Research on the mechanism of mildew contamination affecting the sound quality of analog tape archives

Research on the mechanism of mildew contamination affecting the sound quality of analog tape archives | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Research on the mechanism of mildew contamination affecting the sound quality of analog tape archives Zhihui Jia, Yanan Wang, Shujiao Yu, Quanfeng Cai, Bodian Liu, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8597238/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 8 You are reading this latest preprint version Abstract Since 1947, analog tape recording has become the main method of sound recording, and has been widely used in music dissemination, cultural popularization, scientific research, news interviews, etc. For more than 40 years, giving birth to a large number of important audio recording archives. Due to the presence of a large amount of adhesive in the tape, mildew have been found on the surface of the magnetic layer and the edges of many tapes, posing a serious threat to the long-term preservation and the sound quality of the recordings. In this paper, ATR-FTIR and SEM were used to extensively characterize the chemical and physical of magnetic tape materials. To assess the effects of mildew contamination, the common strain Penicillium and Aspergillus was used to prepare mildew corrosive samples with different coverage degrees, and the corresponding audio samples were analyzed by Praat, a professional speech analysis software. The relationship between wideband spectrogram, sound intensity contours, formants plots, center of gravity, standard deviation, skewness, kurtosis, band energy difference and coverage degree of mildew contaminated samples were analyzed. Additionally, the surface roughness and morphology of the analog tapes were observed using the laser microscopy system. Based on the above information, the mechanism by which the dual effects of mildew coverage and corrosion affect sound quality was revealed. This research provides a theoretical foundation for improved and restored strategies to mitigate mildew damage contaminated analog tape archives in the future. sound quality magnetic audio tape mildew contamination Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 1 Introduction In the early 20th century, people began to attempt to record sound using magnetic materials. In 1935, the AEG company in Germany introduced the world's first tape recorder, utilizing magnetic tape with a paper base coated in iron oxide. This innovation significantly enhanced the recording and playback quality of magnetic tape, marking the official inception of magnetic tape-recording technology [ 1 ] . Analog audio tapes convert sound signals into continuously varying analog electrical signals and subsequently employ magnetic heads to record these electrical signals as variations in magnetism on the coating of the magnetic tape. Since 1947, it has been widely used in music dissemination, cultural popularization, scientific research, news gathering, etc., and has played an indispensable role in all aspects of people's lives and work [ 2 ] . The primary components of the tape consist of the magnetic layer and the substrate, while a back coating was introduced in the 1960s to enhance wear resistance [ 3 ] . The magnetic layer is coated with a polyester-polyurethane adhesive with uniform distribution of magnetic particles. The substrate is the support of the magnetic layer, which provides mechanical properties and is responsible for the physical integrity of the tape, being mostly polyester film (polyethylene terephthalate) [ 4 ] . The polyester polyurethane binders contained in the tapes, as well as other organic substances such as fatty acid esters, paraffin oils, siloxanes and fluorinated lubricants, are the main sources of nutrition for mildew [ 5 , 6 ] . Aspergillus and Penicillium are among the most common varieties in audiovisual materials [ 7 , 8 ] . Other species of mildew including Alternaria , Chaetomium , Stemphylium , Cladosporium , and Streptomyces , are also frequently found on audiovisual materials. Most of the mildew is distributed on the edge of the substrate and on the surface of the magnetic layer (Fig. 1 ). Mildew growth can significantly affect the deterioration of magnetic media, thereby impairing the retrieval of information [ 9 ] . Additionally, mildew has the potential to contaminate other archival materials by releasing spores, which may also pose a health risk to archivists [ 10 ] . Current research on magnetic tapes has centered on the assessment of their chemical and physical properties. For instance, the morphology of the tapes, layer thickness, and physical damage were analyzed using Scanning Electron Microscopy (SEM) and Environmental Scanning Electron Microscopy (ESEM). This methodology was further enhanced by employing X-ray fluorescence (XRF) spectroscopy. The detection of chlorine facilitated the identification of PVC substrates, while the presence of silicon or fluorine provided insights into lubricant composition. This comprehensive approach enabled a preliminary analysis of tapes with incomplete documentation [ 11 – 13 ] . The primary mechanism of tape degradation and limited lifespan is the transformation of metal particles from ferromagnetic to non-ferromagnetic iron oxide [ 14 ] . Spectroscopic analysis has proven instrumental in characterizing magnetic tape materials. an extensive library of ATR spectral data was established to identify characteristic bands for tape substrates and binders [ 15 – 18 ] . Building on this foundation, Hobaia demonstrated that ATR-FTIR spectroscopy could differentiate between SSS and non-SSS audio tapes and serve as a rapid tool for predicting the playability of tapes with no visible signs of aging. A methodology integrating Principal Component Analysis (PCA), Quadratic Discriminant Analysis (QDA), and K-means clustering was employed to classify infrared spectra based on playability. This approach successfully identified 93.78% of unplayable tapes within the calibration set. Thermal analysis provides another avenue for material identification and assessment. TGA degradation temperatures can assist in identifying the materials of unknown tapes. Davis further characterized the degraded tapes using DSC, observing an irreversible endothermic transition at approximately 50 ℃, which is typically absent in playable tapes. The degradation of the adhesive can lead to diseases such as cracking of the magnetic layer and shedding of magnetic powder. Meanwhile, investigations into binder degradation have employed both destructive and non-destructive techniques. Acetone can used to destructively extract soluble compounds formed by binder hydrolysis, thereby studying the kinetics of polyester polyurethane binder breakdown [ 19 , 20 ] . Headspace Solid-Phase Microextraction-Gas Chromatography-Mass Spectrometry (SPME-GC-MS) serves as a non-destructive method for profiling the Volatile Organic Compounds (VOCs) emitted by tapes following decades of both natural and artificial aging [ 21 ] . Despite these advanced diagnostic techniques, a critical gap remains: the impact of various degradation types on the actual audio quality is severely under-researched. Sound recording archives research on the effects of mildew on audio quality is still very limited. There is an urgent need to establish a parameter system for objectively evaluating sound quality changes. Such a system is crucial for assessing the effectiveness of tape restoration methods and, ultimately, for improving the recovery of our audio heritage. Building upon existing research, this study employs an integrated approach combining chemical characterization and acoustic measurement techniques to systematically investigate mildew erosion in audio materials. Through controlled simulation of mildew-contaminated media samples, the research establishes a scientific framework for objectively assessing mildew covers on sound quality preservation. 