pith. sign in

arxiv: 2605.17407 · v1 · pith:I4UNOHCNnew · submitted 2026-05-17 · 📡 eess.AS · cs.SY· eess.SP· eess.SY

Robust Soft-Constrained Spatially Selective Active Noise Control for Hearables Under Secondary Path Variations

Tong Xiao , Reinhild Roden , Matthias Blau , Simon Doclo This is my paper

Pith reviewed 2026-05-19 22:50 UTC · model grok-4.3

classification 📡 eess.AS cs.SYeess.SPeess.SY
keywords active noise controlhearablessecondary path variationrobust optimizationspatial selectivitysoft constraints
0
0 comments X

The pith

A single control filter optimized over multiple measured paths narrows performance variation in hearable noise control despite user and fit differences.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a robust optimization approach for spatially selective active noise control in hearables that must handle changes in the acoustic path from loudspeaker to inner microphone. It finds one fixed filter by minimizing the average cost across a collection of path estimates taken from human subjects and applies soft constraints to keep the solution practical. A reader should care because real devices encounter path variations that degrade noise reduction or risk instability, and the method yields a filter whose outcomes stay more consistent across those mismatches. Simulations plus real-time hardware tests confirm the average result drops only slightly while the range of outcomes tightens markedly. The work therefore supplies a usable design route when exact per-user calibration cannot be performed.

Core claim

The proposed robust soft-constrained optimization framework computes a single control filter by minimizing the average cost over a set of secondary path estimates derived from human measurements. Simulations and experiments on a real-time control platform show that the proposed approach slightly reduces mean performance relative to the matched case but substantially narrows the performance spread under secondary path mismatch. The proposed framework therefore provides a practical design strategy when accurate secondary path estimates are unavailable.

What carries the argument

The soft-constrained optimization that minimizes average cost across a finite set of measured secondary paths.

If this is right

  • Mean performance stays close to the perfectly matched case.
  • The range of outcomes under path mismatch becomes substantially narrower.
  • The resulting filter supports direct implementation on real-time hardware.
  • The method supplies a workable strategy precisely when accurate path estimates cannot be obtained.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach could allow hearables to be manufactured and sold without individual calibration for each user.
  • Averaging over measured variation sets might improve robustness in other adaptive acoustic systems that face changing paths.
  • Expanding the measurement collection to a larger and more varied population could increase coverage of actual real-world differences.

Load-bearing premise

The finite set of secondary path estimates obtained from human measurements sufficiently represents the variations that occur across users and device fits in real use.

What would settle it

Real-world trials with new users and unseen device placements in which the performance spread fails to narrow or the system loses stability would show the robustness benefit does not hold.

read the original abstract

Spatially selective active noise control (SSANC) hearables aim to attenuate noise from certain directions at the eardrum while preserving desired speech arriving from selected directions. Existing SSANC systems typically assume an accurate estimate of the secondary path from the loudspeaker to the inner error microphone. In practice, however, this path varies across users and device fits, which can degrade performance and compromise system stability. This paper proposes a robust soft-constrained optimization framework that computes a single control filter by minimizing the average cost over a set of secondary path estimates derived from human measurements. Simulations and experiments on a real-time control platform show that the proposed approach slightly reduces mean performance relative to the matched case but substantially narrows the performance spread under secondary path mismatch. The proposed framework therefore provides a practical design strategy when accurate secondary path estimates are unavailable.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a robust soft-constrained optimization framework for spatially selective active noise control (SSANC) in hearables. A single control filter is obtained by minimizing an average cost over a finite collection of secondary-path estimates measured from human subjects. Simulations and real-time experiments are reported to show that the resulting filter slightly lowers mean performance relative to a perfectly matched filter but substantially reduces the spread of performance under secondary-path mismatch.

