Pith. sign in

REVIEW 4 major objections 5 minor 37 references

Nosey: Open-source hardware for acoustic nasalance

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Nosey, a low-cost 3D-printed open-source nasalance device, yields raw nasalance scores about 21–22% higher than a commercial nasometer while reproducing the magnitude of nasalance contrasts between phonological environments.

desk verdict Open-source nasalance hardware is a real, useful artifact, but the 'comparable contrasts' claim rests on null interactions and a number typo in the reported raw offset. read the letter →

arxiv 2505.23339 v1 pith:YEWCTN3A submitted 2025-05-29 cs.SD cs.CLeess.AS

Maya Dewhurst , Jack Collins , Justin J. H. Lo , Roy Alderton , Sam Kirkham This is my paper
classification cs.SDcs.CLeess.AS
keywords nasalancenasalityacousticsmeasurementhardwarephoneticsspeechproductionopen-source
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

Nosey is a low-cost, 3D-printed, open-source device for recording nasalance—the ratio of nasal to total oral-plus-nasal acoustic energy during speech. The paper tests Nosey against a commercial nasometer and claims that, although Nosey's raw nasalance scores run about 21–22% higher, the magnitude of nasalance differences between phonological environments (oral versus nasal contexts) is statistically comparable between the two systems. If the claim holds, laboratories can use inexpensive, customizable hardware to study relative patterns of nasal coarticulation without buying a commercial nasometer. The paper is explicit that raw Nosey scores are not interchangeable with commercial norms; only the between-environment contrast magnitudes are claimed to be reliable.

What carries the argument

The load-bearing object is the Nosey hardware itself: a 3D-printed PLA baffle that sits between nose and mouth, with two cardioid electret condenser microphones held in a removable articulated clip on either side, feeding standard analog outputs to an audio interface. The baffle is meant to separate nasal and oral acoustic energy, and nasalance is computed as $A_n/(A_n + A_o)\cdot100$ from intensity extracted at vowel midpoints. The argument is carried by comparing estimated marginal means from linear regressions with an environment-by-system interaction, so the claim is about whether the two devices register the same differences between environments, not the same levels.

What would settle it

Play a steady tone from a mouth simulator while the nasal side is sealed and compare the nasal microphone's output to its baseline; if it captures substantial energy, the baffle is leaking and the apparent contrast match could be coincidental. A complementary check is to block a real speaker's nasal passages and see whether Nosey still shows the same between-environment pattern as the commercial device.

Watch

Extended reading notes

Core claim

The central discovery is that the open-source Nosey system reproduces the contrast structure of a commercial nasometer even though its absolute scores are offset. For two speakers reading words in five vowels and several phonological environments, Nosey gave average nasalance about 22% and 21% higher than the commercial device, yet pairwise comparisons of six environment pairs showed no significant difference between systems in the magnitude of nasalance contrasts. The paper concludes that, under the conditions tested, Nosey can be relied on to represent magnitudes of nasalance contrast across phonological environments comparable to those from commercial devices, making it a viable low-cost alternative for studies that compare environments rather than absolute nasality.

Load-bearing premise

The load-bearing premise is that the plastic baffle and the two directional microphones separate nose and mouth sounds the way the commercial device does; if sound leaks around the baffle, or one microphone catches the other channel, the comparable contrast magnitudes could be an artifact of correlated bleed.

Editorial extensions

If this is right

  • Raw nasalance from Nosey should not be compared directly with commercial device norms or across devices without calibration.
  • Nosey can be used to measure relative differences in nasal coarticulation across phonological environments, which is the main phonetic research use case tested here.
  • Researchers can substitute microphones, adjust clip distance, or change baffle geometry through the editable design files, enabling controlled experiments on how each component affects nasalance.
  • Because Nosey outputs standard microphone-level analog signals, it can be integrated into multichannel acoustic and articulatory recording setups using any audio interface.
  • The paper's two-speaker comparison is a first validation; the authors note that larger speaker samples and time-varying nasalance remain for future work.

