Towards detecting the pathological subharmonic voicing with fully convolutional neural networks

Andrew J. McWhorter; Brad Story; Melda Kunduk; Takeshi Ikuma

REVIEW 3 major objections 4 minor 48 references

Towards detecting the pathological subharmonic voicing with fully convolutional neural networks

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that fully convolutional networks trained on a synthetic corpus of subharmonic voices can assign each 2-ms snapshot a subharmonic period in {1,2,3,4} with over 98% accuracy on held-out synthetic signals, with case studies…

desk verdict A genuinely new FCN approach to subharmonic-period classification with an honest in-distribution synthetic evaluation; real-voice evidence is qualitative, but the paper deserves a serious referee. read the letter →

arxiv 2501.09159 v1 pith:U2JIMB5O submitted 2025-01-15 eess.AS cs.SD

Takeshi Ikuma , Melda Kunduk , Brad Story , Andrew J. McWhorter This is my paper · ORCID

classification eess.AScs.SD

keywords subharmonicphonationvoicedisordersfullyconvolutionalnetworkperiodclassificationsyntheticcorpuskinematicvocalfoldmodelsustainedvowelanalysisacoustic

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper seeks a reliable way to detect subharmonic phonation in voice recordings, where the vocal folds repeat a vibration pattern every two, three, or four glottal cycles instead of every cycle. Existing pitch and roughness measures either miss subharmonics or only handle period-doubling, and subjective spectrogram inspection remains the standard for more complex cases. The proposal is to train fully convolutional neural networks on a large Monte Carlo corpus of synthesized subharmonic vowels so that each network labels short snapshots with the subharmonic period M in {1,2,3,4}. On held-out synthetic signals the longer-window network FCN-785 reaches 98.9% overall accuracy, and every one of the four classes is above 98.5% conditional accuracy. On three real pathological /a/ recordings the networks track period-tripling modulation, partially detect locked biphonation, and mislabel clean segments under severe fo tremor, which the paper reads as evidence that the approach works but the synthetic training distribution needs more realism.

What carries the argument

The load-bearing object is the synthesis-driven FCN pair. On the synthesis side, the kinematic vocal fold model defines fold displacement from a reference vibration r(t; phi), and subharmonic behavior is realized by amplitude and frequency modulation of that reference, so the modulation passes through vocal-fold collision, nonlinear glottal flow, and vocal-tract acoustics rather than being superimposed on the output waveform. On the classifier side, a five-layer fully convolutional network with four max-pooling layers reduces the output rate by a factor of 8 to emit one subharmonic-period estimate per 2-ms snapshot, using sigmoid outputs instead of softmax because the label set is not exhaustive of pathological voices. The training corpus randomizes 21 synthesis parameters, including fo, modulation extents and phases, glottal geometry, tract lengths and areas, lung pressure, and noise level, and the two networks share the same coefficient count while differing in filter length and count at the fourth convolutional layer.

What would settle it

If a laryngoscopy-verified clinical corpus showed that FCN-785 mislabels clean harmonic segments under fo tremor at rates no better than chance while synthetic accuracy stays near 99%, the transfer assumption would be refuted; the paper's own tremor case already points in this direction but with only a single recording.

Watch

Extended reading notes

Core claim

The central claim is that subharmonic period estimation can be cast as a per-snapshot classification problem solved by a fully convolutional network operating on raw, mean-variance-normalized 8 kHz audio, and that a network trained exclusively on synthesized modulation-type subharmonics can classify M=1 through M=4 with over 98% accuracy on held-out synthetic signals and can partially transfer to real clinical recordings. The paper establishes this by generating training data with a kinematic vocal-fold model coupled to a wave-propagation vocal tract, introducing subharmonics through amplitude and frequency modulation of the fold reference vibration, and training two FCN variants: FCN-401 with a 50.1-ms window and FCN-785 with a 98.1-ms window. The longer window performs better on low-fo and low-SHR signals, while the shorter window tracks transitions between subharmonic states slightly better; both ignore weak unlocked modulation, but both mislabel clean harmonic segments in a severe vocal-tremor case because the training corpus fixes fo per signal.

Load-bearing premise

The load-bearing premise is that the 21-parameter Monte Carlo corpus of symmetric-fold, modulation-type subharmonic sustained vowels, each with a fixed fo and a steady state, is representative enough of real pathological subharmonic phonation that the high synthetic accuracy transfers to clinical recordings.

