Euvony
New Member
- Joined
- Mar 24, 2026
- Messages
- 1
- Likes
- 1
I developed a high-resolution AutoEQ optimizer designed specifically for Poweramp parametric EQ.
Unlike magnitude-only fitting approaches, this tool simulates real parametric biquad filters and jointly optimizes gain, center frequency, and Q using robust loss functions and perceptual weighting.
Main goals:
• improve realism of EQ correction
• allow fine control for narrow IEM resonances
• maintain smoother transitions and more natural tonal balance
Current features:
• multi-resolution optimization
• psychoacoustic weighting
• up to 64 parametric filters
• preset export ready for Poweramp
• GUI workflow (measurement → target → preset)
Tested on several IEM measurements with reasonably consistent results.
Repository:
github.com
I would appreciate technical feedback regarding:
• risk of overfitting with high filter counts
• phase and ringing behaviour
• optimization strategy improvements
Unlike magnitude-only fitting approaches, this tool simulates real parametric biquad filters and jointly optimizes gain, center frequency, and Q using robust loss functions and perceptual weighting.
Main goals:
• improve realism of EQ correction
• allow fine control for narrow IEM resonances
• maintain smoother transitions and more natural tonal balance
Current features:
• multi-resolution optimization
• psychoacoustic weighting
• up to 64 parametric filters
• preset export ready for Poweramp
• GUI workflow (measurement → target → preset)
Tested on several IEM measurements with reasonably consistent results.
Repository:
GitHub - DAPAAADF/AutoEQ-Parametric-64band-for-Poweramp: Psychoacoustic-aware 64-band parametric EQ optimizer for Poweramp. Takes stereo IEM measurements and a target curve, outputs a ready-to-use PEQ preset.
Psychoacoustic-aware 64-band parametric EQ optimizer for Poweramp. Takes stereo IEM measurements and a target curve, outputs a ready-to-use PEQ preset. - DAPAAADF/AutoEQ-Parametric-64band-for-Poweramp
I would appreciate technical feedback regarding:
• risk of overfitting with high filter counts
• phase and ringing behaviour
• optimization strategy improvements