Significance. If the measured path collection adequately represents real-world user and fit variations, the method offers a practical route to stable SSANC without per-user secondary-path calibration. This addresses a recognized deployment obstacle for hearable ANC systems and could improve robustness in consumer devices.

major comments (2)
  1. [Abstract / optimization framework] Abstract and the paragraph describing the optimization framework: the central robustness claim rests on the finite set of secondary-path estimates being representative of actual variations across users, ear geometries, and insertion depths, yet no information is supplied on the number of subjects, sampling strategy, or any coverage metric (e.g., principal-component span or worst-case deviation). Without this, it is impossible to judge whether the observed narrowing of performance spread generalizes beyond the training set.
  2. [Results] Results section (simulations and real-time experiments): the abstract states that the method 'substantially narrows the performance spread' but supplies no quantitative values, error bars, or statistical tests. The absence of these numbers prevents assessment of effect size and reproducibility.
minor comments (2)
  1. [Method] Notation for the soft-constraint weighting parameter and the averaging operator over the path set should be introduced explicitly with a single consistent symbol.
  2. [Experiments] Figure captions for the real-time platform should state the exact sampling rate, filter length, and number of independent runs used to compute the reported spread.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major point below and describe the revisions we intend to make.

read point-by-point responses

  1. Referee: [Abstract / optimization framework] Abstract and the paragraph describing the optimization framework: the central robustness claim rests on the finite set of secondary-path estimates being representative of actual variations across users, ear geometries, and insertion depths, yet no information is supplied on the number of subjects, sampling strategy, or any coverage metric (e.g., principal-component span or worst-case deviation). Without this, it is impossible to judge whether the observed narrowing of performance spread generalizes beyond the training set.

    Authors: We agree that explicit details on the secondary-path collection are needed to support the robustness claim. In the revised manuscript we will expand the description of the measurement set to state the number of human subjects, the protocol used to sample variations in ear geometry and insertion depth, and a coverage assessment based on the principal-component span of the collected paths together with the maximum deviation from the mean path. These additions will allow readers to evaluate representativeness directly. revision: yes

  2. Referee: [Results] Results section (simulations and real-time experiments): the abstract states that the method 'substantially narrows the performance spread' but supplies no quantitative values, error bars, or statistical tests. The absence of these numbers prevents assessment of effect size and reproducibility.

    Authors: We accept that the current presentation lacks the quantitative support required for assessing effect size. We will revise the abstract to report concrete figures (e.g., the factor by which the standard deviation of the performance metric is reduced) and will augment the results section with error bars on the relevant plots together with statistical tests (variance-ratio tests) comparing the spread under the proposed filter versus the nominal matched filter. These changes will make the magnitude and statistical significance of the improvement explicit. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected in derivation chain

full rationale

The paper defines a soft-constrained optimization that computes one control filter by minimizing average cost over a finite collection of secondary-path estimates obtained from human measurements. Performance is then assessed via separate simulations and real-time experiments that measure mean degradation and spread under mismatch. This structure does not reduce any reported result to its inputs by algebraic identity or by renaming a fitted quantity as a prediction; the objective and the reported robustness metrics remain distinct. No load-bearing self-citation, imported uniqueness theorem, or ansatz smuggling is present in the provided description. The central claim therefore rests on an empirical assumption about set representativeness rather than on a self-referential derivation.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Only the abstract is available; therefore the ledger is populated at the level of generality stated in the abstract. No explicit free parameters, axioms, or invented entities are named beyond the standard assumption that the collected secondary-path set is representative.

axioms (1)
  • domain assumption The finite collection of secondary-path measurements from human subjects adequately spans the range of variations encountered in practice.
    Invoked when the optimization is defined over that set (abstract description of the framework).

pith-pipeline@v0.9.0 · 5688 in / 1329 out tokens · 42063 ms · 2026-05-19T22:50:57.441266+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages · 1 internal anchor

  1. [1]

    Conventional ANC hearables treat all incoming sounds as noise [4, 5]

    INTRODUCTION Active noise control (ANC) hearables use secondary sources to gen- erate anti-noise to minimize the noise leakage at the eardrum [1–3]. Conventional ANC hearables treat all incoming sounds as noise [4, 5]. This becomes problematic in complex acoustic environments when desired speech is present [6–11]. In these cases, the user may want to focu...

    work page

  2. [2]

    Robust Soft-Constrained Spatially Selective Active Noise Control for Hearables Under Secondary Path Variations

    SIGNAL MODEL As shown in Fig. 1, we consider a hearable withK outer microphones. Without loss of generality, we consider one loudspeaker as the sec- ondary source and one inner error microphone, resulting in a total of K+ 1 microphones. We assume that the acoustic feedback paths between the loudspeaker and the outer microphones are known, such that acoust...