Editorial extensions

If this is right

If the synthetic accuracy transfers to clinical use, subharmonic detection can be automated with a single feed-forward network, giving clinicians a numerical stream of M labels over time instead of requiring subjective spectrogram reading.
The M labels can feed downstream measures such as NSH and SHR and can steer fundamental-frequency estimators away from the subharmonic period, reducing the underestimation of pathology severity that the introduction identifies.
The 98.1% versus 98.9% comparison indicates that the 98-ms window should be preferred for sustained vowels, while the 50-ms window is preferable when transition timing matters, so the two networks cover complementary use cases.
Training on fixed-fo sustained vowels is a known source of failure, as the tremor case shows clean harmonic stretches being mislabeled, so extending the corpus with fo variation, intermittency, and biphonation is the paper's own predicted path to a dependable detector.
Weak unlocked modulation is mostly ignored by both networks, which is desirable if the goal is to report only strongly locked subharmonic periods rather than every cycle-to-cycle irregularity.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A testable extension that the paper does not run is to add slow sinusoidal fo drift to the synthetic corpus; if tremor-case mislabeling disappears while synthetic accuracy stays above 98%, fixed-fo training would be identified as the cause rather than the FCN architecture.
Because the output layer uses sigmoids rather than softmax, the per-class probabilities can be read as confidence values, so a clinical system could flag low-confidence snapshots for human review, a use the paper does not develop.
The same FCN pipeline is input-duration agnostic, so a straightforward extrapolation is to evaluate it on connected speech once the training corpus includes fo variation; the paper only demonstrates sustained vowels, so this is an extension beyond its reported results.
If a larger clinical dataset with simultaneous high-speed videoendoscopy ground truth were available, one could quantify the synthetic-to-real transfer gap in terms of conditional per-M accuracy instead of relying on three case studies.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Quick take: this is a solid proof-of-concept for FCN-based subharmonic-period classification, with an honest synthetic evaluation. The 98.9% number is real but measures learning of the paper's own generator; the real-voice support is qualitative. That is a limitation, but the paper says so itself. I'd send it to peer review.

What's actually new: the authors are the first to frame subharmonic period detection as a four-class FCN classification problem (M=1..4), and they embed AM/FM subharmonic modulation into a kinematic vocal fold model rather than applying cycle-wise rescaling post-hoc. That is a meaningful step beyond prior period-doubling-only synthesis. The synthetic evaluation is well done: 4,000 held-out signals, separate confusion matrices, and a credible analysis of why the shorter-window FCN struggles on M=2 (SHR imbalance and spectral resolution). The two-architecture comparison is useful.

Soft spots, in order of size. First, the central accuracy figure is in-distribution: training and test draw from the same 21-parameter generator and the same uniform ranges. The generator excludes biphonation, intermittent locking, and fo variation. Those are not edge cases for the intended use; the paper's own case studies hit all three. The tremor case is the sharpest evidence: FCN-785 never outputs M=1 even in clean harmonic segments, which the authors attribute to fixed-fo training. So the 98.9% does not transfer as-is to real voices. The paper is transparent about this in the conclusions, but the abstract's 'encouraging outcomes' rests on three qualitative recordings with no ground truth. Second, there is no code or data release, which makes the result harder to build on. The synthesis model is described in detail, so reimplementation is possible, but artifacts would help. Third, the biphonation case is interesting but speculative; the networks were not trained for it, and the 'suspected' locking is interpreted generously.

None of this is fatal. The claims are appropriately scoped, the synthesis is more realistic than prior work, and the failure analysis is honest. This is a conditional accept: it needs external validation on real pathological voices or a clearly stated in-distribution claim, and ideally artifacts.

Who should read it: anyone working on acoustic analysis of type-2 voices, or on DL for pathological speech. I'd bring it to our reading group, and I'd cite it as the current FCN baseline for subharmonic-period classification. Yes to peer review; the correct outcome is likely a revision with more real-voice data or a narrower title.

Referee Report

3 major / 4 minor

Summary. The paper proposes two fully convolutional neural networks, FCN-401 and FCN-785, to classify the subharmonic period M (with M in {1,2,3,4}) of voice signals, where M=1 denotes normal phonation. The networks are trained exclusively on synthetic signals generated from a kinematic vocal fold model with subharmonic amplitude and frequency modulation, using random draws over the 21 synthesis parameters in Table I. Evaluation on held-out synthetic signals reports 98.1% and 98.9% overall accuracy for FCN-401 and FCN-785, respectively, with an analysis of performance versus subharmonic-to-harmonic ratio and fundamental frequency. The paper then presents three qualitative case studies on sustained /a/ recordings from a disordered voice database, showing encouraging but mixed results, including the failure of both networks on a voice with severe vocal tremor.

