you want to use for recording. Input sources are no longer chosen in Mixer Toolbar.

In the image above, "Microsoft Sound Mapper - Input" appears when the Windows "MME" Host is selected. Similarly to the "Mapper" output, the "Mapper" input is the device from the input list that is the current default Windows recording device. When "Windows DirectSound" Host is selected, the device that maps to the current Windows recording device is called "Primary Sound Capture Driver".

In most cases (for example, the inbuilt computer sound device), each entry for input device consists of the input type (such as microphone), followed by the name of the audio device the input belongs to.

If a device only has one possible input for selection by the computer (for example, a USB microphone), only the name of the device will be listed, with no input type.

If you attach or disconnect an external device while Audacity is open, use Transport > Rescan Audio Devices to update the device lists before playing or recording.

Input Channels

1 (Mono), 2 (Stereo) or the number of channels that are provided by the drivers of your sound device. On most inbuilt sound devices, especially on Windows, only mono or stereo will be available. For some devices on Windows, choosing Windows DirectSound in "Host" above may be more likely to reveal options for recording more than two channels. On some devices capable of recording more than two channels, an explicit "multi" device may appear in the "Device" drop-down for recording all the channels simultaneously.

• If your input is mono, such as most microphone ports for the inbuilt sound device, selecting "2 (Stereo)" in Input Channels duplicates the mono source to both channels, merely producing a dual mono recording.
• If your input is stereo, then as well as setting Input Channels to "2 (Stereo)", ensure any settings in the system or sound device control panels are stereo. Many input devices on Windows Vista, Windows 7 and Windows 8 default to mono recording even if they are stereo devices. To correct this, follow these instructions.