Reading between the lines

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

  • My inference: the roughly constant 21–22% offset suggests a per-speaker gain calibration could map Nosey scores onto a commercial-like scale, but the paper does not test any calibration procedure.
  • My inference: if the modular geometry transfers across labs, identical printable hardware could make multi-site nasal coarticulation studies more reproducible than commercial units, which differ headset by headset; this goes beyond the paper's single-site comparison.
  • My inference: the authors' planned comparisons of hypercardioid microphones and denser baffle materials imply that cross-channel acoustic bleed is the main threat to validity; a direct bleed measurement (sealing one side and playing a known source) would sharpen the validation.
  • My inference: a testable extension would be recording the same wordlist with Nosey while swapping the nasal and oral microphone channels; if the environment-contrast pattern follows the channels rather than the phonology, the baffle is not the effective separator.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper introduces Nosey, a low-cost, 3D-printed open-source hardware system for acoustic nasalance measurement, and compares it with a commercial icSpeech nasometer. Two male speakers read 60 monosyllabic English words in varied phonological environments, with six repetitions per system; nasalance was computed at vowel midpoints using the standard ratio An/(An+Ao). The authors report that Nosey gives consistently higher raw nasalance scores than icSpeech, but that pairwise comparisons of phonological-environment contrasts show no significant system-by-environment interactions after Bonferroni correction. They interpret this as evidence that Nosey captures the magnitude of nasalance contrasts comparably to a commercial device, and conclude that Nosey is a viable flexible and cost-effective alternative, while acknowledging limitations including manual gain calibration, a small speaker pool, and untested baffle and microphone factors.

Significance. The open-hardware contribution is genuinely valuable: CAD files, assembly instructions, and a versioned repository with a DOI are publicly released; the signal chain uses standard audio interfaces; and the data-processing pipeline is described in enough detail to be reproduced. If the comparability claim were robustly established, the device would fill a real gap for low-cost, customizable nasalance measurement in phonetic fieldwork and laboratory research. The manuscript is also honest about several limitations. However, the central claim of comparable contrast magnitudes currently rests on non-significant interactions without equivalence bounds or power analysis; the reported effect-size arithmetic in Section 3.4 is internally inconsistent; and the acoustic-isolation assumptions of the custom baffle and microphone geometry are explicitly left untested. These issues are load-bearing for the paper's main conclusion, though they appear fixable with reanalysis or a more cautious framing.

major comments (4)
  1. [§3.4 and §4.1] The central conclusion that Nosey 'can be relied upon to represent magnitudes of contrasts in nasalance across phonological environments comparable to those reported using commercially available devices' rests entirely on the absence of significant system-by-environment interactions (all adjusted p > .1, confidence intervals crossing zero). A non-significant interaction does not establish equivalence: with only two speakers, no pre-specified equivalence margin, and no power analysis, the result may simply reflect low sensitivity. Please add an equivalence analysis (for example, two one-sided tests with a phonetically meaningful bound) or report the smallest contrast difference the design could detect at a given power, and in either case soften the definitive 'can be relied upon' wording to match the evidentiary strength.
  2. [§3.4] The reported raw offset is arithmetically inconsistent: with SE = 0.0030 and t = 7.37, the estimate is 0.022, i.e. about 2.2 percentage points, not 22%; similarly, 0.0033 × 6.36 = 0.021 for Speaker 2. If the intended unit is percentage points, the text should read 'an increase of 2.2%' and '2.1%'; if 22% is intended, the standard errors or t values must be corrected. The magnitude of the raw offset is part of the headline result and must be reported consistently.
  3. [§2.2–2.4 and §4.2] The comparability claim assumes that the 3D-printed PLA baffle and the AKG CK99L cardioid microphones separate nasal and oral acoustic energy in the same way as the commercial device. The manuscript itself lists baffle materials, microphone polar patterns, and microphone distance as untested future work. If acoustic bleed around the baffle or differences in channel gain depend on phonological environment, the observed non-significant interactions could occur even if the two systems capture different underlying contrasts. Please add a basic acoustic-isolation check (for example, an occlusion/leak test with a nasal-only or oral-only excitation, or a comparison with the same microphones mounted in the commercial headset geometry), or explicitly restrict the conclusion to the specific untested hardware configuration.
  4. [§3.3] The statistical model is not specified in enough detail to assess the standard errors that drive the main null result. The regression is described as a linear regression per speaker with environment, system, and their interaction, plus a control for vowel, but it is not stated whether the model accounts for the nesting of tokens within words and repetitions, or whether the six repetitions and 60 words are treated as independent observations. If repeated measures are pooled without appropriate clustering or random effects, the interaction standard errors may be anti-conservative, making the non-significant interactions appear more informative than they are. Please clarify the model specification and, if necessary, reanalyze with item-level or repetition-level random effects.
minor comments (5)
  1. [§3.3] The phrase 'applied top-values to p-values' appears to contain a typo; it should read 'applied to p-values'.
  2. [References] Reference [22] lists the year as '20167'; this should be '2016'.
  3. [Figure 6] The figure legend and caption do not state the confidence level of the error bars or the units of the difference-of-differences; these should be stated (e.g., 95% CI in nasalance percentage points).
  4. [§3.2] The paper does not state whether the order of devices was counterbalanced across speakers or sessions, or whether the two devices were recorded in the same session; please clarify, since order or session effects could contribute to the raw offset.
  5. [§3.4] The phrase 'both p < .01' could be replaced with exact p-values, which would make the strength of the raw system difference easier to evaluate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the nasalance comparison is a direct empirical measurement with no fitted parameter or self-citation chain driving the central claim.

