SuperEQ

Equalization (British: equalisation, EQ for short) is the process of adjusting the balance between frequency components within an electronic signal. The most well known use of equalization is in sound recording and reproduction but there are many other applications in electronics and telecommunications. The circuit or equipment used to achieve equalization is called an equalizer. These devices boost or cut the energy of specific frequency bands. (source: wikipedia:Equalization)

SuperEQ (Shibatch Super Equalizer) is a graphic equalizer, originally created as a plugin for Winamp by Naoki Shibata. SuperEQ uses 16383th order FIR filters using FFT. Its equalization is very precise. Note that unlike traditional equalizers, there is very little roll-off or overlap between the bands.

SuperEQ(clip, string filename)

string  filename =

Set EQ bands from a preset file (see Preset Files below).

SuperEQ(clip, int band1 [, int band2, ..., int band18 ] )

int  band1, ..., band18 =

Set EQ bands within your script (see Band Gain Arguments below).

In both cases, the audio will be converted to float before processing.

The 18 bands cover the following frequency ranges:

Band #	f (low)	f (high)
1	0 Hz	65 Hz
2	65 Hz	93 Hz
3	93 Hz	131 Hz
4	131 Hz	185 Hz
5	185 Hz	262 Hz
6	262 Hz	370 Hz
7	370 Hz	523 Hz
8	523 Hz	740 Hz
9	740 Hz	1047 Hz
10	1047 Hz	1480 Hz
11	1480 Hz	2093 Hz
12	2093 Hz	2960 Hz
13	2960 Hz	4186 Hz
14	4186 Hz	5920 Hz
15	5920 Hz	8372 Hz
16	8372 Hz	11840 Hz
17	11840 Hz	16744 Hz
18	16744 Hz	22000±Hz

(adapted from avisynth/src/audio/supereq.h)

SuperEQ can accept a foobar2k equalizer Preset (.feq) file. Foobar's equalizer can be found in the DSPManager, and presets are adjusted and saved from there as well.

A Preset file is a text file with a positive or negative integer on each line. Nothing else is allowed on the line, except white space. Empty (blank) lines are ignored. Each line sets the gain in decibels for a frequency band. All text after the first 18 non-empty lines is ignored, allowing you to add comments to the file.

Preset files can be found in your \AviSynth25\Examples folder (Equalizer Presets.zip), available online here, or may be found by searching the Web for "foobar2000 presets."

Example: Using a Preset file

Example 1: Using a Preset file

SuperEq("C:\Equalizer Presets\Loudness.feq")

Contents of Loudness.feq:

4
4
4
2
-2
-2
-2
-2
-2
-2
-2
-4
-10
-7
0
3
4
4

*comment* 
(text after the first 18 non-empty lines ignored)

To give a very rough indication of the frequency response of the above example, here is a screenshot of an Audacity spectrum analysis of processed white noise (which should be approximately flat without equalization, allowing for low-frequency variability)

From Avisynth version 2.60, band gain arguments allow all 18 bands to be set within your script, instead of requiring a preset file. The values are specified in decibels (in integer format).

Example: Using band arguments

Example 2: Boost midrange (speech) using band arguments

SuperEQ(
\   0,  [*     0 -    65  *]
\   0,  [*    65 -    93  *]
\   0,  [*    93 -   131  *]
\   0,  [*   131 -   185  *]
\   0,  [*   185 -   262  *]
\   0,  [*   262 -   370  *]
\   0,  [*   370 -   523  *]
\   5,  [*   523 -   740  *]
\  10,  [*   740 -  1047  *]
\  10,  [*  1047 -  1480  *]
\  10,  [*  1480 -  2093  *]
\   5,  [*  2093 -  2960  *]
\   0,  [*  2960 -  4186  *]
\   0,  [*  4186 -  5920  *]
\   0,  [*  5920 -  8372  *]
\   0,  [*  8372 - 11840  *]
\   0,  [* 11840 - 16744  *]
\   0   [* 16744 - 22000± *]
\ )
Normalize ## avoid possible overload
ConvertAudioTo16bit 

To give a very rough indication of the frequency response of the above example, here is a screenshot of an Audacity spectrum analysis of processed white noise (which should be approximately flat without equalization, allowing for low-frequency variability)

Example: Audio Spectrum Analyzer

Example 3: Audio Spectrum Analyzer

This is a rather extreme example of what can be done with SuperEQ. It implements a spectrum analyzer by displaying the audio level in each of SuperEQ's 18 frequency bands, using Histogram("audiolevels"). See this doom9.org thread for details.

(shown being fed a 400 Hz tone)

This plugin is optimized for processors which have cache equal to or greater than 128k bytes (16383*2*sizeof(float) = 128k). This plugin won't work efficiently with K6 series processors.