Significance. If the synthetic results were to transfer to clinical recordings, the proposed approach would fill a real gap in acoustic voice analysis: automatic and reliable detection of subharmonic phonation, which is known to degrade fundamental-frequency estimators and voice parameter measurements. The paper has notable strengths: the use of a physiologically motivated kinematic vocal fold model rather than simple AM/FM-cycle manipulation, a clearly described and reproducible Monte Carlo training procedure, and a credible analysis of the FCN-401 M=2 weakness in terms of SHR imbalance and window size. The central claim, however, is conditional: the 98% figure is an in-distribution result on the same generative model and parameter ranges used for training, and the real-voice evidence is qualitative and limited. The paper is transparent about its limitations, but the abstract and title do not carry that conditionality.

major comments (3)

[IV and III (Table I), Abstract] The reported classification accuracies (98.1% for FCN-401 and 98.9% for FCN-785) are measured on held-out signals drawn from the same synthesis model and the same uniform parameter ranges used for training. This is an in-distribution evaluation: it demonstrates that the networks learn the specific generator distribution, not that they detect pathological subharmonic voicing in general. The abstract's over 98% classification accuracy is presented without this caveat. The paper should either reframe the central claim as applying only to the synthetic distribution, or add out-of-distribution tests (e.g., parameter ranges outside Table I, biphonation, fo variability) that would support broader generalization.
[V, Fig. 8] The third case study (severe vocal tremor) directly contradicts the practical claim of encouraging outcomes at the level of segment classification: FCN-785 never outputs M=1 even in clean harmonic segments (t=0.05-0.13 s and 0.77-0.84 s), and FCN-401 also fails to hold M=1 for the entirety of these segments. The authors attribute this to training with fixed fo, which is a reasonable hypothesis, but the failure is not quantified and no ground truth is available for the real recordings. These qualitative cases cannot by themselves establish clinical utility. The paper should either provide quantitative evaluation on a larger real dataset with annotation, or explicitly limit the practical claims to a feasibility demonstration.
[III and VI] The training dataset contains only symmetric-fold subharmonic modulation with fixed fo per signal, uniform independent parameter draws, and steady-state sustained vowels. The paper's own conclusions list biphonation, intermittency, and fo variation as necessary additions. This is not merely a future-work item: the current synthetic evaluation omits exactly the phenomena that the three case studies encounter (unlocked modulation, suspected biphonation, fo tremor). Consequently, the synthetic accuracy figure cannot be interpreted as a measure of performance on the clinically relevant population of subharmonic voices, and the paper's central claim needs to be scoped accordingly.

minor comments (4)

[Throughout] There are numerous typographical errors, for example JANURAY in the header, wholistic in the abstract, trainig in Section IV, Sythetic in the Fig. 3 caption, acuracy in Section IV, tranditional in the text before Eq. (23), radition in the lip radiation description, simultor in Section III, constititute in Section V, and entirity in Section V. These should be corrected before submission.
[II, Fig. 1] The description of the network variants is difficult to parse: different allocations of its coefficients for the fourth convolution layer does not clearly explain how the same architecture yields different window sizes. A short explanation of how the receptive field is controlled would improve reproducibility.
[III, Eq. (6)] The notation r_M(t; phi) is used with subscript M, but the definition of r(t; phi) in Eq. (8) is clear only after reading both equations; consider writing r_M(t; phi) explicitly in the modulation definition and clarifying that M is the subharmonic period of the reference vibration.
[IV, Fig. 4] The SHR analysis is informative, but the paper does not report the SHR computation parameters (e.g., whether the Hamming window is applied to the whole 1-s signal or to analysis frames, and how K_s and K_h are defined precisely). Adding these details would strengthen the reproducibility of the analysis.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the 98.9% synthetic accuracy is a genuine held-out evaluation on a separate test set from the same generator, and the real-voice case studies are external, unfitted data; the paper explicitly flags remaining distributional limitations.

full rationale

The paper's central claim is an empirical, not analytical, result: FCN-401/FCN-785 are trained on 32,000 synthesized signals and evaluated on 4,000 separately generated signals (Section IV: 'another 4000 signals were generated'). The held-out test signals share the Section III/Table I generator, so the 98.9% figure is an in-distribution generalization result, but it is not a fitted parameter renamed as a prediction; the test labels M are not used in training, and there is no equation in the paper that defines the classifier output in terms of the training label or vice versa. Real sustained-vowel recordings from the KayPENTAX database (Section V) are external to training and are not fine-tuned, providing a limited but real transfer check; the paper explicitly notes that biphonation, intermittency, and fo variation are not in the synthetic corpus (Section VI), so it does not overclaim clinical generality. Self-citations [5], [12], [13] support prevalence/motivation statements but are not load-bearing; no uniqueness theorem or ansatz is imported from prior work. Therefore no circular step can be quoted, and the appropriate score is 0.

