autoeq

AutoEq

AutoEq is a tool for equalizing headphone frequency responses automatically, and it achieves this by parsing frequency response measurements and producing equalization settings which correct the headphone to a neutral sound. AutoEq provides methods for reading data, equalizing it to a given target response and saving the results for usage with equalizers. It's possible to use different target curves, apply tilt for making the headphones brighter/darker and adding a bass boost. It's even possible to make one headphone sound (roughly) like another headphone. For more info about usage see Usage.

AutoEq Github page also serves as a database for headphone frequency response measurements, pre-computed results and has documentation about different equalizers and how the implementation works.

Updates

4.1.1

• Updated changelog

4.1.0

• Added gain_range option to optimize_fixed_band_eq() for limiting gain of each filter to maximum distance from the equalization target

4.0.0

BREAKING changes included!

• Several FrequencyResponse class methods renamed and arguments renamed or dropped
  • Improved duplicate frequency check
  • All kwargs in reset() default to False
  • Renamed read_from_csv() as read_csv() and write_to_csv() as write_csv()
  • Dropped Rockbox support (has been added to webapp)
  • Moved generate_frequencies() to utils.py
  • Moved _tilt() to utils.py as log_tilt()
  • Changed all mentions of "compensation" to "target"
  • Renamed smoothen_fractional_octave() as smoothen()
  • Removed iterations from smoothing
  • Moved _sigmoid() to utils.pyas log_f_sigmoid()
  • Moved log_log_gradient() to utils.py as log_log_gradient()
  • Renamed plot_graph() as plot
  • Changed plot() kwargs show and close to show_fig and close_fig, respectively
  • Changed plot colors to match webapp
  • Added missing default values to process() kwargs
• Removed some CLI arguments

3.0.1

Updated dependencies.

3.0.0

• Added --input-file, --max-slope and --sound-signature-smoothing-window-size parameters.
• Fixed crashing with non-standard sized compensation and measurement data.
• Fixed parametric eq optimizer producing filters outside of specified frequency range.
• Fixed parametric eq optimizer crashing without any free optimizeable parameters.
• Added defaul values for some args in parametric eq optimizer.
• Added more parametric eq optimizer configs.
• Introduced API breaking naming changes.

2.2.0

Added --preamp parameter.

2.1.1

Fixed README in PyPi package.

2.1.0

Fixed dependencies for Apple Silicon and added --treble-boost parameter.

2.0.0

Restructured the project and published in PyPi. Source code moved under autoeq directory and command line usage changed from python autoeq.py to python -m autoeq with underscores _ replaced with hyphens - in the parameter names.

Parametric eq optimizer reworked. The new optimizer supports shelf filters, has a powerful configuration system, run 10x faster, has limits for Fc, Q and gain value ranges and treats +10 kHz range as average value instead of trying to fix it precisely.

Installing

AutoEq requires Python 3 and should work with any decently recent version of Python 3.

pip install autoeq

You may need to install libsndfile

On Windows you may need to install Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, and 2019

Usage

AutoEq has command line interface in addition to Python methods. See python -m autoeq --help for arguments.

The full functionality with file input and output can be used with batch_processing:

from autoeq.batch_processing import batch_processing

batch_processing(
  input_dir='path/to/measurements', output_dir='path/to/results', new_only=False, standardize_input=False,
  compensation='path/to/target.csv', parametric_eq=True, fixed_band_eq=True,
  ten_band_eq=True, parametric_eq_config='8_PEAKING_WITH_SHELVES', fixed_band_eq_config='10_BAND_GRAPHIC_EQ',
  convolution_eq=True, fs=44100, bit_depth=16, phase='minimum', f_res=10, bass_boost_gain=6,
  bass_boost_fc=105, bass_boost_q=0.7, treble_boost_gain=0, treble_boost_fc=10000, treble_boost_q=0.7, tilt=None,
  sound_signature=None, max_gain=12, thread_count=0)

The main functionalities of AutoEq are in frequency_response which implements FrequencyResponse class. Parametric equalizer optimization and frequency response computations are implemented in peq.

from autoeq.frequency_response import FrequencyResponse
from autoeq.constants import PEQ_CONFIGS

harman_target = FrequencyResponse.read_csv('path/to/Harman over-ear 2018.csv')

fr = FrequencyResponse.read_csv('path/to/measurement.csv')
fr.interpolate()  # Creates standard logarithmic sampling when no argument is passed
fr.center()  # Centers the frequency response around 0 dB
fr.compensate(harman_target)  # Creates target and error data for the FR
fr.smoothen()  # Smoothens the FR data and error
fr.equalize(concha_interference=True)  # Creates equalization target
peqs = fr.optimize_parametric_eq(PEQ_CONFIGS['8_PEAKING_WITH_SHELVES'], 44100)
for filt in peqs[0].filters:
    print(f'{filt.gain:.2f} db, {filt.fc:.2f} Hz, {filt.q:.2f} Q')