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  3. [3]

    A soft- constrained optimization has been proposed to balance the noise reduc- tion and the speech distortion [17]

    SSANC FORMULA TION AND EV ALUA TION CASES The objective of the SSANC system is to minimize the power of the inner error microphone signal while preserving the delayed desired speech component of an outer reference microphone signal. A soft- constrained optimization has been proposed to balance the noise reduc- tion and the speech distortion [17]. The cost...

    work page

  4. [4]

    0| {z } ∆ 1 0. . .0| {z } Lh+L−1−∆ ]T ∈R La+Lh+L−1,(11c) where Hk is the convolution matrix of the ReIR for thek-th channel with respect to a chosen reference microphone, with La and Lh denoting the length of the anti-causal and causal parts of the ReIR, respectively [16]. α is a real-valued positive amplification factor of the desired speech signal. It s...

    work page

  5. [5]

    To ensure consistency across a diverse set of conditions, we propose a robust optimization framework that minimizes the average cost over a set of secondary path estimates

    CASE 3: ROBUST OPTIMIZA TION Significant secondary path variations can lead to performance degrada- tion when the control filter is optimized for a single nominal secondary path. To ensure consistency across a diverse set of conditions, we propose a robust optimization framework that minimizes the average cost over a set of secondary path estimates. The o...

    work page

  6. [6]

    Setup For the evaluation, we considered a pair of closed-fitting hearables inserted into both ears of a GRAS 45BB-12 KEMAR Head & Torso simulator, as illustrated in Fig

    EV ALUA TION 5.1. Setup For the evaluation, we considered a pair of closed-fitting hearables inserted into both ears of a GRAS 45BB-12 KEMAR Head & Torso simulator, as illustrated in Fig. 2a. We used four outer microphones (entrance microphones and concha microphones at the left and right ears, labeled as #1–#4), two inner error microphones (located at th...

    work page

  7. [7]

    Simulation results showed that the matched case, assuming or- acle knowledge of the secondary path, achieves the best overall perfor- mance

    CONCLUSION This paper examined the impact of secondary path variations on the performance of spatially selective active noise control systems for hear- ables. Simulation results showed that the matched case, assuming or- acle knowledge of the secondary path, achieves the best overall perfor- mance. However, when the estimated secondary path does not match...

    work page

  8. [8]

    S. M. Kuo and D. R. Morgan,Active noise control systems: Algorithms and DSP implementations. Wiley, 1996

    work page 1996

  9. [9]

    S. J. Elliott,Signal processing for active control. Academic Press, 2000

    work page 2000

  10. [10]

    Hansen, S

    C. Hansen, S. Snyder, X. Qiu, L. Brooks, and D. Moreau,Active control of noise and vibration, 2nd ed. CRC Press, Nov. 2012

    work page 2012

  11. [11]

    Optimization of a fixed virtual sensing feedback ANC controller for in-ear headphones with multiple loudspeakers,

    P. R. Benois, R. Roden, M. Blau, and S. Doclo, “Optimization of a fixed virtual sensing feedback ANC controller for in-ear headphones with multiple loudspeakers,” inProc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Singapore, 2022, pp. 8717–8721

    work page 2022

  12. [12]

    Data-driven uncertainty modeling for robust feedback active noise control in headphones,

    F. Hilgemann, E. Chatzimoustafa, and P. Jax, “Data-driven uncertainty modeling for robust feedback active noise control in headphones,”Journal of the Audio Engineering Society, vol. 72, no. 12, pp. 873–883, Apr 2024

    work page 2024

  13. [13]

    Listening in a noisy environment: Integration of active noise control in audio products,

    C.-Y . Chang, A. Siswanto, C.-Y . Ho, T.-K. Yeh, Y .-R. Chen, and S. M. Kuo, “Listening in a noisy environment: Integration of active noise control in audio products,”IEEE Consumer Electronics Magazine, vol. 5, no. 4, pp. 34–43, 2016

    work page 2016

  14. [14]