Assumptions & free parameters 21 free parameters · 4 assumptions · 0 invented entities

The central result rests on the synthetic training distribution. The 21 synthesis parameters are hand-chosen and not fitted to clinical data; their ranges shape every accuracy number. The synthesis model's realism is a domain assumption, and the paper itself flags missing biphonation, intermittency, and fo variability as gaps.

free parameters (21)

fo = [100, 300) Hz
Speaking fundamental frequency; hand-picked uniform range in Table I.
epsilon_AM = [0.1, 1.0) log-uniform
AM extent in Eq. (6); hand-picked range shapes subharmonic depth.
epsilon_FM = [0.005, 0.1) log-uniform
FM extent in Eq. (6); hand-picked range shapes frequency modulation.
phi_AM = [-pi/2, pi/2)
AM phase; uniform draw in Table I.
phi_FM = [-pi/2, pi/2)
FM phase; uniform draw in Table I.
S_nu(0) = [100, 2500) (cm^3/s)^2/Hz
DC PSD level of aspiration noise; hand-picked range.
delta = [0.2, 0.6)
Noise suppression factor during subcritical phase; hand-picked range.
L = [0.738, 1.562) cm
Vibrating vocal fold length; hand-picked range.
T = [0.18, 0.33) cm
Vibrating vocal fold thickness; hand-picked range.
xi_m = [0.09, 0.132) cm
Maximum glottal half width; hand-picked range.
Qa = [0.27, 0.33)
Abduction quotient; hand-picked range.
Qs = [1.8, 2.2)
Shape quotient; hand-picked range.
Qb = [0.45 Qs, 0.55 Qs)
Bulging quotient relative to Qs; hand-picked multiplicative range.
Qp = [0.18, 0.22)
Phase quotient; hand-picked range.
Rzn = [0.63, 0.77)
Nodal point ratio; hand-picked range.
alpha = [0.9983, 0.9985) 1/cm
Propagation gain per unit length; hand-picked range.
L_T_supra = [11.111, 15.873) cm
Supraglottal tract length; hand-picked range.
Ae = [1.0, 5.0) cm^2
Supraglottal tract area; hand-picked range.
L_T_sub = [6.349, 9.524) cm
Subglottal tract length; hand-picked range.
As = [1.0, 3.0) cm^2
Subglottal tract area; hand-picked range.
PL = [7056, 8624) dyn/cm^2
Lung pressure; hand-picked range.

assumptions (4)

domain assumption The kinematic vocal fold and two-port vocal tract synthesis model produces acoustic subharmonic signals that match the structure of real pathological phonation.
Section III uses this model exclusively for training and evaluation; Section V calls it 'crude' and the Conclusions list missing behaviors (biphonation, intermittency, fo variation).
ad hoc to paper Uniform independent draws over the Table I ranges cover the clinically relevant diversity of subharmonic voices.
No clinical distribution data are used; the paper says a 'better stochastical representation of human voice production system is crucial'.
domain assumption A snapshot must contain more than M+1 glottal cycles and M only needs to be from {1,2,3,4}.
Section II restricts M to 4 based on window size and fo 100-300 Hz, while acknowledging higher-period subharmonics exist in pathology.
domain assumption Subharmonic phonation can be represented as AM/FM modulation of the single-source reference vibration r_M(t;phi).
Section III generates only modulation with symmetric folds; real case 2 contains biphonation, which is outside this model.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Towards detecting the pathological subharmonic voicing with fully convolutional neural networks." pith.science (2026). https://pith.science/paper/U2JIMB5O

@misc{pith2026250109159,
  author       = {Pith},
  title        = {Pith review of: Towards detecting the pathological subharmonic voicing with fully convolutional neural networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U2JIMB5O}},
  note         = {Machine review of arXiv:2501.09159}
}

read the original abstract

Many voice disorders induce subharmonic phonation, but voice signal analysis is currently lacking a technique to detect the presence of subharmonics reliably. Distinguishing subharmonic phonation from normal phonation is a challenging task as both are nearly periodic phenomena. Subharmonic phonation adds cyclical variations to the normal glottal cycles. Hence, the estimation of subharmonic period requires a wholistic analysis of the signals. Deep learning is an effective solution to this type of complex problem. This paper describes fully convolutional neural networks which are trained with synthesized subharmonic voice signals to classify the subharmonic periods. Synthetic evaluation shows over 98% classification accuracy, and assessment of sustained vowel recordings demonstrates encouraging outcomes as well as the areas for future improvements.

Figures

Figures reproduced from arXiv: 2501.09159 by the authors.