full rationale

The paper's central claim—that Nosey and the icSpeech nasometer capture comparable magnitudes of nasalance contrast between phonological environments—is supported by a direct empirical comparison, not by a derivation that reduces to its own inputs. Nasalance is computed from the standard definition in Equation (1), An/(An+Ao)*100, which is taken from external clinical literature [15]; the same equation is applied to both systems. No parameter is fitted to make the conclusion true, and the higher raw nasalance scores for Nosey are reported as an observed offset rather than forced by normalization or calibration. The statistical analysis uses linear regression with an interaction between phonological environment and system, and the conclusion of comparable contrast magnitudes rests on non-significant interaction terms; that is a legitimate evidentiary weakness (absence-of-evidence rather than a demonstrated equivalence), but it is not circularity. The only self-citations in the reference list ([6] and [17]) support background statements about coarticulation and nasalance research and are not load-bearing for the validity of the hardware or the outcome of the comparison. There is no self-definitional step, no fitted input relabeled as a prediction, no imported uniqueness theorem, and no ansatz smuggled in via citation. The paper is self-contained against the commercial device as an external benchmark, so the appropriate finding is no significant circularity.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

No theoretical entities or fitted model parameters are introduced. The listed assumptions are measurement-level premises inherited from standard nasometry; the main hand-set setting is microphone gain, which affects raw scores but not the reported contrast pattern.

free parameters (1)
  • Microphone gain settings on Nosey = not reported numerically; gains set equal on Focusrite Scarlett 2i2
    Chosen by the experimenters rather than calibrated against a reference. The authors attribute the 21-22 percent raw nasalance offset to this manual calibration in Sections 3.4 and 4.1, but the absolute gain values are not given.
assumptions (3)
  • domain assumption Equation (1) is a valid operationalization of nasality as the ratio of nasal energy to total oral plus nasal energy.
    Section 1, Equation (1): the paper adopts Fletcher et al.'s nasalance definition without deriving or validating it against articulatory ground truth.
  • domain assumption The baffle and microphone arrangement separate nasal and oral acoustic radiation with negligible cross-talk.
    Sections 2.2-2.4: the design assumes cardioid microphones and the baffle create isolated channels; cross-signal bleed is named as a risk in Section 4.2 but not measured.
  • domain assumption A single intensity sample at vowel midpoint is representative of each phonological environment's nasalance.
    Section 3.3: intensity is extracted at the midpoint of each vowel token; the authors note time-varying nasalance as future work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Nosey: Open-source hardware for acoustic nasalance." pith.science (2026). https://pith.science/paper/YEWCTN3A

@misc{pith2026250523339,
  author       = {Pith},
  title        = {Pith review of: Nosey: Open-source hardware for acoustic nasalance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YEWCTN3A}},
  note         = {Machine review of arXiv:2505.23339}
}
read the original abstract

We introduce Nosey (Nasalance Open Source Estimation sYstem), a low-cost, customizable, 3D-printed system for recording acoustic nasalance data that we have made available as open-source hardware (http://github.com/phoneticslab/nosey). We first outline the motivations and design principles behind our hardware nasalance system, and then present a comparison between Nosey and a commercial nasalance device. Nosey shows consistently higher nasalance scores than the commercial device, but the magnitude of contrast between phonological environments is comparable between systems. We also review ways of customizing the hardware to facilitate testing, such as comparison of microphones and different construction materials. We conclude that Nosey is a flexible and cost-effective alternative to commercial nasometry devices and propose some methodological considerations for its use in data collection.