    Augmented/mixed reality audio for hearables: Sensing, control, and rendering,

    R. Gupta, J. He, R. Ranjan, W.-S. Gan, F. Klein, C. Schnei- derwind, A. Neidhardt, K. Brandenburg, and V . V ¨alim¨aki, “Augmented/mixed reality audio for hearables: Sensing, control, and rendering,”IEEE Signal Processing Magazine, vol. 39, no. 3, pp. 63–89, 2022

    work page 2022

  15. [15]

    Integrated active noise control and noise reduction in hearing aids,

    R. Serizel, M. Moonen, J. Wouters, and S. H. Jensen, “Integrated active noise control and noise reduction in hearing aids,”IEEE Transactions on Audio, Speech, and Language Processing, vol. 18, no. 6, pp. 1137–1146, 2010

    work page 2010

  16. [16]

    Influence of secondary path estimation errors on the performance of ANC-motivated noise reduction algorithms for hearing aids,

    D. Dalga and S. Doclo, “Influence of secondary path estimation errors on the performance of ANC-motivated noise reduction algorithms for hearing aids,” inProc. IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), New Paltz, USA, 2013, pp. 1–4

    work page 2013

  17. [17]

    Design and implementation of an active noise control headphone with directional hear-through capability,

    V . Patel, J. Cheer, and S. Fontana, “Design and implementation of an active noise control headphone with directional hear-through capability,”IEEE Transactions on Consumer Electronics, vol. 66, no. 1, pp. 32–40, Feb. 2020

    work page 2020

  18. [18]

    Spatially selective active noise control systems,

    T. Xiao, B. Xu, and C. Zhao, “Spatially selective active noise control systems,”The Journal of the Acoustical Society of America, vol. 153, no. 5, pp. 2733–2744, May 2023

    work page 2023

  19. [19]

    Beamforming: A versatile approach to spatial filtering,

    B. Van Veen and K. Buckley, “Beamforming: A versatile approach to spatial filtering,”IEEE ASSP Magazine, vol. 5, no. 2, pp. 4–24, 1988

    work page 1988

  20. [20]

    A consolidated perspective on multimicrophone speech enhancement and source separation,

    S. Gannot, E. Vincent, S. Markovich-Golan, and A. Ozerov, “A consolidated perspective on multimicrophone speech enhancement and source separation,”IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 25, no. 4, pp. 692–730, 2017

    work page 2017

  21. [21]

    Multichannel signal enhancement algorithms for assisted listening devices: Exploiting spatial diversity using multiple microphones,

    S. Doclo, W. Kellermann, S. Makino, and S. E. Nordholm, “Multichannel signal enhancement algorithms for assisted listening devices: Exploiting spatial diversity using multiple microphones,”IEEE Signal Processing Magazine, vol. 32, no. 2, pp. 18–30, Mar. 2015

    work page 2015

  22. [22]

    Effect of target signals and delays on spatially selective active noise control for open-fitting hearables,

    T. Xiao and S. Doclo, “Effect of target signals and delays on spatially selective active noise control for open-fitting hearables,” inProc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Seoul, Republic of Korea, 2024, pp. 1056–1060

    work page 2024

  23. [23]

    Spatially selective active noise control for open-fitting hearables with acausal optimization,

    ——, “Spatially selective active noise control for open-fitting hearables with acausal optimization,” inProc. F orum Acusticum Euronoise 2025, M´alaga, Spain, Jun. 2025, pp. 117–124

    work page 2025

  24. [24]

    Soft-constrained spatially selective active noise control for open-fitting hearables,

    T. Xiao, R. Roden, M. Blau, and S. Doclo, “Soft-constrained spatially selective active noise control for open-fitting hearables,” inProc. IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), Tahoe City, USA, 2025, pp. 1–5

    work page 2025

  25. [25]

    Robust single-and multi-loudspeaker least-squares-based equalization for hearing devices,