**Figure 2.** Block diagram of the transmission-line voice synthesis [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Sythetic classification confusion matrices: (a) FCN [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 5.** Figure 5: Synthetic classification performance vs. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

**Figure 4.** Figure 4: Synthetic classification performance vs. SHR ( [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 6.** Figure 6: Case Study 1 (modulation): Spectrogram and outputs of [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

**Figure 8.** Figure 8: Case Study 3 (severe vocal tremor): Spectrogram and [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 48 canonical work pages

[1]

Microphone and electroglottographic data from dyspho- nic patients: Type 1, 2 and 3 signals,

A. Behrman, C. J. Agresti, E. Blumstein, and N. Lee, “Microphone and electroglottographic data from dyspho- nic patients: Type 1, 2 and 3 signals,” J. V oice, vol. 12, no. 2, pp. 249–260, Jan. 1998

work page 1998
[2]

Diplophonia reappraised,

L. Cavalli and A. Hirson, “Diplophonia reappraised,” J. V oice, vol. 13, no. 4, pp. 542–556, 1999

work page 1999
[3]

[Spectrographic study of voice disorders: subharmonics],

F. Núñez Batalla, C. Suárez Nieto, C. Muñoz Pinto, L. Baragaño Río, M. J. Alvarez Zapico, and A. Martínez Ferreras, “[Spectrographic study of voice disorders: subharmonics],” Acta Otorrinolaringol. Esp., vol. 51, no. 1, pp. 52–56, 2000

work page 2000
[4]

A study of subharmonics in connected speech material,

E. Kramer, R. Linder, and R. Schönweiler, “A study of subharmonics in connected speech material,” J. V oice, vol. 27, no. 1, pp. 29–38, Jan. 2013

work page 2013
[5]

Investigation of vocal bifurcations and voice patterns induced by asymmetry of pathological vocal folds,

T. Ikuma, A. J. McWhorter, L. Adkins, and M. Kunduk, “Investigation of vocal bifurcations and voice patterns induced by asymmetry of pathological vocal folds,” J. Speech Lang. Hear . Res., vol. 66, no. 1, pp. 48–60, Jan. 2023

work page 2023
[6]

On the nature of vocal fry,

H. Hollien, P. Moore, R. W. Wendahl, and J. F. Michel, “On the nature of vocal fry,” J. Speech Hear . Res., vol. 9, no. 2, pp. 245–247, Jun. 1966

work page 1966
[7]

Freddie Mercury—acoustic analysis of speaking fundamental frequency, vibrato, and subhar- monics,

C. T. Herbst, S. Hertegard, D. Zangger-Borch, and P.- Å. Lindestad, “Freddie Mercury—acoustic analysis of speaking fundamental frequency, vibrato, and subhar- monics,” Logoped. Phoniatr . V ocol., vol. 42, no. 1, pp. 29–38, Jan. 2017

work page 2017
[8]

Acoustic characteristics of rough voice: Subharmonics,

K. Omori, H. Kojima, R. Kakani, D. H. Slavit, and S. M. Blaugrund, “Acoustic characteristics of rough voice: Subharmonics,” J. V oice, vol. 11, no. 1, pp. 40–47, Mar. 1997

work page 1997

Show all 48 references

[9]

Perception of pitch and roughness in vocal signals with subharmonics,

C. C. Bergan and I. R. Titze, “Perception of pitch and roughness in vocal signals with subharmonics,” J. V oice, vol. 15, no. 2, pp. 165–175, Jun. 2001

2001
[10]

Perceived pitch of synthesized voice with alternate cycles,

X. Sun and Y . Xu, “Perceived pitch of synthesized voice with alternate cycles,” J. V oice, vol. 16, no. 4, pp. 443– 459, Dec. 2002

2002
[11]

Perception and imitation of period-doubled phonation: Pitch and voice quality,

Y . Huang, “Perception and imitation of period-doubled phonation: Pitch and voice quality,” J. Acoust. Soc. Am. , vol. 156, no. 2, pp. 1391–1412, Aug. 2024

2024
[12]

Evaluation of machine-learning pitch estimation algorithms,

T. Ikuma, A. J. McWhorter, and M. Kunduk, “Evaluation of machine-learning pitch estimation algorithms,” in 13th ICVPB , Erlangen, Germany, Jul. 2024, pp. 28–9. [Online]. Available: https://www.icvpb-2024.de/ documents/8/Program_ICVPB2024.pdf

2024
[13]

Com- parison of fundamental frequency estimators with subharmonic voice signals,

T. Ikuma, M. Kunduk, and A. J. McWhorter, “Com- parison of fundamental frequency estimators with subharmonic voice signals,” 2025, arXiv preprint, https://arxiv.org/abs/2501.04789

2025 arXiv
[14]