Figures

Figures reproduced from arXiv: 2505.23339 by the authors.

Figure 2
Figure 2. shows an assembled view of the 3D model design. The bolt at (1) fixes the handle (2) to the baffle (3). The re￾movable microphone clip at (4) slots into the baffle and can be replaced with alternative versions to allow swapping different 1https://github.com/phoneticslab/nosey All versioned releases are archived at: https://doi.org/10.5281/zenodo.15543852 [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. 3D model of the removable, articulated, dual micro￾phone holder. plate (M3 nut / bolt). This also facilitates the articulated micro￾phone clip, which can be angled by loosening the bolts. 2.5. Customization We provide editable FreeCAD models and 3D Manufacturing Format files for all model components. The repository includes details on how to modify the baffle shape and distance of the microphone cutout in the baffle… view at source ↗
Figure 5
Figure 5. Baffle sketch with highlighted measurement control￾ling the distance of the cutout to the front of the baffle. speaker_1 speaker_2 −0.025 0.000 0.025 −0.025 0.000 0.025 CVC − CVN CVC − NVC CVC − NVN CVN − NVC CVN − NVN NVC − NVN Difference of Differences (icSpeech − NOSEY) Environment Pair [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figures from the paper (1 more)
Figure 6
Figure 6. Figure 6: Pairwise comparison of selected phonological envi￾ronment pairs. Values below zero suggest a greater contrast magnitude for Nosey; values above zero suggest a greater con￾trast magnitude for icSpeech. An error bar that does not cross zero suggests a significant differe…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

37 extracted references · 37 canonical work pages

  1. [1]

    Introduction Nasality is an integral part of human speech that underpins many important questions in phonetics and phonology [1, 2]. Achieved by velum lowering, nasality can spread from phono- logically nasal segments to nearby oral segments via coarticula- tion in language-, dialect- and speaker-specific ways [3, 4, 5, 6]. Such variation in coarticulatio...

    work page Pith review arXiv 2025

  2. [2]

    Motivations We report the design of an open-source, customizable, 3D- printed nasalance system

    Open-source nasalance hardware 2.1. Motivations We report the design of an open-source, customizable, 3D- printed nasalance system. Our aims were to develop a reliable system that is easy to use, cheap to manufacture and provides additional flexibility as a testing and development platform. For example, our system allows for customized microphone selec- t...

    work page

  3. [3]

    M3 nut and bolt (12 mm long) [×2]

    work page

  4. [4]

    plate (M3 nut / bolt)

    M8 bolt (12 mm long) [×1] These bolts are used to fix the baffle to the handle (M8 bolt) and to attach the microphone clips to the removable mounting Figure 3: 3D model of the removable, articulated, dual micro- phone holder. plate (M3 nut / bolt). This also facilitates the articulated micro- phone clip, which can be angled by loosening the bolts. 2.5. Cu...

    work page

  5. [5]

    Overview We assessed the performance of Nosey by comparing it to a commercial nasometer: the Nasality Microphone developed by icSpeech (Canterbury, UK)

    Comparison with commercial system 3.1. Overview We assessed the performance of Nosey by comparing it to a commercial nasometer: the Nasality Microphone developed by icSpeech (Canterbury, UK). The basic designs of the two sys- tems are very similar (and also very similar to other commercial devices), with two small microphones separated by an acoustic baff...

    work page

  6. [6]

    Language-specific and individual variation in anticipatory nasal coarticulation: A comparative study of American English, French, and German,

    M. Pouplier, F. Rodriquez, J. J. H. Lo, R. Alderton, B. G. Evans, E. Reinisch, and C. Carignan, “Language-specific and individual variation in anticipatory nasal coarticulation: A comparative study of American English, French, and German,”Journal of Phonetics, vol. 107, p. 101365, 2024

    work page 2024

  7. [7]

    Summary and limitations We have outlined the development and release of open-source hardware for acoustic nasalance data collection in a way that is accessible and customizable

    Discussion 4.1. Summary and limitations We have outlined the development and release of open-source hardware for acoustic nasalance data collection in a way that is accessible and customizable. While a comprehensive valida- tion of Nosey is beyond the scope of this paper, we show that performance is comparable with one commercial system. Raw nasalance is ...