2. Material and methods 2.1 Material An analog magnetic audiotape, distributed by Tianjin Audiovisual Co., Ltd. in 1991, was used. Potato Dextrose Agar (PDA) medium from Beijing Aobo Star Biotechnology Co., LTD. Lactic acid phenol cotton blue staining agent from Beijing Solaibao Technology Co., LTD. Penicillium and aspergillus versicolor were isolated and purified from the mildew on the surface of the audio tape. 2.2 Analog sample preparation for mildew Preparation for the bacterial suspension Penicillium and Aspergillus were cultured on PDA medium for a duration of 5 days. The mildew colony, along with a small amount of agar, was carefully scraped into a sample tube using an inoculation loop. Subsequently, normal saline was added to the tube, and the conidia of the mildew were thoroughly dispersed in the saline by shaking vigorously for 5 minutes. Following this process, the conidia were uniformly suspended in normal saline, and a bacterial suspension was obtained through simple filtration using gauze. The inclusion of a small amount of agar serves to provide essential nutrients, promote mildew growth on the tape, and reduce the preparation time required for creating mock samples intended for mildew attack testing. Preparation for mildew simulation samples First, play the audio tape and select playback time than 5 s and label D0. The audio sample is labeled R-D0. Subsequently, the bacterial suspension was applied to sample D0. The tape was laid flat on the bench with the magnetic layer facing upward. Using a cotton swab, the bacterial suspension was carefully applied to the sample. Care was taken to control the volume of the applied solution and to prevent droplets from running off due to the smooth surface of the tape. After treatment, placed the sample in a well-ventilated area to dry. The same procedure was repeated on the audio tape at one-day intervals. Audio tapes exhibiting progressively increasing mildew coverage were designated as samples D1, D2, D3, D4, and D5, as illustrated in Fig. 2 . The corresponding audio recordings were labeled R-D1, R-D2, R-D3, R-D4, and R-D5. 2.3 Surface topography characterization of mildew simulation samples Epson scanner (Epson Perfection V850 Pro Seiko Epson Co., Ltd.) was used to perform 8-bit scanning of the surface morphology of D0, D1, D2, D3, D4 and D5, with a background plate selected from a black cardboard sheet, to record the surface topography of the audio tape with different coverage degrees was recorded, and to analyze the gray-scale histograms extracted from the scanned photographs. Laser confocal scanning microscopy was used to characterize the mildew covered samples by roughness analysis. The VK-H1XMC multi-file analysis software was used to measure surface roughness parameters of mildew covered samples D0, D1, D2, D3, D4 and D5. Five points were taken for each sample, observed at every 1/6 of the length of samples, avoiding the edge of the tape. Each sample measurement data was subsequently analyzed using five key surface roughness parameters: texture aspect ratio of the surface (Str), peak density of the surface (Spc), developed interfacial area ratio (Sdr), maximum height (Sz), and arithmetic mean height (Sa). In the equations, z (x, y) denotes the signed normal distance between the reference surface and the scale-limited surface. Symbol A denotes the evaluated area's numerical measure. $$\:\varvec{S}\varvec{a}=\frac{1}{\varvec{A}}\underset{\varvec{A}}{\overset{}{\iint\:}}\left|\varvec{Z}\left(\varvec{x},\varvec{y}\right)\right|\varvec{d}\varvec{x}\varvec{d}\varvec{y}$$ $$\:\varvec{S}\varvec{z}=\varvec{S}\varvec{p}+\varvec{S}\varvec{v}$$ $$\:\varvec{S}\varvec{p}\varvec{c}=-\frac{1}{2}\frac{1}{\varvec{n}}\sum\:_{\varvec{k}=1}^{\varvec{n}}(\frac{{\partial\:}^{2}\varvec{z}\left(\varvec{x},\varvec{y}\right)}{\partial\:{\varvec{x}}^{2}}+\frac{{\partial\:}^{2}\varvec{z}\left(\varvec{x},\varvec{y}\right)}{\partial\:{\varvec{y}}^{2}})$$ $$\:\varvec{S}\varvec{t}\varvec{r}=\frac{\underset{{\varvec{\tau\:}}_{\varvec{x},}{\varvec{\tau\:}}_{\varvec{y}}\in\:\varvec{R}}{\mathbf{min}}\sqrt{{{\varvec{\tau\:}}_{\varvec{x}}}^{2}+{{\varvec{\tau\:}}_{\varvec{y}}}^{2}}}{\underset{{\varvec{\tau\:}}_{\varvec{x},}{\varvec{\tau\:}}_{\varvec{y}}\in\:\varvec{Q}}{\mathbf{max}}\sqrt{{{\varvec{\tau\:}}_{\varvec{x}}}^{2}+{{\varvec{\tau\:}}_{\varvec{y}}}^{2}}}$$ $$\:\varvec{S}\varvec{d}\varvec{r}=\frac{1}{\varvec{A}}\underset{\varvec{A}}{\overset{}{\iint\:}}（\sqrt{\left[1+（{\frac{\partial\:\varvec{z}\left(\varvec{x},\varvec{y}\right)}{\partial\:\varvec{x}}）}^{2}+（{\frac{\partial\:\varvec{z}\left(\varvec{x},\varvec{y}\right)}{\partial\:\varvec{y}}）}^{2}\right]}-1）\varvec{d}\varvec{x}\varvec{d}\varvec{y}$$ Arithmetic Mean Height (Sa ) is defined as the arithmetic mean of the absolute deviations in vertical distance between the measured surface profile and the reference plane within the sampling area. As a height parameter, a higher Sa value indicates greater amplitude variation and increased surface roughness. Maximum Height (Sz), represents the sum of the highest peak and deepest valley within the sampling area. This parameter quantifies the extreme vertical deviations of the surface. Texture Aspect Ratio (Str) is a spatial parameter that characterizes surface isotropy or anisotropy, where λ min and λ max denote the dominant texture wavelengths perpendicular and parallel to the lay direction. Values approaching 0 indicate striated patterns, Str ≈ 1 denotes isotropic topography. \(\:\varvec{S}\varvec{t}\varvec{r}=\frac{\varvec{\lambda\:}\varvec{m}\varvec{i}\varvec{n}}{\varvec{\lambda\:}\varvec{m}\varvec{a}\varvec{x}}\) (0 \(\:\le\:\varvec{S}\varvec{t}\varvec{r}\le\:1\) ) Arithmetic Mean Peak Curvature (Spc) calculates the average principal curvature of the surface summits. As a feature shape parameter, lower Spc values indicate blunt or rounded asperities, whereas higher values correspond to sharp surface features that influence contact mechanics. Developed Interfacial Area Ratio (Sdr) is a hybrid parameter that quantifies the percentage increase in surface area relative to an ideal flat plane: $$\:\varvec{S}\varvec{d}\varvec{r}=\left(\frac{\varvec{A}\varvec{a}\varvec{c}\varvec{t}\varvec{u}\varvec{a}\varvec{l}-\varvec{A}\varvec{p}\varvec{r}\varvec{o}\varvec{j}\varvec{e}\varvec{c}\varvec{t}\varvec{e}\varvec{d}}{\varvec{A}\varvec{p}\varvec{r}\varvec{o}\varvec{j}\varvec{e}\varvec{c}\varvec{t}\varvec{e}\varvec{d}}\right)\times\:100\varvec{\%}$$ Where A actual is the true surface area and A projected is the projected area. An Sdr value > 0 reflects surface porosity and roughness complexity, while Sdr = 0 indicates perfect flatness. 