Clicking the Learn button creates a noise signature based on the audio segment currently selected on screen. There are two Learn modes: Learn First Audio mode and Learn Last Audio mode.
 |
Learn First Audio mode is the default Learn mode. It is designed for use with audio that has an identifiable noise-only section that you can locate and pre-select. To use this mode, locate and select the noise-only portion of the audio, click the Learn button, start playback, and BNR will build a noise signature based on the first 16 milliseconds of audio playback. First Audio Learn mode can be thought of as a trigger-learn mode, since noise capturing is triggered by the first audio that DINR receives.
Learn Last Audio mode is designed to let you locate and identify a segment of noise on-the-fly as you listen to audio playback. In this mode, you first Alt-click (Windows) or Option-click (Mac) the Learn button, then initiate audio playback. When you hear the portion of audio that contains the noise you want to identify and remove, click the Learn button a second time. BNR will build a noise signature based on the last 16 milliseconds of audio playback. The Spectral Graph displays data in real-time in Learn Last Audio mode.
The Fit button computes a noise Contour Line with approximately 30 breakpoints to fit the shape of the current noise signature. The Contour Line can then be edited to more closely fit the noise signature or to reduce specific frequency bands by dragging, adding or deleting breakpoints.
 |
Pressing the Up Arrow or Down Arrow keys on your computer keyboard lets you raise or lower the entire Contour Line, or a selected portion of the Contour Line. The Left/Right arrows lets you move a selection left or right. To select a portion of the Contour Line with multiple breakpoints, Control-drag (Windows) or Command-drag (Mac) to highlight the desired area.
After you use the Fit function, BNR will automatically boost the entire Contour Line 6 dB above the noise signature so that all noise components of the audio file are below the Contour Line. You may want to adjust the Contour Line downwards as needed to modify the character of the noise reduction.
The Super Fit button creates a noise Contour Line consisting of over five hundred breakpoints in order to follow the shape of the noise signature more precisely.
 |
The Auto Fit function is designed to generate a noise curve for audio that lacks a noise-only portion for DINR to learn. Clicking Auto Fit computes this generic noise curve based on the points contained within the currently selected audio, then fits the Contour Line to it. To use the Auto Fit function, you must first make a selection in the Spectral Graph by Control-dragging (Windows) or Command-dragging (Mac).
If the selected audio has both noise and desired sound components, you can generate an approximate noise-only Contour Line by selecting a frequency range that appears to be mostly noise, then pressing the auto fit button. You can then edit the resulting noise Contour Line to optimize the noise reduction.
These arrows behave differently depending on whether or not there is a selection of points along the Contour Line.
 |