    work page

  8. [8]

    and UKRI/AHRC fellowship AH/Y002822/1 to S.K

    Acknowledgements This research was supported by UKRI/ESRC Doctoral Fel- lowship ES/P000665/1 to M.D. and UKRI/AHRC fellowship AH/Y002822/1 to S.K

    work page

Show all 37 references
  1. [9]

    Nasals and nasalization,

    J. Hajek, “Nasals and nasalization,” in Manual of Romance Pho- netics and Phonology , C. Gabriel, R. Gess, and T. Meisenburg, Eds. Berlin: De Gruyter, 2022, pp. 215–241

    2022

  2. [10]

    Coarticulation and phonology,

    J. J. Ohala, “Coarticulation and phonology,” Language and Speech, vol. 36, no. 2, 3, pp. 155–170, 1993

    1993

  3. [11]

    Acoustic investigation of anticipatory vowel nasalization in a Caribbean and a non-Caribbean dialect of Span- ish,

    S. Bongiovanni, “Acoustic investigation of anticipatory vowel nasalization in a Caribbean and a non-Caribbean dialect of Span- ish,” Linguistics Vanguard, vol. 7, no. 1, p. 20200008, 2021

    2021

  4. [12]

    Patterns of soft palate movements in six languages,

    H. Clumeck, “Patterns of soft palate movements in six languages,” Journal of Phonetics, vol. 4, no. 4, pp. 337–351, 1976

    1976

  5. [13]

    The physiological basis of the phonologization of vowel nasalization: A real-time MRI analysis of American and Southern British En- glish,

    C. Cunha, P. Hoole, D. V oit, J. Frahm, and J. Harrington, “The physiological basis of the phonologization of vowel nasalization: A real-time MRI analysis of American and Southern British En- glish,” Journal of Phonetics, vol. 105, p. 101329, 2024

    2024

  6. [14]

    A practical method of estimating the time-varying degree of vowel nasalization from acoustic features,

    C. Carignan, “A practical method of estimating the time-varying degree of vowel nasalization from acoustic features,”The Journal of the Acoustical Society of America, vol. 149, no. 2, pp. 911–922, 2021

    2021

  7. [15]

    A coarticulatory path to sound change,

    P. S. Beddor, “A coarticulatory path to sound change,” Language, vol. 85, no. 4, pp. 785–821, 2009

    2009

  8. [16]

    Oppa-Ng Gamsahamnita-Ng∼∼∼: The phonetics of nasal cuteness in Korean aegyo,

    D. Crosby, “ Oppa-Ng Gamsahamnita-Ng∼∼∼: The phonetics of nasal cuteness in Korean aegyo,” Ph.D. dissertation, University of South Carolina, 2023

    2023

  9. [17]

    Listeners’ evaluations of voice quality in Australian English speakers,

    J. Pittam, “Listeners’ evaluations of voice quality in Australian English speakers,” Language and Speech, vol. 30, no. 2, pp. 99– 113, 1987

    1987

  10. [18]

    Impressions of personality from intentional voice quality in Arabic-speaking and native Finnish-speaking listeners,

    T. Waaramaa, P. Lukkarila, K. J ¨arvinen, A. Geneid, and A.-M. Laukkanen, “Impressions of personality from intentional voice quality in Arabic-speaking and native Finnish-speaking listeners,” Journal of Voice, vol. 35, no. 2, pp. 326.e21–326.e28, 2021

    2021

  11. [19]

    Nasalance in the speech of children with normal hearing and children with hearing loss,

    S. G. Fletcher, F. Mahfuzh, and H. Hendarmin, “Nasalance in the speech of children with normal hearing and children with hearing loss,” American Journal of Speech-Language Pathology , vol. 8, no. 3, pp. 241–248, 1999

    1999

  12. [20]

    The relation- ship between nasalance and nasality in children with cleft palate,

    T. Watterson, S. C. McFarlane, and D. S. Wright, “The relation- ship between nasalance and nasality in children with cleft palate,” Journal of Communication Disorders , vol. 26, no. 1, pp. 13–28, 1993

    1993

  13. [21]

    On the acoustical features of vowel nasality in English and French,

    W. Styler, “On the acoustical features of vowel nasality in English and French,” The Journal of the Acoustical Society of America , vol. 142, no. 4, pp. 2469–2482, 2017

    2017

  14. [22]