2.4 Acoustic characterization of simulation samples Portable tape recorder (6503, Nanjing Panda Electronics Co., Ltd.) was used to record audio in MP3 format before and after processing. Praat speech analysis software (version Intel64) developed by the Institute of Phonetic Sciences at the University of Amsterdam's Faculty of Humanities was used. Praat is a comprehensive speech analysis software package developed and maintained by the Institute of Phonetic Sciences at the University of Amsterdam [ 22 ] . It enables a wide range of acoustic-phonetic analyses, including but not limited to fundamental frequency (F0), spectral characteristics, and formant tracking, in addition to voice analysis. Wideband spectrogram is a type of spectrogram optimized for high time resolution at the expense of frequency resolution. The center of gravity ( \(\:{\text{f}}_{\text{c}}\) ) is a measure for how high the frequencies in a spectrum are on average, expressed as \(\:{\text{f}}_{\text{c}}\) in Hz. $$\:{\varvec{f}}_{\varvec{c}}={\int\:}_{0}^{\varvec{\infty\:}}{\varvec{f}\mid\:\varvec{S}\left(\varvec{f}\right)\mid\:}^{\varvec{P}}\varvec{d}\varvec{f}$$ The standard deviation ( \(\:\text{S}\text{D}\) )is a measure for how much the frequencies in a spectrum can deviate from the center of gravity. Standard deviation represents the standard deviation in the spectrum, denoted by SD, and is used to measure the degree of dispersion of the frequency with respect to. The skewness ( \(\:{\text{S}}_{\text{k}}\) ) is a measure for how much the shape of the spectrum below the center of gravity is different from the shape above the mean frequency. Skewness represents the skewness in the spectrum, and is used to measure the direction and degree of skewness of the data distribution (relative to the standard normal distribution). The kurtosis ( \(\:{K}_{u}\) ) is a measure for how much the shape of the spectrum around the center of gravity is different from a Gaussian shape. Kurtosis denotes the degree of kurtosis in the spectrum, and is used as a measure of the sharpness of the data distribution (relative to the standard normal distribution), with a high kurtosis usually indicating a sharper distribution curve, often with more extreme values. The band energy difference ( \(\:\text{B}\text{E}\text{D}\) ) represents the energy difference between the low-frequency band (20 ~ 3000 Hz) and the high-frequency band (3000 ~ 6000 Hz). 3. Results and Discussion In Fig. 2 a, the characteristic peak at 630.8 cm − 1 in the FTIR spectrum is attributed to the lattice vibration of Fe-O, indicating that the audio tape contains γ-Fe 2 O 3 particles. The peak at 1060.4 cm − 1 is due to the stretching vibration of the C-O bond, and the peak at 1170.4 cm − 1 is caused by the stretching vibration of the C-O bond connected to the acetyl group. The peak at 1725.5 cm − 1 is attributed to the stretching vibration of the ester group (C = O), and the peak at 2921.6 cm − 1 is the absorption peak caused by the -CH 2 in the polyester [ 23 , 24 ] , suggesting that the binder is a polyester type. Combined with the manufacturing process of the audio tape, it is speculated that the audio tape selected for this experiment contains a polyester polyurethane binder. It can be seen from Fig. 2 b, the peak at 723.6 cm − 1 is attributed to the out-of-plane bending vibration of the benzene ring, the peak at 1097.5 cm − 1 is due to the symmetrical stretching vibration of the C-O bond in the ethylene glycol segment, the peak at 1244.5 cm − 1 is attributed to the asymmetrical stretching vibration of C-O-C, and the peak at 1714.1 cm − 1 is due to the stretching vibration of the ester group (C = O), indicating that the audio tape selected for this experiment is made of PET (polyethylene terephthalate). Figure 2 c and 2 d shows the SEM micrographs of the cross-section of the audio tape. It shows that the audio tape is composed of three layers with an average thickness of approximately 3.5 µm, 8.0 µm, and 2.5 µm (from left to right). The energy spectrum indicates that the main element of the first layer structure is Fe, while the main elements of the second- and third-layer structures are C. From the comprehensive Fig. 2 a-d, it can be concluded that the three layers from left to right are the magnetic layer, the base layer, and the back coating layer, respectively. Figure 3 shows the 2D surface morphology of localized areas of audio samples after mildew erosion during different stages of fungal erosion. The surface of the magnetic layer without mildew erosion (D0) is relatively smooth and contains fine black magnetic particles. When the mildew begins to erode the magnetic layer (D1), long and thin hyphae emerge, arranged in a grid pattern [ 25 ] . As time progresses and spore concentration increases (D2-D5), patchy aggregated colonies gradually form, leading to an increase in coverage of the magnetic layer. Notably, the mildew colonies on the D5 sample nearly encompass the entire surface of the magnetic layer [ 26 ] . As illustrated in Fig. 4 , the grayscale histograms representing the gray levels of the scanned image and the corresponding number of pixels was used to represent the varying coverage degrees of the mildew simulation samples. The pixel distribution of the sample U0 was the most concentrated, mainly around the gray level of 25–50, the corresponding number of pixels was about 14 x 10 5 . When the sample is covered with mildew (D1), there is a noticeable decrease in the number of pixels within the gray level range of 25–50. This change is accompanied by a shift in the peak shape towards higher gray values, while an increase in the number of pixels within the gray level range of 50–100 becomes evident. These observations indicate that the overall image brightness increases, resulting in diminished detail in darker areas. As the degree of coverage progresses from slight (D2) to severe (D5), there is a gradual decline in pixel count for the gray level range of 25–50, contrasted by a steady rise in pixel count for the gray level range of 50–100. This phenomenon can be attributed to the higher whiteness associated with mildew compared to that of the black magnetic layer found on audio tape. Consequently, as mildew coverage intensifies, there is an overall increase in sample whiteness, leading to enhanced brightness within the image [ 25 ] . As shown in Fig. 5 , the trend of average surface roughness change of mildew-covered samples at varying coverage levels can be clearly seen. The average Sa value of the surface of sample D0 is 2.34 µm (Fig. 5 a). As the degree of mildew coverage increases from slight (D1) to severe (D5), the Sa value exhibits a monotonically increasing trend, indicating greater amplitude variation and enhanced surface roughness. Figure 5 b illustrates the variation in Sz values for samples with differing degrees of mildew coverage. As the degree of mildew coverage progresses from D0 to D5, Sz demonstrates a clear upward trend. This suggests that increased mildew coverage results in an elevation in both the maximum height and deepest valley on the surface, revealing a strong positive correlation between Sz, Sa. Upon examining Fig. 5 c, it is evident that the Str values for all mildew-covered samples (D0 through D5) exceed 0.5 [ 27 ] , signifying that these sample surfaces exhibit more isotropic characteristics. Figure 5 d and 5 e