Filter item in the menu.

Filters are used to remove a range of frequencies or amplitudes from a sound and can produce a variety of effects. The submenu contains several filter related effects.

Use Auto Offset Removal automatically remove a vertical shift or dc offset in the waveform. Offsets occur when audio is wired through several external devices that do not share a common ground. Computers with integrated sound hardware often have significant offsets as well. If the waveform is constantly above or below zero during silence, then it has an offset that must be removed. Offsets can cause pops and clicks between edit points and other problems.

Unlike , this effect does not need to scan the entire selection and offsets cannot be set to specific values. Values to cancel out existing offsets are automatically calculated and used. Even varying offsets are removed.

Auto Offset Removal Setting
Setting Description
Duration for offset calculation (s) Sets the length of time to analyze the audio for an offset. For constant, stable offsets, set this value to the maximum. For fluctuating offsets, use lower values.

See Also: , ,

Bandpass filters block all outside a specified range, keeping only frequencies within the range. Bandstop filters block all frequencies within a specified range, keeping all other frequencies outside the range.

Bandpass/Stop Settings
Group Setting Description
Initial frequency range From (Hz),
To (Hz)
Sets the initial frequency range of the pass or stop band.
Final frequency range From (Hz),
To (Hz)
Sets the final frequency range of the pass or stop band. The band range fades from the initial setting to the final setting over the length of the selection. Dynamics must be checked to set these values.
Settings Bandpass,
Bandstop
Sets the type of filter to use. Select Bandpass to keep only the frequencies within the range. SelectBandstop to block all the frequencies in the range.
Static,
Dynamic,
Steepness
See filter for details.
To remove hum and hiss from a voice recording:
  1. Set From to 200.
  2. Set To to 3800.
  3. Choose the Bandpass option.
  4. Choose the Static option.
  5. Set Steepness to 10 or higher.
  6. Choose OK.

See Also: , , , ,

Equalizers are commonly found on stereo systems. They boost or reduces certain . Simple equalizers control only treble and bass. GoldWave's equalizer controls 7 bands.

Center frequencies for each of the 7-bands are given at the top of each fader. Adjust the faders to boost or reduce a band between +12dB to -24dB.

To change bass, adjust the two or three left-most bands. To change treble, adjust the two or three right-most bands. Several are included to demonstrate bass, mid, and treble changes.

More detailed equalization is possible using the or .

See Also: , , , ,

Lowpass filters block high pitched, (treble) but allow low pitched frequencies (bass) to pass. They can be used to reduce high end hiss noise or remove unwanted sounds above the given cutoff frequency. If you were to apply a lowpass filter with a cutoff frequency of 1000Hz on speech, it would make it sound muffled and deep. Lowpass filters can also be used to eliminate aliasing noise when used before downsampling.

Highpass filters block low pitch frequencies, but allow high pitched frequencies to pass. They can remove deep rumbling hum or remove unwanted sounds below the given cutoff frequency. If you were to apply a highpass filter with a cutoff frequency of 1000Hz on speech, it would make it sound thin and hollow.

Low/Highpass Settings
Group Setting Description
Cutoff frequency Initial cutoff (Hz) Sets the constant cutoff frequency for static filtering or the initial frequency for dynamic filtering.
Final cutoff (Hz) Sets the cutoff frequency at the end of the selection, allowing you to fade from one cutoff to another over the selection. Dynamics must be checked to set this value.
Settings Lowpass,
Highpass
Sets the type of filter to use. Select Lowpass to keep only the frequencies below the cutoff frequency. Select Highpass to keep only the frequencies above the cutoff frequency.
Static,
Dynamic
Sets whether the filter is constant or changing. Select Static to keep the cutoff frequency constant throughout the selection during processing. Select Dynamic to fade the cutoff frequency from the initial value to the final value over the duration of the selection.
Steepness Sets how sharply the filter cuts off frequencies outside the cutoff frequency. A higher steepness makes the filter sharper, but it also increases processing time. In technical terms, the steepness specifies the number of second order cascade filters used.
To make speech gradually become more hollow and thin:
  1. Set Initial cutoff to 60.
  2. Choose Dynamic.
  3. Set Final cutoff to 1000.
  4. Choose Highpass.
  5. Choose OK.
Remove all high end hiss noise from a voice recording.
  1. Set Initial cutoff to 4000.
  2. Choose Lowpass.
  3. Choose Static.
  4. Set Steepness to 20.
  5. Choose OK.

See Also: , , , ,

Use Noise Gate to reduce or eliminate noise in quiet or silent parts of a recording. Brief noises, such as mouse clicks or microphone bumps can be muted as well. A noise gate cannot remove noise from louder parts of the recording. Use for that instead. To completely delete silences from a recording, use .

Use before using this effect to ensure volume is optimal for the presets.