I. R. Titze, Workshop on Acoustic V oice Analysis: Summary Statement . Denver, CO, USA: National Center for V oice and Speech, 1994. [Online]. Available: ncvs.org/archive/freebooks/summary-statement.pdf

1994
[15]

Multi-Dimensional V oice Program (MDVP) model 5105 software instruction manual,

KayPENTAX, “Multi-Dimensional V oice Program (MDVP) model 5105 software instruction manual,” Jun. 2008

2008
[16]

Acoustic model and evaluation of patho- logical voice production,

D. Deliyski, “Acoustic model and evaluation of patho- logical voice production,” in Eurospeech 1993 , Berlin, Germany, 1993, pp. 969–1972

1993
[17]

A pitch determination algorithm based on subharmonic-to-harmonic ratio,

X. Sun, “A pitch determination algorithm based on subharmonic-to-harmonic ratio,” in Proc. 6th ICSLP , vol. 4, Beijing, China, 2000, pp. 676–679

2000
[18]

Acoustic tracking of pitch, modal, and subhar- monic vibrations of vocal folds in Parkinson’s Disease and Parkinsonism,

J. Hlavni ˇcka, R. ˇCmejla, J. Klempí ˇr, E. R˚ užiˇcka, and J. Rusz, “Acoustic tracking of pitch, modal, and subhar- monic vibrations of vocal folds in Parkinson’s Disease and Parkinsonism,” IEEE Access , vol. 7, pp. 150 339– 150 354, 2019

2019
[19]

Fundamental frequency tracking in diplophonic voices,

P. Aichinger, M. Hagmüller, I. Roesner, B. Schneider- Stickler, J. Schoentgen, and F. Pernkopf, “Fundamental frequency tracking in diplophonic voices,” Biomed. Sig- nal Process. Control , vol. 37, pp. 69–81, Aug. 2017

2017
[20]

Tracking of mul- tiple fundamental frequencies in diplophonic voices,

P. Aichinger, M. Hagmüller, B. Schneider-Stickler, J. Schoentgen, and F. Pernkopf, “Tracking of mul- tiple fundamental frequencies in diplophonic voices,” IEEEACM Trans. Audio Speech Lang. Process. , vol. 26, no. 2, pp. 330–341, Feb. 2018

2018
[21]

A two-stage cepstral analysis procedure for the classification of rough voices,

S. N. Awan and J. A. Awan, “A two-stage cepstral analysis procedure for the classification of rough voices,” J. V oice, vol. 34, no. 1, pp. 9–19, Jan. 2020. SUBMITTED TO IEEE TRANS AUDIO SPEECH LANG PROCESS, JANURAY 2025 9

2020
[22]

Validation of subharmonics quantification using two-stage cepstral analysis,

I. Kitayama, K. Hosokawa, S. Iwaki, M. Yoshida, A. Miyauchi, M. Ogawa, and H. Inohara, “Validation of subharmonics quantification using two-stage cepstral analysis,” J. V oice, p. S0892199723003892, Dec. 2023

2023
[23]

Fully-convolutional net- work for pitch estimation of speech signals,

L. Ardaillon and A. Roebel, “Fully-convolutional net- work for pitch estimation of speech signals,” in Inter- speech 2019 , Sep. 2019, pp. 2005–2009

2019
[24]

Crepe: A convolutional representation for pitch estimation,

J. W. Kim, J. Salamon, P. Li, and J. P. Bello, “Crepe: A convolutional representation for pitch estimation,” in IEEE ICASSP 2018 , Calgary, AB, Apr. 2018, pp. 161– 165

2018
[25]

Fully convo- lutional networks for semantic segmentation,

J. Long, E. Shelhamer, and T. Darrell, “Fully convo- lutional networks for semantic segmentation,” in Proc IEEE CVPR 2015 . Boston, MA, USA: IEEE, Jun. 2015, pp. 3431–3440

2015
[26]

Time series classifica- tion from scratch with deep neural networks: A strong baseline,

Z. Wang, W. Yan, and T. Oates, “Time series classifica- tion from scratch with deep neural networks: A strong baseline,” in Proc. IJCNN 2017 . Anchorage, AK, USA: IEEE, May 2017, pp. 1578–1585

2017
[27]

Deep learning for time series classi- fication: A review,

H. Ismail Fawaz, G. Forestier, J. Weber, L. Idoumghar, and P.-A. Muller, “Deep learning for time series classi- fication: A review,” Data Min. Knowl. Discov. , vol. 33, no. 4, pp. 917–963, Jul. 2019

2019
[28]

An analysis of the diplo- phonia phenomenon,

P. Dejonckere and J. Lebacq, “An analysis of the diplo- phonia phenomenon,” Speech Commun., vol. 2, no. 1, pp. 47–56, May 1983