have similar variation patterns. The changes from D1 to D3 are relatively gentle, and the overall trend is monotonically increasing. These results suggest that the growth of mildew on the magnetic layer surface is uneven, and the masking effect on the local magnetic layer is different [ 28 ] . Figure 6 illustrates the wideband spectrograms of audio samples subjected to varying degrees of mildew coverage. These spectrograms are commonly employed to depict the energy distribution of sound signals across different frequencies over time. The depth of color typically signifies the intensity of energy, with darker hues indicating higher energy levels and lighter shades representing lower energy levels. From the wideband spectrogram corresponding to R-D0, it is evident that color depth fluctuates across various frequency bands. Specifically, in the low-frequency range of 1000–2000 Hz, the coloration is darker compared to that in the high-frequency range, suggesting a predominant concentration of energy within this lower frequency spectrum. This phenomenon can be attributed to low-frequency sound waves possessing longer wavelengths, which render them more susceptible to diffraction or reflection; thus, they retain their energy during propagation [ 29 ] . When mildew begins to cover the sample (R-D1), all frequency bands exhibit a noticeable lightening in color within a span of 5 seconds, accompanied by a decrease in total or average sound energy. This observation indicates that mildew coverage diminishes both overall amplitude and volume of sound-transforming it from "full-bodied" (characterized by rich harmonics typical in music) into a "thin" quality-and results in reduced speech clarity. As surface coverage escalates from slight (R-D0) to severe (R-D5), there is an increasingly pronounced lightening effect observed across all frequency bands within 5 seconds. Notably, for sample R-D5, nearly all high-frequency energies ranging from 2000–6000 Hz dissipate entirely while some degree of intensity persists within the low-frequency range. This suggests that high-frequency waves-with their shorter wavelengths-are more readily obstructed or scattered by mildew particles than their low-frequency counterparts, consequently leading to a swifter loss of energy [ 30 ] . High-frequency sounds contribute significantly to attributes such as brightness, clarity, and detail within auditory perception. The absence of energetic presence in this high-frequency domain manifests as continued existence but with timbral alterations shifting from "bright" towards "dull," further accompanied by diminished clarity and loss of intricate details. As illustrated in Fig. 7 , the intensity curves of audio samples were exhibited with varying mildew-covered levels. In comparison to R-D0, the presence of mildew coverage diminishes the amplitude of the sound wave. The severity of this coverage correlates with a reduction in audio energy, with these changes being particularly pronounced at both peaks and troughs. This phenomenon occurs because as the magnetized audio tape moves toward the playback head, the fluctuating magnetic field generates magnetic flux, resulting in an output electrical signal that continuously varies to record information. When there is close contact between the head and the magnetic tape, magnetic coupling is enhanced. Consequently, more magnetic field lines traverse through the head coil, generating a stronger induced signal. However, increased mildew coverage elevates the distance between the playback head and the magnetic layer, thereby decreasing the number of magnetic field lines passing through the head coil. This results in a weakened induced signal and subsequently reduces sound volume. Furthermore, mildew coverage can alter both peak height and shape within intensity curves. For instance, as mildew coverage intensifies, a curve peak occurring around 1s becomes higher and sharper. It was indicated that sound transitions from low to high levels during this interval which may lead to noise formation. Moreover, the concentrated frequency point (peak) of energy at 2s shifts to the low-frequency direction, which may cause the attack time to be advanced. Using Praat software to perform a Fourier transform allows for the decomposition of complex sound waves into simpler components, which are then distributed across various frequency bands to create a formant chart [ 22 ] . The horizontal axis represents time, while the vertical axis indicates the frequency of the sound wave. Formants correspond to specific regions in the sound spectrum where energy is relatively concentrated, and they appear as thick black "horizontal bars" in the spectrogram [ 31 ] . Figure 8 illustrates the formant charts of audio samples exhibiting varying degrees of mildew coverage. The audio sample R-D0 displays relatively concentrated distributions of sound wave frequencies at 900–1100 Hz, 2900–3100 Hz, 3900–4100 Hz, and 5000–5200 Hz. As mildew coverage increases (D1 - D5), the high-frequency changes are obvious, but the shape and distribution range of the formants do not change much overall. On one hand, mildew presence obstructs direct contact between the magnetic head and magnetic layer, leading to attenuation of remanent magnetic signals read by the magnetic head. Given that high-frequency signals possess shorter wavelengths, these changes become more pronounced at higher frequencies. On the other hand, the presence of hypha reduces the smoothness of the surface of the audio tape, and the local magnetic field distribution changes during playback, causing changes in the frequency distribution. However, since the magnetic domain disorder is random, the impact on signals of different frequencies is relatively consistent, so the shape and distribution range of the formants do not change much overall. Figure 9 a illustrates the variations in 5s spectrogram parameters (center of gravity, standard deviation, skewness, kurtosis and band energy difference) of mildew-covered audio samples under varying coverage levels. As the coverage changes from slight (D0) to severe (D5), center of gravity and standard deviation of the 5s audio spectrogram decrease monotonically, and the skewness and kurtosis increase monotonically. This is because the overlay process attenuates the remanence signal read by the magnetic head, some of the sound information lost. The loss of part of the frequency information during reading directly leads to a decreasing trend in the standard deviation of the spectrogram, and the curve becomes sharper and sharper, which is manifested as higher and higher kurtosis. Due to the short wavelength of the high frequency, the coverage treatment has a greater impact on the high frequency region, so the average frequency shows a decreasing trend with the increase of mildew coverage [ 32 ] . In addition, due to the higher energy in the low-frequency region, more information in the low-frequency region is lost due to the coverage processing, so the peak area in the spectrogram is positively biased and the skewness increases. As evidence from the Fig. 10 b, with the increase of coverage, the energy difference between the low frequency region (20-3000 Hz) and the high frequency region (3000–6000 Hz) becomes larger, indicating that the treatment of surface anisotropy will increase the energy