Unformatted text preview: through the selection, the plug-in has captured a noiseprint and switched to
Noise Reduction is reducing noise in real-time, but only reducing the selected noise.
3. Select the entire audio portion you want to process and click the Play button. The entire file will play, allowing you to change parameters and simultaneously listen to the results.
4. From the File menu, choose Save EditList File As to save the processed file. Operation basics (General page)
Once you have obtained a noiseprint, the hard part is done and for most applications you
will not have to capture another noiseprint until you are ready to process a different file.
The three most important processing controls are the Reduce Noise by slider, the Reduction
type drop-down list, and the Noise bias slider. In most cases, using the default presets and slightly adjusting these three controls will give
you excellent results. However, we encourage you to try the many adjustment sliders
available like the FFT Size or Release Speed. CHP. 2 NOISE REDUCTION 16 Using the Reduce noise by slider
The Reduce noise by slider is the control that you will use the most. This slider controls how much the plug-in will reduce the noise by. A higher number will do
a better job of reducing the noise, but will most likely add unwanted artifacts to the material
you wish to keep.
Settings of 10 to 20 dB will usually provide the best results on a single pass, but depending
on the source material, you may be able to use higher settings.
A good strategy on stubborn noise is to do multiple passes of Noise Reduction at the 10 to 20
dB range by processing the file, getting a new noiseprint, and then processing the file again.
Two passes of 20 dB will usually give you better results than a single pass at 40 dB. Using the Attack speed and Release speed sliders
The Attack speed and Release speed sliders determine how quickly the plug-in reacts to
changes in noise level. The Attack speed determines how fast the noise reduction algorithm reacts to a noiseless
signal. The default value is recommended. Very low values often remove fast transients from
a sound, while very fast values can create audible artifacts during fast attacks. When using a
very large FFT size, the Attack speed should be higher.
The Release speed determines how fast the noise reduction algorithm reacts to a noisy signal.
The default value is recommended. Higher values can cause the algorithm to trim the ends
off of long decaying sounds. When using a very large FFT size, the Release speed should be
To change either of these settings, you may adjust the slider left or right or type a value into
the edit box. NOISE REDUCTION CHP. 2 17 Compensating for lost high frequencies
A high-shelf filter is included to brighten the sound after removing noise. Frequently, noise
reduction decreases the high frequencies in the source material. Enabling the high-shelf
filter and boosting the high frequencies can compensate for this loss.
To enable the filter, select the High-shelf start freq. check box. To vary the frequency above
which the filter will be applied to the source material, click and drag the slider left or right,
or type a value in the edit box. Below the High-shelf start freq. slider is the High-shelf gain slider. To adjust the level by which
the filter boosts or cuts high frequencies, click and drag the slider to the left or right, or type
a value (in decibels) in the edit box. Adjusting the precision of frequency analysis
The Windowing section provides settings to determine the precision of the frequency analysis
of the noiseprint. indicates the size of the samples that the algorithm uses to analyze the frequency
spectrum of the sound. To adjust the FFT size setting, choose a new setting from the dropdown list. FFT size controls the amount of overlap between frequency analysis windows. Overlapping
sometimes allows for more accuracy of the analysis. To adjust the Overlap, click and drag the
slider to the left or right, or type a value in the edit box. The default setting of 75 is
recommended. For faster processing, set the value to a lower value. Overlap CHP. 2 NOISE REDUCTION 18 Seeing the removed noise
The Keep residual output check box inverts the Noise Reduction process: instead of hearing
the source material you are keeping, you will hear the noise you are removing. This can be a
valuable tool when adjusting parameters because you can hear whether you are removing
any of the source material you want to keep. For example, if you are reducing some noise from the background of a speech and you hear
some of the speech in the residual output, you know you are adversely affecting the quality of
the source material. You can then change the settings so that none or as little of the residual
output contains the material you want to keep. Noiseprint modification and navigation
The Noiseprint page is used to manually modify the noiseprint envelope points. By making
modifications on this page, you may t...
View Full Document
This note was uploaded on 05/15/2013 for the course EMUS 201 taught by Professor Pardal during the Winter '10 term at Life.
- Winter '10