Vocal violence in actors: An investigation into its acoustic consequences and the effects of hygienic laryngeal release training

Acoustic analysis techniques were used to investigate the short-term consequences of vocally violent behavior, and to compare voice production before and after training in hygienic laryngeal release (HLR) techniques. Twenty-seven actors ranging in age from 17 to 48 years were audiorecorded before and after multiple productions of 4 vocally violent behaviors: grunting, groaning, sobbing, and shouting. After training in HLR techniques, the experimental protocol was repeated. Audiorecordings of vowels (produced at 3 pitch levels: modal F0, minimum F0, maximum F0) before and after vocal violence, and before and after HLR training, were analyzed using the Multidimensional Voice Program (4305, Kay Elemetrics Corp, Lincoln Park, NJ). After vocal violence, no consistent acoustic changes were detected for voice generated at modal and minimum F0; however, significant increases in both fundamental frequency range and maximum F0 were observed. After training in HLR techniques, acoustic measures sensitive to pitch and amplitude perturbation, and non-harmonic noise, improved across pitch levels. The results also indicated that vocal training does defend the laryngeal system from undesirable changes related to vocally violent maneuvers that might surface at the extremes of an actor's pitch range. Because the HLR technique used in this investigation was multimodal, interesting questions are raised regarding which aspect of training is primarily responsible for the observed effects. Further study is required to identify such factors.     

The effects of preventive vocal hygiene education on the vocal hygiene habits and perceptual vocal characteristics of training singers

The purpose of the present study was to determine the effects of vocal hygiene education on the vocal hygiene behaviors and perceptual vocal characteristics of untrained singers. Eleven adult untrained singers served as subjects. They attended four 1-hour class sessions on vocal hygiene, including anatomy and physiology of the phonatory mechanism, vocally abusive behaviors, voice disorders commonly seen in singers, and measures to prevent voice disorders. Pre- and postinstruction surveys were used to record subjects' vocal abuses and their perceptions of their speaking and singing voice. They also rated their perceived value of vocal hygiene education. Results revealed minimal changes in vocal hygiene behaviors and perceptual voice characteristics. The subjects did report a high degree of benefit and learning, however.     

Phonomicrosurgical management of vocal fold polyps: The subepithelial microflap resection technique

Vocal fold polyps are typically caused by acute and chronic trauma to the microvasculature of the superficial lamina propria (SLP). Shearing stresses that are induced by hyperfunctional glottal sound production lead to bleeding into the SLP and malformed neo-vascularized masses. Because the primary process does not involve the epithelium, the authors designed a technique to resect hemorrhagic polyps by epithelial cordotomy with partial or complete preservation of the vocal fold epithelium. This approach is different from the traditional microsurgical resection of hemorrhagic polyps by amputation with or without the carbon dioxide laser. Forty patients who underwent microlaryngoscopic resection of hemorrhagic polyps from 1996 through 1998 were reviewed retrospectively. Thirty-six of the 40 procedures were by epithelial cordotomy and subepithelial removal of the polyp contents. Sixteen of 36 were assisted by a subepithelial infusion of saline and epinephrine, and all were 3 mm to 6 mm. Four of 40 polyps were amputated; all of these were less than 3 mm and were pedicled on a narrow base. Cold instruments were used exclusively in all 40 patients. Postoperative laryngeal stroboscopy within 2 weeks revealed improved mucosal wave propagation and improved glottal closure in all 33 patients in whom postoperative strobovideolaryngoscopy was available. The epithelial cordotomy technique was introduced to minimize disturbance of normal SLP and epithelium. Despite the hemorrhagic nature of these lesions, cold instruments could be used exclusively with facility due to careful microdissection between the polyp and the residual normal SLP and the enhanced hemostasis provided by the subepithelial infusion of saline and epinephrine. The rapid return to improved glottal function is the result of this ultra tissue-sparing technique.     

Voice characteristics, effects of voice therapy, and long-term follow-up of contact granuloma patients

This study evaluates the laryngoscopic findings and voice characteristics of male contact granuloma patients before and after voice therapy and at a follow-up about 9 years later. Pre- and posttherapy recordings as well as follow-up recordings were made for 19 granuloma patients. Pretherapy revealed the most salient perceptual voice characteristics were low pitch, monotony, and a high degree of vocal fry and hyperfunction. Interjudge reliability for these traits was high. Immediately following therapy the healed patients (n = 10) had a decrease in hyperfunction, vocal fry, and monotony, while the unhealed patients (n = 9) had an increase in hyperfunction and vocal fry decreased only marginally. Monotony decreased significantly in this group. As regards the acoustic analyses, no significant differences were found in mean fundamental frequency (F₀) or perturbation. At the follow-up assessment 4 patients had granuloma while 15 had normal laryngeal status. Perceptually their voice characteristics resembled those pretherapy independently of the laryngeal findings. The results suggest that reduced hyperfunction and decreased vocal fry may create better circumstances for the healing process at the posterior glottis.     

一种基于奇异谱的语音激活检测方法

为了提高语音激活检测在低信噪比环境中的检测性能,提出了一种基于奇异谱的语音激活检测方法。首先用多窗口方法计算每一帧语音信号的相关矩阵;然后对相关矩阵进行奇异值分解;利用奇异值可以反映有用信号和噪声分布情况的特性,将每一帧语音信号经过加权处理后的最大奇异值与自适应阈值进行比较进行语音激活检测。该方法原理简单,易于硬件实现,通过实验仿真表明,在低信噪比环境下,和基于对数能量方法相比,本文方法也能够很好的区分语音段和非语音段,有良好的检测性能。