difference to a certain extent. As shown in Fig. 10 , surface morphology of the mildew corrosion sample (Fig. 10 a and a 1 ) and sample after removing mildew (Fig. 10 b and b 1 ) reveal notable. In Fig. 10 a and a 1 , fungal hyphae proliferate across the tape surface, forming an interconnected network. The colonies form pitting on the surface of the magnetic layer. It is worth noting that after removing the mildew, we observe not only localized pitting but also extensive irregular line-like corrosion in severely affected areas on the magnetic layer's surface (Fig. 10 b). As can be seen from the 3D surface morphology of sample after removing mildew (Fig. 10 b1), these pitting or linear corrosion patterns exhibit a concave structure. Based on the SEM cross-section image of mildew sample (Fig. 10a2), it can be concluded that during its reproductive process, mildew mycelium penetrates into the magnetic layer, resulting in both pitting and linear concave structures due to corrosion. This suggests that as mildew metabolizes, it produces organic acids which corrode the recording medium, thereby compromising the durability of the magnetic carrier. Mildew degradation affects components of this medium leading to structural damage. Additionally, it causes surface pitting and movement of magnetic particles which ultimately results in distortion and attenuation of magnetic recording signals. 4. Conclusions In this study, a combination of chemical characterization and acoustic measurement to systematically evaluate the impact of mildew contamination on audio materials was explored. The common strain Penicillium and Aspergillus was used to prepare mildew corrosive samples with different coverage degrees, and the corresponding audio samples were analyzed by Praat. Refine the data on surface structure and changes in acoustic information, it was found that the coverage of mycelium on the magnetic layer surface was uneven and would cause the image to brighten. On the one hand, the energy was mainly concentrated in the low-frequency range (1000–2000). The coverage of mildew would cause the total energy or average energy of the sound would decrease. This indicates that the covering effect of mildew would reduce the overall amplitude or volume of the sound and decrease the clarity of speech. Notably, the energy loss in the high-frequency range was faster than that in the low-frequency range, resulting in detail loss and reduced intelligibility. On the other hand, it would cause changes in the peak values of the sound intensity curve or shifts in the peak shape, leading to noise or alterations in the attack time. From the parameter change graph of the 5-second spectrum of the sample, it can be concluded that the average frequency and standard deviation monotonically decreased, while the skewness and kurtosis monotonically increased. The above information indicates that the covering effect of mildew increases the distance between the magnetic head and the magnetic layer, resulting in weakened induction signals and the loss of some sound information. Combining the surface and cross-sectional morphology and roughness information, mildew degradation affects the components of this medium, leading to surface pitting and movement of magnetic particles, which ultimately results in distortion and attenuation of magnetic recording signals. Based on the information presented above, the dual effects of mildew coverage and corrosion cause a decrease in sound clarity and distortion. This work highlights the necessity of mildew removal and provides valuable theoretical references for the evaluation of sound information after the removal of contaminants from magnetic materials in the future. Declarations Competing Interests The authors declare no competing interests. Funding The authors acknowledge financial support from the National Natural Science Foundation of China (22572112), National Natural Science Foundation of China (22002080), Science and Technology Project of the National Archives Administration (2022-B-005), The Key Research and Development Program of Shaanxi Province (2021GY-172), The Key Scientific Research Project at the Museum Level of Hubei Provincial Museum (25A05) Author Contribution J.Z.H. and C.L. provided research design, research guidance, data analysis and writing Original Draft; X.H.P. and L.Y.H. provided research guidance; W.Y.N. and Z.Y.H. participated in data analysis; Y.S.J. was responsible for research design and data collection; C.Q.F. and L.B.D involved in data collection. References Bressan F, Hess RL, Sgarbossa P. Chemistry for Audio Heritage Preservation: A Review of Analytical Techniques for Audio Magnetic Tapes[J]. Heritage. 2019;2(2):1551–87. Luo LH. Development and Applications of Magnetic Recording[J]. Wuli (Physics). 1984;13(3):145–9. Bereijo A. The conservation and Preservation of Film and Magnetic Materials (2): Magnetic Materials[J]. Libr Rev. 2004;53(7):372–8. Lantz MA, Furrer S, Peterman M. Magnetic Tape Storage Technology[J]. ACM Trans Storage. 2025;21(1):1–70. Bogart JWV. Magnetic Tape Storage and Handling: A Guide for Libraries and Archives[M]. ERIC. 1995;1:34. Heitkamp MA, Freeman JP, McMillan DC. Fungal Metabolism of tert-Butylphenyl Diphenyl Phosphate[J]. Appl Environ Microbiol. 1985;50(2):265–73. Cappitelli F, Sorlini C. From Papyrus to Compact Disc: The Microbial Deterioration of Documentary Heritage[J]. Crit Rev Microbiol. 2005;31(1):1–10. Vivar I, Borrego SGE. Fungal Biodeterioration of Color Cinematographic Films of the Cultural Heritage of Cuba[J]. Int Biodeterior Biodegrad. 2013;84(3):372–80. Soleymani S, Russ L. Mould on Magnetic Media: What Are the Current Preservation Practices by Audiovisual Conservation Practitioners?[J]. Stud Conserv. 2023;68(7):704–19. Richardson E, Giachet M, Schilling M. Assessing the Physical Stability of Archival Cellulose Acetate Films by Monitoring Plasticizer Loss[J]. Polym Degrad Stab. 2013;107(2):231–6. Bressan F, Bertani R, Furlan C, An ATR-FTIR. ESEM Study on Magnetic Tapes for the Assessment of the Degradation of Historical Audio Recordings[J]. J Cult Herit. 2016;18(1):313–20. Bressan F, Canazza S, Bertani R. A study on Thermal Treatment for the Recovery of Magnetic Tapes Affected by Soft Binder Syndrome-Sticky Shed Syndrome[J]. IASA J. 2015;44(1):53–64. Davis AR, Monroe E, France FG. Understanding Magnetic Tape Degradation by Polymeric and Material Testing[C]. Proceedings of 2018 AES International Conference on Audio Archiving, Preservation & Restoration, Audio Engineering Society , 2018. Judge JS, Schmidt RG, Weiss RD. Media Stability and Life Expectancies of Magnetic Tape for Use with IBM 3590 and Digital Linear Tape Systems[C]. Proceedings of the 20th IEEE/11th NASA Goddard Conference on Mass Storage Systems and Technologies (MSST 2003) , 2003, 97–100. Gómez-Sánchez E. ATR-FTIR Spectroscopy for the Characterization of Magnetic Tape Materials[J]. E Preservation Sci. 2011;8(1):2–9. Cassidy BM, Lu Z, Fuenffinger NC. Minimally Invasive Identification of Degraded Polyester-Urethane Magnetic Tape Using Attenuated Total Reflection Fourier Transform Infrared Spectroscopy and Multivariate Statistics[J]. Anal Chem. 2015;87(18):9265–72. Hobaica S. Analysis of Audio Magnetic Tapes with Sticky Shed Syndrome by ATR-FTIR[J]. J Appl Polym Sci. 2013;128(3):1962–73. Yang WG, Ha JH, Kim SG. Spectroscopic Determination of Alkyl Resorcinol Concentration in Hydroxyapatite Composite[J]. J Anal Sci Technol. 