1983
[29]

Performance evaluation of subharmonic- to-harmonic ratio (SHR) computation,

C. T. Herbst, “Performance evaluation of subharmonic- to-harmonic ratio (SHR) computation,” J. V oice, vol. 35, no. 3, pp. 365–375, May 2021

2021
[30]

Parameterization of the glottal area, glottal flow, and vocal fold contact area,

I. R. Titze, “Parameterization of the glottal area, glottal flow, and vocal fold contact area,” J. Acoust. Soc. Am. , vol. 75, no. 2, pp. 570–580, 1984

1984
[31]

A four-parameter model of the glottis and vocal fold contact area,

——, “A four-parameter model of the glottis and vocal fold contact area,” Speech Commun. , vol. 8, no. 3, pp. 191–201, Sep. 1989

1989
[32]

Physiologically-Based Speech Simulation Us- ing an Enhanced Wave-Reflection Model of the V ocal Tract,

B. Story, “Physiologically-Based Speech Simulation Us- ing an Enhanced Wave-Reflection Model of the V ocal Tract,” Ph.D. dissertation, University of Iowa, Iowa City, IA, May 1995

1995
[33]

Speech Synthesis with Reflection-Type Line Analog,

J. Liljencrants, “Speech Synthesis with Reflection-Type Line Analog,” Ph.D. dissertation, Royal Institute of Tech- nology, Stockholm, Sweden, 1985

1985
[34]

Relation of structural and vibratory kinematics of the vocal folds to two acoustic measures of breathy voice based on computational mod- eling,

R. A. Samlan and B. H. Story, “Relation of structural and vibratory kinematics of the vocal folds to two acoustic measures of breathy voice based on computational mod- eling,” J. Speech Lang. Hear . Res. , vol. 54, no. 5, pp. 1267–1283, Oct. 2011

2011
[35]

Relation of perceived breathiness to laryngeal kinematics and acoustic measures based on computational modeling,

R. A. Samlan, B. H. Story, and K. Bunton, “Relation of perceived breathiness to laryngeal kinematics and acoustic measures based on computational modeling,” J. Speech Lang. Hear . Res., vol. 56, no. 4, pp. 1209–1223, Aug. 2013

2013
[36]

Acoustic and perceptual effects of left–right laryn- geal asymmetries based on computational modeling,

R. A. Samlan, B. H. Story, A. J. Lotto, and K. Bunton, “Acoustic and perceptual effects of left–right laryn- geal asymmetries based on computational modeling,” J. Speech Lang. Hear . Res., vol. 57, no. 5, pp. 1619–1637, Oct. 2014

2014
[37]

Synthesis of voiced sounds from a two-mass model of the vocal cords,

K. Ishizaka and J. L. Flanagan, “Synthesis of voiced sounds from a two-mass model of the vocal cords,” Bell Syst. Tech. J. , vol. 51, no. 6, pp. 1233–1268, 1972

1972
[38]

The physics of small-amplitude oscillation of the vocal folds,

I. R. Titze, “The physics of small-amplitude oscillation of the vocal folds,” J. Acoust. Soc. Am. , vol. 83, no. 4, pp. 1536–1552, 1988

1988
[39]

High-speed digital image analysis of vocal cord vibration in diplo- phonia,

S. Kiritani, H. Hirose, and H. Imagawa, “High-speed digital image analysis of vocal cord vibration in diplo- phonia,” Speech Commun. , vol. 13, no. 1-2, pp. 23–32, 1993

1993
[40]

The mechanisms of subharmonic tone generation in a synthetic larynx model,

S. Kniesburges, A. Lodermeyer, S. Becker, M. Traxdorf, and M. Döllinger, “The mechanisms of subharmonic tone generation in a synthetic larynx model,” J. Acoust. Soc. Am., vol. 139, no. 6, pp. 3182–3192, Jun. 2016

2016
[41]

Synthetic multi-line kymographic analysis: A spatiotem- poral data reduction technique for high-speed videoen- doscopy,

T. Ikuma, M. Kunduk, D. Fink, and A. J. McWhorter, “Synthetic multi-line kymographic analysis: A spatiotem- poral data reduction technique for high-speed videoen- doscopy,” J. Acoust. Soc. Am. , vol. 140, no. 4, pp. 2703– 2713, Oct. 2016

2016
[42]

Irregular vocal- fold vibration—High-speed observation and modeling,

P. Mergell, H. Herzel, and I. R. Titze, “Irregular vocal- fold vibration—High-speed observation and modeling,” J. Acoust. Soc. Am., vol. 108, no. 6, pp. 2996–3002, 2000

2000
[43]

