11.0

EFFECTS MENU 529
Noise reduction parameters
Noise sample: Selects the wave project with the noise test. It must be opened in
Samplitude as a normal wave project beforehand.
Rate: This parameter allows you to set the resolution (low, medium, or high) used by the
noise reduction algorithm. The higher the rate, the longer the algorithm will take to process.
A higher resolution does not always guarantee the best results. For example, low resolution
often works best when applying noise reduction to short percussive sounds, and speech.
Precision and real-time precision: The exactness of the algorithm can be selected here
in three steps. Processing times increase with increasing algorithm precision, and the
quality of the results also improves.
Correction: The effectiveness of the algorithm can be controlled in three steps with "Low",
"Middle" and "High" settings, followed by the "Fine Tuning" slider. With this parameter the
level of the noise sample's spectrum will be lowered or raised. This effect is visualized in
the graphic display: the yellow curve represents the original noise sample spectrum,
and the blue curve is the corrected spectrum, which will be used by the algorithm to
remove noise.
Low settings can lead to incomplete noise removal or increased artifacts. High settings can
cause discoloration of the original sample, phase-like distortion, or chirping modulation
effects. The occurrence of these effects is dependent on the type of noise as well as the
consistency of the original material. A higher-level noise does not necessarily call for a
higher correction level.
Mix (Reduction): This parameter determines the amount (in dB units) by which the noise is
reduced.
In which situations does it make sense to leave some level of noise?
When working with gramophone recordings, it can be desirable to leave some of that
"gramophone feeling" in. Background noise from "on location" reports does not need to be
removed completely. When noise is not completely eliminated, the occurrence of artifacts
or discoloration is reduced. Think of the process in terms of reduction, not 100%
elimination.
Parameters for artifact suppression: These settings serve to suppress artifacts as they
may occur in broadband distortions at higher levels. If you select a very high value, the
quality of the results is reduced. For this reason, you should select the value at which the
artifacts are just about no longer audible.
With uncritical distortions, no artifacts occur. In this case, high settings would produce poor
results. If you're not sure, reset the values to zero.
Smooth attack: This parameter controls the attack when applying noise reduction. A
higher value will reduce artifacts, but alters the response time of the audio. Generally
speaking, this parameter is most effective for diminishing artifacts in musical tracks. In the
case of vocals or speech, higher values lead to less than optimal results.