2016;7(1):9–16. Katayama K, Chinda Y, Shimizu O. Long-Term Stability of Magnetic Tape for Data Storage Under an Accelerated Condition[J]. IEEE Trans Magn. 2016;52(7):1–4. Thiébaut B, Vilmont LB, Lavédrine B. Characterization of U-matic videotape Deterioration by Size Exclusion Chromatography and Pyrolysis Gas Chromatography/mass Spectrometry and the Role of Adipic Acid[J]. J Cult Herit. 2009;10(2):183–97. Thiébaut B, Lattuati-Derieux A, Hocevar M. Application of Headspace SPME-GC-MS in Characterisation of Odorous Volatile Organic Compounds Emitted from Magnetic Tape Coatings Based on Poly(urethane-ester) After Natural and Artificial Ageing[J]. Polym Test. 2007;26(2):243–56. Boersma P, Weenink D. Praat, a System for Doing Phonetics by Computer[J]. Glot Int. 2001;5(3):341–5. Bressan F, Bertani R, Furlan C. An ATR-FTIR and ESEM Study on Magnetic Tapes for the Assessment of the Degradation of Historical Audio Recordings[J]. J Cult Herit. 2016;18(3):313–20. Cassidy BM, Lu Z, Fuenffinger NC. Minimally Invasive Identification of Degraded Polyester-Urethane Magnetic Tape Using Attenuated Total Reflection Fourier Transform Infrared Spectroscopy and Multivariate Statistics[J]. Anal Chem. 2015;87(18):9265–72. Watsky L. Mildew Treatment for Analog Audio Tapes: A Technical Guide[C]. Proceedings of the Association for Recorded Sound Collections (ARCS) Annual Conference , 2014. Pelaez J, Torres C, Ortiz JE. Efecto de Dos Campos Magnéticos sobre el Crecimiento Micelial y Propiedades Sensoriales del Hongo Pleurotus Ostreatus[J]. Revista Colombiana de Ciencias Hortícolas. 2013;7(1):89–97. Hussein I, Hasan R. The Effect of Zirconium Oxide Nanoparticle on the Tear Strength of Maxillofacial Silicone[J]. Al-Rafidain Dent J. 2021;21(3):193–201. Rashidi M. Effect of Zirconium Oxide-Titanium Dioxide Nanoparticles on Mechanical and Physical Properties of Soft Denture Lining Materials[J]. J Nanostruct. 2022;12(1):1–8. Oelze T. Sound Reflection and Energy Loss at Boundaries[J]. Appl Acoust. 2018;137(3):211–7. Inspur General Software Co. Ltd. Lightweight High-Frequency Speech Restoration Method, System, Device, and Medium: CN118072750A[P]. 2024. King R, Grau-Bové J, Curran K. Plasticiser Loss in Heritage Collections: Its Prevalence, Cause, Effect, and Methods for Analysis[J]. Herit Sci. 2020;8(1):123–31. Liu QQ, Li XG. Research Progress on Mildew Corrosion of Metals and Their Protective Layers in Atmospheric Environment[J]. J Univ Sci Technol Beijing. 2017;39(10):1463–9. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 09 Mar, 2026 Reviews received at journal 18 Feb, 2026 Reviewers agreed at journal 27 Jan, 2026 Reviewers agreed at journal 23 Jan, 2026 Reviewers invited by journal 23 Jan, 2026 Editor assigned by journal 19 Jan, 2026 Submission checks completed at journal 19 Jan, 2026 First submitted to journal 13 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8597238","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":580630644,"identity":"e1f58ff5-db0f-4927-adf4-37222b80c595","order_by":0,"name":"Zhihui Jia","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIiWNgGAWjYBACxgYehgMfKmzkoHxmorQwHpxxJs2YeC0MDDzMh3nbDic2EK2Fuf/sgcM8Z9LS58/uPSbBUGGd2MB+9gABh51LODinwiZ3w51zaRIMZ9ITG3jyEvBraewxOPDmTFruBokcMwlGkAsleAzwa2nmMTgA9Eu6/AyQln/EaGnjMTgI1JLAcAOkpYEYLT1ALcBANtxwI8fYIuFYunEbTw5+LYb9Z4w/AKNSHugwwxsfaqxl+9nPENDSgMxLAGI2vOqBQJ6QglEwCkbBKBgFDADx+EklnyrdQQAAAABJRU5ErkJggg==","orcid":"","institution":"Shaanxi Normal University","correspondingAuthor":true,"prefix":"","firstName":"Zhihui","middleName":"","lastName":"Jia","suffix":""},{"id":580630645,"identity":"c64a7c92-b314-42f9-8f24-25700068630b","order_by":1,"name":"Yanan Wang","email":"","orcid":"","institution":"Shaanxi Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yanan","middleName":"","lastName":"Wang","suffix":""},{"id":580630647,"identity":"34121954-6016-4f31-b0ce-35e6a83c9ed1","order_by":2,"name":"Shujiao Yu","email":"","orcid":"","institution":"Shaanxi Normal University","correspondingAuthor":false,"prefix":"","firstName":"Shujiao","middleName":"","lastName":"Yu","suffix":""},{"id":580630648,"identity":"3dbaf27c-6f40-4328-acc4-3186ad8ac76a","order_by":3,"name":"Quanfeng Cai","email":"","orcid":"","institution":"Shaanxi Normal University","correspondingAuthor":false,"prefix":"","firstName":"Quanfeng","middleName":"","lastName":"Cai","suffix":""},{"id":580630649,"identity":"a7340602-4be0-4628-8492-cb823e7d842c","order_by":4,"name":"Bodian Liu","email":"","orcid":"","institution":"Shaanxi Normal University","correspondingAuthor":false,"prefix":"","firstName":"Bodian","middleName":"","lastName":"Liu","suffix":""},{"id":580630650,"identity":"dbb58b8e-7917-4924-a000-d24642653be7","order_by":5,"name":"Huiping Xing","email":"","orcid":"","institution":"Shaanxi Normal University","correspondingAuthor":false,"prefix":"","firstName":"Huiping","middleName":"","lastName":"Xing","suffix":""},{"id":580630651,"identity":"bb725d12-9144-4cee-b7e5-90811b581146","order_by":6,"name":"Yuhu Li","email":"","orcid":"","institution":"Shaanxi Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yuhu","middleName":"","lastName":"Li","suffix":""},{"id":580630652,"identity":"b0b04a3b-e135-4e2a-a6bd-62d1178143a4","order_by":7,"name":"Long Chen","email":"","orcid":"","institution":"tate Key Laboratory of Loess Science, Institute of Earth Environment, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Long","middleName":"","lastName":"Chen","suffix":""},{"id":580630653,"identity":"90fac763-071c-40b2-955b-e53cce58fd98","order_by":8,"name":"Yanhong Zhao","email":"","orcid":"","institution":"Hubei Provincial Museum","correspondingAuthor":false,"prefix":"","firstName":"Yanhong","middleName":"","lastName":"Zhao","suffix":""}],"badges":[],"createdAt":"2026-01-14 04:08:35","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8597238/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8597238/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101301196,"identity":"fe886028-635d-42b5-9282-2d46f4db2430","added_by":"auto","created_at":"2026-01-28 09:51:09","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":63427,"visible":true,"origin":"","legend":"\u003cp\u003eThe\u003cstrong\u003e \u003c/strong\u003ephotos of Mildewy audio tapes (a: at the edge of the substrate: b: on the surface of the magnetic layer)\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/6944ec320f8f19be465f7e38.jpg"},{"id":101302586,"identity":"a6caee40-d54c-432a-97ec-67c708be9fe1","added_by":"auto","created_at":"2026-01-28 09:54:22","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":98902,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 3.\u003c/strong\u003e Schematic diagram of mildew simulation samples\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/9d0fa52cfa89c8ba9622139c.jpg"},{"id":101301476,"identity":"03d04e12-2055-4dae-bbd2-60c2a23b887d","added_by":"auto","created_at":"2026-01-28 09:51:56","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":88188,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 2. \u003c/strong\u003eFTIR spectra\u003cstrong\u003e \u003c/strong\u003eand\u003cstrong\u003e \u003c/strong\u003eSEM micrographs of tape (a: magnetic layer; b: substrate; c: the cross-section of the audio tape; d: The mapping of elements in the cross-section of the audio tape)\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/cd7a4ae96fc52127b5396549.jpg"},{"id":101301383,"identity":"51a75793-37ab-49a2-baf0-e28ff71d9a50","added_by":"auto","created_at":"2026-01-28 09:51:41","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":246245,"visible":true,"origin":"","legend":"\u003cp\u003eFig. 3. 