    Earbuds: A method for analyzing nasality in the field,

    J. Stewart and M. Kohlberger, “Earbuds: A method for analyzing nasality in the field,” Language Documentation & Conservation, vol. 11, no. 1, pp. 49–80, 20167

  15. [23]

    Quantitative and graphic analysis of prosthetic treatment for “nasalance

    S. G. Fletcher, I. Sooudi, and S. D. Frost, “Quantitative and graphic analysis of prosthetic treatment for “nasalance” in speech,” Journal of Prosthetic Dentistry, vol. 32, no. 3, pp. 284– 291, 1974

    1974

  16. [24]

    Using ultrasound and nasalance to separate oral and nasal contributions to formant frequencies of nasalized vowels,

    C. Carignan, “Using ultrasound and nasalance to separate oral and nasal contributions to formant frequencies of nasalized vowels,” Journal of the Acoustical Society of America, vol. 143, no. 5, pp. 2588–2601, 2018

    2018

  17. [25]

    Nasal coarticulation in Lombard speech,

    J. J. H. Lo, “Nasal coarticulation in Lombard speech,” Speech Communication, p. 103205, 2025

    2025

  18. [26]

    Normative data on nasalance scores for Swedish as measured on the Nasometer: Influence of dialect, gender, and age,

    K. Brunneg ˚ard and J. van Doorn, “Normative data on nasalance scores for Swedish as measured on the Nasometer: Influence of dialect, gender, and age,” Clinical Linguistics & Phonetics , vol. 23, no. 1, pp. 58–69, 2009

    2009

  19. [27]

    Normative nasalance scores for Estonian children,

    L. Lehes, R. Horn, P. Lippus, M. Padrik, P. Kasen ˜omm, and T. Jagom¨agi, “Normative nasalance scores for Estonian children,” Clinical Linguistics & Phonetics, vol. 32, no. 11, pp. 1054–1066, 2018

    2018

  20. [28]

    Intensity was extracted at the mid- point of each vowel token, and nasalance was calculated using Equation 1

    written by author M.D. Intensity was extracted at the mid- point of each vowel token, and nasalance was calculated using Equation 1. All statistical analyses were conducted in R. To investi- gate nasalance differences across phonological environments and recording systems, for...

  21. [29]

    Comparison of nasalance scores obtained with the Nasometers 6200 and 6450,

    G. De Boer and T. Bressmann, “Comparison of nasalance scores obtained with the Nasometers 6200 and 6450,” The Cleft Palate Craniofacial Journal, vol. 51, no. 1, pp. 90–97, 2014

    2014

  22. [30]

    Differences in nasalance scores obtained with different nasometer headsets,

    T. Bressmann and B. H. Y . Tang, “Differences in nasalance scores obtained with different nasometer headsets,” Clinical Linguistics & Phonetics, pp. 1–11, 2024

    2024

  23. [31]

    Ground-truth validation of the “earbuds method

    C. Carignan, “Ground-truth validation of the “earbuds method” for measuring acoustic nasalance,” Journal of the Acoustical So- ciety of America, vol. 156, no. 2, pp. 851–864, 2024

    2024

  24. [32]

    J. C. Wells, Accents of English: Volumes 1–3. Cambridge: Cam- bridge University Press, 1982

    1982

  25. [33]

    Speaker identification on the SCOTUS corpus,

    J. Yuan and M. Liberman, “Speaker identification on the SCOTUS corpus,” The Journal of the Acoustical Society of America , vol. 123, pp. 3878–3878, May 2008

    2008

  26. [34]

    Coretta, speakr: A Wrapper for the Phonetic Software ’Praat’, 2025, R package version 3.2.4

    S. Coretta, speakr: A Wrapper for the Phonetic Software ’Praat’, 2025, R package version 3.2.4. [Online]. Available: https://CRAN.R-project.org/package=speakr

    2025

  27. [35]

    [Online]

    R Core Team, R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria, 2025. [Online]. Available: https://www.R-project.org/

    2025

  28. [36]

    Praat: doing phonetics by computer,

    P. Boersma and D. Weenink, “Praat: doing phonetics by computer,” 2025. [Online]. Available: http://www.praat.org/

    2025

  29. [37]

    R. V . Lenth, emmeans: Estimated Marginal Means, aka Least-Squares Means, 2025, R package version 1.10.7. [Online]. Available: https://CRAN.R-project.org/package=emmeans

    2025

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.