Spatio-temporal analysis of irregular vocal fold oscil- lations: Biphonation due to desynchronization of spatial modes,

J. Neubauer, P. Mergell, U. Eysholdt, and H. Herzel, “Spatio-temporal analysis of irregular vocal fold oscil- lations: Biphonation due to desynchronization of spatial modes,” J. Acoust. Soc. Am. , vol. 110, no. 6, pp. 3179– 3192, 2001

2001
[44]

Regulating glottal airflow in phonation: Application of the maximum power transfer theorem to a low dimensional phonation model,

I. R. Titze, “Regulating glottal airflow in phonation: Application of the maximum power transfer theorem to a low dimensional phonation model,” J. Acoust. Soc. Am. , vol. 111, no. 1, pp. 367–376, Jan. 2002

2002
[45]

Analysis, synthesis, and perception of voice quality variations among female and male talkers,

D. H. Klatt and L. C. Klatt, “Analysis, synthesis, and perception of voice quality variations among female and male talkers,” J. Acoust. Soc. Am. , vol. 87, no. 2, pp. 820–857, 1990

1990
[46]

Glottal airflow and transglottal air pressure measurements for male and female speakers in soft, normal, and loud voice,

E. B. Holmberg, R. E. Hillman, and J. S. Perkell, “Glottal airflow and transglottal air pressure measurements for male and female speakers in soft, normal, and loud voice,” J. Acoust. Soc. Am. , vol. 84, no. 2, pp. 511–529, Aug. 1988

1988
[47]

Adam: A Method for Stochastic Optimization,

D. P. Kingma and J. Ba, “Adam: A Method for Stochastic Optimization,” in ICLR 2015 . San Diego, CA: arXiv, 2015

2015
[48]

Disordered V oice Database and Program [Model 4337],

KayPENTAX and Massachusetts Eye and Ear Infirmary, “Disordered V oice Database and Program [Model 4337],” 2006

2006

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

Microphone and electroglottographic data from dyspho- nic patients: Type 1, 2 and 3 signals,

A. Behrman, C. J. Agresti, E. Blumstein, and N. Lee, “Microphone and electroglottographic data from dyspho- nic patients: Type 1, 2 and 3 signals,” J. V oice, vol. 12, no. 2, pp. 249–260, Jan. 1998

work page 1998

[2] [2]

Diplophonia reappraised,

L. Cavalli and A. Hirson, “Diplophonia reappraised,” J. V oice, vol. 13, no. 4, pp. 542–556, 1999

work page 1999

[3] [3]

[Spectrographic study of voice disorders: subharmonics],

F. Núñez Batalla, C. Suárez Nieto, C. Muñoz Pinto, L. Baragaño Río, M. J. Alvarez Zapico, and A. Martínez Ferreras, “[Spectrographic study of voice disorders: subharmonics],” Acta Otorrinolaringol. Esp., vol. 51, no. 1, pp. 52–56, 2000

work page 2000

[4] [4]

A study of subharmonics in connected speech material,

E. Kramer, R. Linder, and R. Schönweiler, “A study of subharmonics in connected speech material,” J. V oice, vol. 27, no. 1, pp. 29–38, Jan. 2013

work page 2013

[5] [5]

Investigation of vocal bifurcations and voice patterns induced by asymmetry of pathological vocal folds,

T. Ikuma, A. J. McWhorter, L. Adkins, and M. Kunduk, “Investigation of vocal bifurcations and voice patterns induced by asymmetry of pathological vocal folds,” J. Speech Lang. Hear . Res., vol. 66, no. 1, pp. 48–60, Jan. 2023

work page 2023

[6] [6]

On the nature of vocal fry,

H. Hollien, P. Moore, R. W. Wendahl, and J. F. Michel, “On the nature of vocal fry,” J. Speech Hear . Res., vol. 9, no. 2, pp. 245–247, Jun. 1966

work page 1966

[7] [7]

Freddie Mercury—acoustic analysis of speaking fundamental frequency, vibrato, and subhar- monics,

C. T. Herbst, S. Hertegard, D. Zangger-Borch, and P.- Å. Lindestad, “Freddie Mercury—acoustic analysis of speaking fundamental frequency, vibrato, and subhar- monics,” Logoped. Phoniatr . V ocol., vol. 42, no. 1, pp. 29–38, Jan. 2017

work page 2017

[8] [8]

Acoustic characteristics of rough voice: Subharmonics,

K. Omori, H. Kojima, R. Kakani, D. H. Slavit, and S. M. Blaugrund, “Acoustic characteristics of rough voice: Subharmonics,” J. V oice, vol. 11, no. 1, pp. 40–47, Mar. 1997

work page 1997