2D surface morphology of mildew simulation samples\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/984b13700e87178dc1c86914.jpg"},{"id":101301515,"identity":"51ab3063-f9e9-4c82-94b9-94774bbaddce","added_by":"auto","created_at":"2026-01-28 09:51:58","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":95149,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 4. \u003c/strong\u003eGrayscale histograms of mildew simulation samples\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/640c2fd93be62bc86fceddbe.jpg"},{"id":101301379,"identity":"8bd95f9f-c521-48a0-a08c-30cff6a038f2","added_by":"auto","created_at":"2026-01-28 09:51:41","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":104853,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 5.\u003c/strong\u003e Average surface roughness of mildew simulation samples\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/17d5534825b73a217d9ed3c3.jpg"},{"id":101301458,"identity":"96b10e9c-a1a7-4630-ae77-0e744b9e6e2a","added_by":"auto","created_at":"2026-01-28 09:51:53","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":152339,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 6.\u003c/strong\u003e Wideband spectrogram of mildew simulation samples\u003c/p\u003e","description":"","filename":"7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/7d5bffe179bd8a3f855642b2.jpg"},{"id":101301292,"identity":"109d75e0-3fe1-497e-89d3-b00209c47fed","added_by":"auto","created_at":"2026-01-28 09:51:30","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":189118,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 7.\u003c/strong\u003e Sound intensity\u003cstrong\u003e \u003c/strong\u003econtours of mildew simulation samples\u003c/p\u003e","description":"","filename":"8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/5488fc1a6c14ffcd80e18060.jpg"},{"id":101301449,"identity":"e8ca1c44-31b7-4bb1-ae03-fc5d7954ef65","added_by":"auto","created_at":"2026-01-28 09:51:44","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":267579,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 8.\u003c/strong\u003e Formants plots of mildew simulation samples\u003c/p\u003e","description":"","filename":"9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/faeb65281a132f5ef755df36.jpg"},{"id":101302562,"identity":"173bf58a-dfa4-439d-aa5d-cba5673f667d","added_by":"auto","created_at":"2026-01-28 09:54:21","extension":"jpg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":80626,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 9.\u003c/strong\u003e Spectrogram parameters of mildew simulation samples (a: Center of gravity, Standard deviation, Skewness, Kurtosis; b: Band energy difference)\u003c/p\u003e","description":"","filename":"10.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/b2204f505aa593bdf5f5823c.jpg"},{"id":101301324,"identity":"8383cea9-ded7-49ce-ba86-1c71e01bb6d9","added_by":"auto","created_at":"2026-01-28 09:51:33","extension":"jpg","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":179626,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFig. 10. \u003c/strong\u003eSurface morphology of the audio samples before and after treated (a: 2D surface morphology of mildew sample; a1: 3D surface morphology of mildew sample; a2: SEM cross-section image of mildew sample; b: 2D surface morphology of sample after removing mildew; b1: 3D surface morphology of sample after removing mildew)\u003c/p\u003e","description":"","filename":"11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/5c5ba6470c547d8c7399534a.jpg"},{"id":101942879,"identity":"c4b7455d-f48a-4ce5-8b3d-883a4a131b08","added_by":"auto","created_at":"2026-02-05 09:39:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2239583,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8597238/v1/229e601f-c337-41ad-99ce-721325f6d207.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Research on the mechanism of mildew contamination affecting the sound quality of analog tape archives","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eIn the early 20th century, people began to attempt to record sound using magnetic materials. In 1935, the AEG company in Germany introduced the world's first tape recorder, utilizing magnetic tape with a paper base coated in iron oxide. This innovation significantly enhanced the recording and playback quality of magnetic tape, marking the official inception of magnetic tape-recording technology \u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]\u003c/sup\u003e. Analog audio tapes convert sound signals into continuously varying analog electrical signals and subsequently employ magnetic heads to record these electrical signals as variations in magnetism on the coating of the magnetic tape. Since 1947, it has been widely used in music dissemination, cultural popularization, scientific research, news gathering, etc., and has played an indispensable role in all aspects of people's lives and work\u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. The primary components of the tape consist of the magnetic layer and the substrate, while a back coating was introduced in the 1960s to enhance wear resistance\u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e. The magnetic layer is coated with a polyester-polyurethane adhesive with uniform distribution of magnetic particles. The substrate is the support of the magnetic layer, which provides mechanical properties and is responsible for the physical integrity of the tape, being mostly polyester film (polyethylene terephthalate)\u003csup\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe polyester polyurethane binders contained in the tapes, as well as other organic substances such as fatty acid esters, paraffin oils, siloxanes and fluorinated lubricants, are the main sources of nutrition for mildew \u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/sup\u003e. \u003cem\u003eAspergillus\u003c/em\u003e and \u003cem\u003ePenicillium\u003c/em\u003e are among the most common varieties in audiovisual materials \u003csup\u003e[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e. Other species of mildew including \u003cem\u003eAlternaria\u003c/em\u003e, \u003cem\u003eChaetomium\u003c/em\u003e, \u003cem\u003eStemphylium\u003c/em\u003e, \u003cem\u003eCladosporium\u003c/em\u003e, and \u003cem\u003eStreptomyces\u003c/em\u003e, are also frequently found on audiovisual materials. Most of the mildew is distributed on the edge of the substrate and on the surface of the magnetic layer (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Mildew growth can significantly affect the deterioration of magnetic media, thereby impairing the retrieval of information \u003csup\u003e[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]\u003c/sup\u003e. Additionally, mildew has the potential to contaminate other archival materials by releasing spores, which may also pose a health risk to archivists\u003csup\u003e[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eCurrent research on magnetic