    H. Schepker, F. Denk, B. Kollmeier, and S. Doclo, “Robust single-and multi-loudspeaker least-squares-based equalization for hearing devices,”EURASIP Journal on Audio, Speech, and Music Processing, vol. 2022, no. 1, pp. 1–14, 2022

    work page 2022

  26. [26]

    Cstr vctk corpus: English multi-speaker corpus for cstr voice cloning toolkit, 2019

    C. Veaux, J. Yamagishi, and K. MacDonald, “CSTR VCTK corpus: English multi-speaker corpus for CSTR voice cloning toolkit,” 2017. [Online]. Available: https://doi.org/10.7488/ds/2645

    work page doi:10.7488/ds/2645 2017

  27. [27]

    Sound sample 07025055,

    British Broadcasting Corporation, “Sound sample 07025055,” BBC Sound Effects Archive, 2024, accessed: March 04, 2026. [Online]. Available: https://sound-effects.bbcrewind.co.uk/search?q=07025055

    work page 2024

  28. [28]

    A toolbox for rendering virtual acoustic environments in the context of audiology,

    G. Grimm, J. Luberadzka, and V . Hohmann, “A toolbox for rendering virtual acoustic environments in the context of audiology,”Acta acustica united with acustica, vol. 105, no. 3, pp. 566–578, 2019

    work page 2019

  29. [29]

    Pub environment,

    G. Grimm, M. Hendrikse, and V . Hohmann, “Pub environment,” Sep. 2021. [Online]. Available: https://doi.org/10.5281/zenodo.5886987

    work page doi:10.5281/zenodo.5886987 2021

  30. [30]

    A one-size-fits-all ear- piece with multiple microphones and drivers for hearing device research,

    F. Denk, M. Lettau, H. Schepker, S. Doclo, R. Roden, M. Blau, J.- H. Bach, J. Wellmann, and B. Kollmeier, “A one-size-fits-all ear- piece with multiple microphones and drivers for hearing device research,” inProc. AES International Conference on Headphone Technology, San Francisco, USA, Aug. 2019, pp. 1–9

    work page 2019

  31. [31]

    The hearpiece database of individual transfer functions of an in-the-ear earpiece for hearing device research,

    F. Denk and B. Kollmeier, “The hearpiece database of individual transfer functions of an in-the-ear earpiece for hearing device research,”Acta Acustica, vol. 5, no. 2, pp. 1–16, 2021

    work page 2021

  32. [32]

    Spatially pre-processed speech distortion weighted multi-channel Wiener filtering for noise reduction,

    A. Spriet, M. Moonen, and J. Wouters, “Spatially pre-processed speech distortion weighted multi-channel Wiener filtering for noise reduction,”Signal Processing, vol. 84, no. 12, pp. 2367–2387, 2004

    work page 2004

  33. [33]

    Frequency- domain criterion for the speech distortion weighted multichannel Wiener filter for robust noise reduction,

    S. Doclo, A. Spriet, J. Wouters, and M. Moonen, “Frequency- domain criterion for the speech distortion weighted multichannel Wiener filter for robust noise reduction,”Speech Communication, vol. 49, no. 7, pp. 636–656, 2007

    work page 2007

  34. [34]

    Methods for Calculation of the Speech Intelligibility Index,

    Acoustical Society of America (ASA), “Methods for Calculation of the Speech Intelligibility Index,” American National Stan- dards Institute (ANSI), ANSI/ASA S3.5-1997 Standard, 1997

    work page 1997

  35. [35]

    Perceptual evaluation of speech quality (PESQ)–a new method for speech quality assessment of telephone networks and codecs,

    A. Rix, J. Beerends, M. Hollier, and A. Hekstra, “Perceptual evaluation of speech quality (PESQ)–a new method for speech quality assessment of telephone networks and codecs,” inProc. IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings (ICASSP), Salt Lake City, USA, May 2001, pp. 749–752

    work page 2001

  36. [36]

    An algorithm for predicting the intelligibility of speech masked by modulated noise maskers,

    J. Jensen and C. H. Taal, “An algorithm for predicting the intelligibility of speech masked by modulated noise maskers,” IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 24, no. 11, pp. 2009–2022, 2016

    work page 2009