Noise Gate Settings
Setting Description
Threshold (dB) Sets the level of audio that is considered noise. Audio below the threshold is reduced. The threshold should be set as low as possible so that only noise is removed. If the threshold is set too high, short and quiet sections of audio may be muted too.
Attack (s) Sets the amount of time required to fade out to complete silence when the audio is below the threshold.
Release (s) Sets the amount of time required to fade in to full volume when the audio rises above the threshold again.
Reduction (%) Sets the level of reduction in the quiet parts. Normally this value would be set to 100 to replace the quiet parts with complete silence. A value of 25 reduces the noise only slightly (-3dB).
Ignore (s) Sets the duration of brief sounds to ignore, even if they are above the threshold. Any sound less than this duration is muted. Use this to remove clicks, pops, and other very short sounds. Set this to zero to ensure that all sounds above the threshold are retained.

See Also: , ,

Noise Reduction uses frequency analysis techniques to remove unwanted noise, such as a background hiss, a power hum, or any continuous, consistent sound. It cannot be used to separate or remove complex or brief sounds, such as coughs, laughter and applause. It cannot remove instruments or vocals from music.

If the noise to be removed contains pops and clicks, use to eliminate those before using Noise Reduction. Any pops or clicks in the noise profile/envelope may cause a significant reduction in quality.

A frequency analysis window with controls displays frequency in Hertz on the x axis and frequency magnitudein on the y axis. Figure: Noise Reduction
Noise Reduction

The frequency analysis provides graphical information about all frequencies within the sound at the given Time. For multichannel sounds, each channel is shown it's own colour. Moving the scroll bar located below the analysis window changes the time, showing the frequency analysis of a different part of the sound. The detail of the analysis depends on the FFT size setting, explained below.

Noise is removed using a reduction envelope. The shape of the envelope must closely match the shape of the noise to remove. The frequency analysis graph can help determine that shape. Change the analysis time so that it coincides with a time in the sound where only the noise is heard (use the button to play the file to find such a place and time). Once you have isolated the noise in the analysis graph, you can then create the envelope.

The envelope can be created in four different ways, depending on the eeduction envelope setting. Use clipboard (see below) generally gives the best results, but requires the noise to the clipboard before using Noise Reduction. Also try the .

Noise Reduction Settings
Group Setting Description
Reduction envelope Use shape Creates an envelope based on the shape defined by the . Click the mouse in the analysis graph to draw the shape. A simple horizontal line at about 80dB removes a quiet hiss from a sound. In some cases, you'll need to trace the outline of the analysis graph or draw completely different shapes to reduce the noise.
Use current spectrum Creates an envelope based on the shape of the graph shown in the frequency analysis window. Use this to remove a complex buzz or hum. The analysis Time must be set to a place where the noise is heard by itself.
Use average Creates an averaging envelope throughout noise reduction processing. The envelope is continuously updated, based on the average frequency analysis of the sound. Use this setting if the noise changes frequently throughout the sound.
Use clipboard Creates an envelope based on an analysis of the waveform in the clipboard. This is the most flexible option and usually gives the best results. Before you can use this option, you must a piece of noise into the clipboard. The piece must contain only the noise to be removed from the rest of the file. The noise can even be copied from a different file. After copying the noise, remember to change the to the part of the file you want to apply the noise reduction.
Shape controls Point,
X,
Y,
Time
See .
Settings FFT size Sets the detail of the frequency analysis and the noise reduction envelope. Usually values of 11 or higher give the best results. See in for more information.
Overlap Sets the amount of audio to overlap from one calculation to the next. A value of 4x works well. See in for more information.
Scale (%) Sets reduction envelope scale. A value of 100 uses the envelope as it is. A value of 200 doubles the envelope amplitude, which doubles the amount audio removed from the sound. A value of 50 halves the envelope amplitude, which halves the amount removed. Normally it should be set to 100.
Output noise only Makes the effect perform the exact opposite of noise reduction so that only the noise remains in the output. This is useful when previewing the effect to hear what is actually being removed from the audio. Do not check this box when removing noise.

See Also: , ,

A parametric equalizer (shown below) is a flexible tool for reducing or enhancing ranges of . GoldWave presents an easy to use interface where all the parameters for up to 40 bands can be configured quickly. contain some commonly used templates. Figure: Parametric Equalizer
Parametric Equalizer

Graph window
The graph shows frequency on the x-axis in Hertz and the gain on the y-axis in decibels. Each enabled band is displayed in the graph as a diamond shaped box located at its center frequency and gain. The width of the box shows the bandwidth. The currently selected band is shown in blue and its exact settings are given in edit box controls.

A short time frequency analysis graph is drawn with the left channel in green and the right channel in red. The time of the analysis can be changed using the scroll bar located at the bottom of the graph. The analysis helps determine what frequencies to boost or reduce. A high pitched squeal, for example, would appear as a spike near the right side of the graph. Whereas a low pitched hum would be a spike or bump near the left side.

Controls
A band is configured by selecting its number from the Band box and adjusting the Gain, Width, and Center faders. A quicker method is to drag-and-drop the band to a new location on the graph. Note that because of the logarithmic frequency scale, the width of a diamond changes as you move it left or right. The bandwidth, however, remains constant.

