Configure Dither Algorithms in Audacity

Dithering is an essential process in digital audio editing that masks quantization distortion with low-level randomized noise whenever you downsample bit depth. In Audacity, while you cannot select dither algorithms directly inside the standard Export dialog, you can configure the global dither algorithm through the application preferences. This guide explains step-by-step how to navigate Audacity's settings to select custom dither algorithms, such as Triangle or Shaped, and how to ensure they are properly applied to your exported audio files.

Accessing Dither Settings in Audacity

Audacity manages its dithering algorithms globally under its internal conversion settings rather than on an export-by-export prompt. To change the algorithm:

1. Open Audacity.
2. Access the Preferences menu:
   • Windows / Linux: Go to Edit > Preferences (or press Ctrl + P).
   • macOS: Go to Audacity > Preferences (or press Cmd + ,).
3. In the left-hand navigation pane, select Audio Settings (in Audacity 3.2 and newer) or Quality (in Audacity 3.1 and older).
4. Locate the section titled High-quality Conversion.
5. Find the Dither dropdown menu.
6. Select your preferred algorithm from the list: None, Rectangle, Triangle, or Shaped.
7. Click OK at the bottom of the window to save your changes.

Choosing the Right Dither Algorithm

Audacity provides three active dither algorithms under High-quality Conversion:

• Triangle (Triangular): Produces a linear probability density function. It completely eliminates quantization distortion without introducing noise modulation. Triangle is the most versatile, standard choice, particularly if the audio will undergo further processing or encoding later.
• Shaped: Applies psychoacoustic noise shaping. Instead of spreading noise evenly across the entire frequency spectrum, it pushes the noise floor into higher, less audible frequencies (typically above 14–15 kHz). This yields the lowest perceived background hiss, making it the preferred choice for the final master of 16-bit audio at 44.1 kHz or 48 kHz.
• Rectangle (Rectangular): The simplest form of dither. While it reduces distortion, it can cause slight noise modulation depending on the signal level. It is rarely recommended for final music or spoken-word delivery.

Applying the Dither to Your Export

Once configured in Preferences, the selected dither algorithm is automatically applied whenever a bit-depth reduction occurs:

1. Ensure your default working track is set to 32-bit float (Audacity’s native processing format).
2. Go to File > Export Audio.
3. Set your target format to a lower bit depth, such as WAV (Microsoft) signed 16-bit PCM or MP3.
4. Click Export. Audacity will automatically apply the Triangle or Shaped dither algorithm chosen in your High-quality Conversion preferences during the render.