Use the diamond plus button to add more bands. Use the diamond minus button remove existing bands. The current (blue) band given in the Band box is the one removed.

The "Notch" is effective for removing a simple tone from a sound, such as a 60Hz hum or telephone dial tones. The "Bass boost" and "Treble boost" presets work the same way as the bass and treble controls on a stereo system. Adjust the gain up or down to control them.

See Also: , , ,

A pop/click filter is a specially designed filter that searches for abrupt changes in the sound and eliminates them. Such a filter is often used to remove pops and clicks caused by dust and scratches when recording old vinyl records.

When a click is detected, the filter attempts to reconstruct the damaged waveform based on the surrounding waveform shape making the repair almost imperceptible. However with excessive pops and clicks or at low tolerance levels, reconstructed waveforms may overlap and sound distorted. The tolerance setting should be kept as high as possible. Using a very low setting may introduce more distortion than existed in the original. This is most noticeable in voice recordings and instrument solos, particularly trumpet solos. Always start with the maximum tolerance setting for those types of sounds.

Pop/Click Setting
Setting Description
Tolerance (%) Sets how abrupt a change in amplitude can be before it is considered a click. It is best to start with a value of 1000% or higher. Using a lower value will detect more clicks, but may eliminate natural clicks such as drum sticks tapping together or a conductor tapping the baton. Values less than 500% should be used on short selections only.

The filter requires a minimum of 4000 samples (about one tenth of a second at CD quality) to establish a baseline. Using the filter on a shorter selection has no effect.

See Also: , ,

Use Silence Reduction to automatically remove silences from a sound. Use it to save storage space or to remove long pauses in speeches or police/airport radio recordings. and are recommended before using this effect to ensure a consistent silence level.

Silence Reduction Settings
Setting Description
Silence threshold (dB) Sets the level for the silence. Any audio below this level is considered silence and is subject to removal, provided it has a long enough duration. The Duration specifies how long the silence must be before it is reduced. Any silences short than this remain unchanged.
Duration (s) Specifies how long the silence must be before it is reduced. Any silences shorter than this remain unchanged.
Reduce to (%) Sets the length of the reduced silence relative to its original length. A value of 75, for example, reduces a 10 second silence to 7.5 seconds.
Maximum length (s) Sets the maximum length for reduced silences. This setting overrides the Reduce to setting if the reduced length still exceeds the maximum. If this is set to 5 seconds in the above example, then the 10 second silence is reduced to 5 seconds.
Full crossfade If is checked, the ends of the audio where silences were removed are gradually crossfaded over the entire length the remaining silence. If unchecked, a short crossfade of one-tenth of a second is used to join together non-silent sections where silences were removed. If there is a high level of background noise that varies, then full crossfade is recommended. However, if you hear unexpected overlapping fragments after removing silences, then this option should not be used.

See Also: , , ,

Use Smoother to reduce hiss and crackle.

Smoother Setting
Setting Description
Length Sets the length of the smoother filter, which is the number of samples to be averaged. The larger the value, the more averaging is applied to the audio and the duller it will sound.
Volume Increases the of the processed audio to offset the the loss caused by averaging. Larger Length values require a higher volume.

See Also: , ,

A spectrum filter is a general purpose audio filter similar to , but with much greater control. Instead of using individual bands, the entire frequency spectrum is controlled using to adjust the . This allows many kinds of filters to be designed, such as lowpass, highpass, bandpass, bandstop, notch, peak, comb, and more. Filtering is performed in the frequency domain using Fast Fourier Transforms (FFTs).

A spectral analysis window displays a shape line and several other controls. The X and Y coordinates are updated when you click-and-drag a shape point. The X coordinate is the frequency in Hertz and the Y coordinate is the magnitude in decibels. The time of the spectral analysis shown is given in the Time box. Moving the time scroll bar, located below the analysis window, changes the graph to show the spectral analysis of a different part of the sound. Each channels is shown in a different colour.

Master gain sets the overall of the filter, which is equivalent to shifting the entire shape up or down.

Initially the shape line is horizontal at 0dB, which means that no changes in gain are made at any frequency. Alter the shape line up or down to increase or decrease the gain at a particular frequency. In technical terms, the shape line represents the frequency response function.

Spectrum Filter Settings
Group Setting Description
Graph range Min (dB) Sets the lower range of the y axis of the graph.
Max (dB) Sets the upper range of the y axis of the graph. Setting Min to -5 and Max to 5, for example, sets the graph to show a narrower range between -5dB to 5dB, allowing shape points to be set more precisely with the mouse within that range.
Settings Sets the detail of the spectral analysis and the resolution of the filter. Higher settings provide a higher resolution, allowing the filter to follow the shape more accurately, with sharper cutoffs. When processing high sampling rate files, such as 88kHz or 192kHz, the FFT size must be set higher for the filter to follow the shape. Using too high a value may cause overshoot and oscillations in the gain (
Sets the amount of audio data to overlap from one calculation to the next. The lowest value gives the fastest processing and generally works well.

See Also: , , ,