tapes has centered on the assessment of their chemical and physical properties. For instance, the morphology of the tapes, layer thickness, and physical damage were analyzed using Scanning Electron Microscopy (SEM) and Environmental Scanning Electron Microscopy (ESEM). This methodology was further enhanced by employing X-ray fluorescence (XRF) spectroscopy. The detection of chlorine facilitated the identification of PVC substrates, while the presence of silicon or fluorine provided insights into lubricant composition. This comprehensive approach enabled a preliminary analysis of tapes with incomplete documentation \u003csup\u003e[\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/sup\u003e. The primary mechanism of tape degradation and limited lifespan is the transformation of metal particles from ferromagnetic to non-ferromagnetic iron oxide \u003csup\u003e[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/sup\u003e. Spectroscopic analysis has proven instrumental in characterizing magnetic tape materials. an extensive library of ATR spectral data was established to identify characteristic bands for tape substrates and binders \u003csup\u003e[\u003cspan additionalcitationids=\"CR16 CR17\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]\u003c/sup\u003e. Building on this foundation, Hobaia demonstrated that ATR-FTIR spectroscopy could differentiate between SSS and non-SSS audio tapes and serve as a rapid tool for predicting the playability of tapes with no visible signs of aging. A methodology integrating Principal Component Analysis (PCA), Quadratic Discriminant Analysis (QDA), and K-means clustering was employed to classify infrared spectra based on playability. This approach successfully identified 93.78% of unplayable tapes within the calibration set. Thermal analysis provides another avenue for material identification and assessment. TGA degradation temperatures can assist in identifying the materials of unknown tapes. Davis further characterized the degraded tapes using DSC, observing an irreversible endothermic transition at approximately 50 ℃, which is typically absent in playable tapes.\u003c/p\u003e \u003cp\u003eThe degradation of the adhesive can lead to diseases such as cracking of the magnetic layer and shedding of magnetic powder. Meanwhile, investigations into binder degradation have employed both destructive and non-destructive techniques. Acetone can used to destructively extract soluble compounds formed by binder hydrolysis, thereby studying the kinetics of polyester polyurethane binder breakdown \u003csup\u003e[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]\u003c/sup\u003e. Headspace Solid-Phase Microextraction-Gas Chromatography-Mass Spectrometry (SPME-GC-MS) serves as a non-destructive method for profiling the Volatile Organic Compounds (VOCs) emitted by tapes following decades of both natural and artificial aging \u003csup\u003e[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]\u003c/sup\u003e. Despite these advanced diagnostic techniques, a critical gap remains: the impact of various degradation types on the actual audio quality is severely under-researched. Sound recording archives research on the effects of mildew on audio quality is still very limited. There is an urgent need to establish a parameter system for objectively evaluating sound quality changes. Such a system is crucial for assessing the effectiveness of tape restoration methods and, ultimately, for improving the recovery of our audio heritage. Building upon existing research, this study employs an integrated approach combining chemical characterization and acoustic measurement techniques to systematically investigate mildew erosion in audio materials. Through controlled simulation of mildew-contaminated media samples, the research establishes a scientific framework for objectively assessing mildew covers on sound quality preservation.\u003c/p\u003e"},{"header":"2. Material and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Material\u003c/h2\u003e \u003cp\u003eAn analog magnetic audiotape, distributed by Tianjin Audiovisual Co., Ltd. in 1991, was used. Potato Dextrose Agar (PDA) medium from Beijing Aobo Star Biotechnology Co., LTD. Lactic acid phenol cotton blue staining agent from Beijing Solaibao Technology Co., LTD. Penicillium and aspergillus versicolor were isolated and purified from the mildew on the surface of the audio tape.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Analog sample preparation for mildew\u003c/h2\u003e \u003cp\u003e \u003cb\u003ePreparation for the bacterial suspension\u003c/b\u003e \u003c/p\u003e \u003cp\u003ePenicillium and Aspergillus were cultured on PDA medium for a duration of 5 days. The mildew colony, along with a small amount of agar, was carefully scraped into a sample tube using an inoculation loop. Subsequently, normal saline was added to the tube, and the conidia of the mildew were thoroughly dispersed in the saline by shaking vigorously for 5 minutes. Following this process, the conidia were uniformly suspended in normal saline, and a bacterial suspension was obtained through simple filtration using gauze. The inclusion of a small amount of agar serves to provide essential nutrients, promote mildew growth on the tape, and reduce the preparation time required for creating mock samples intended for mildew attack testing.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePreparation for mildew simulation samples\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFirst, play the audio tape and select playback time than 5 s and label D0. The audio sample is labeled R-D0. Subsequently, the bacterial suspension was applied to sample D0. The tape was laid flat on the bench with the magnetic layer facing upward. Using a cotton swab, the bacterial suspension was carefully applied to the sample. Care was taken to control the volume of the applied solution and to prevent droplets from running off due to the smooth surface of the tape. After treatment, placed the sample in a well-ventilated area to dry. The same procedure was repeated on the audio tape at one-day intervals. Audio tapes exhibiting progressively increasing mildew coverage were designated as samples D1, D2, D3, D4, and D5, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003e. The corresponding audio recordings were labeled R-D1, R-D2, R-D3, R-D4, and R-D5.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Surface topography characterization of mildew simulation samples\u003c/h2\u003e \u003cp\u003e \u003cb\u003eEpson scanner\u003c/b\u003e (Epson Perfection V850 Pro Seiko Epson Co., Ltd.) was used to perform 8-bit scanning of the surface morphology of D0, D1, D2, D3, D4 and D5, with a background plate selected from a black cardboard sheet, to record the surface topography of the audio tape with different coverage degrees was recorded, and to analyze the gray-scale histograms extracted from the scanned photographs.\u003c/p\u003e \u003cp\u003e \u003cb\u003eLaser confocal scanning microscopy\u003c/b\u003e was used to characterize the mildew covered samples by roughness analysis. The VK-H1XMC multi-file analysis software was used to measure surface roughness parameters of mildew covered samples D0, D1, D2, D3, D4 and D5. Five points were taken for each sample, observed at every 1/6 of the length of samples, avoiding the edge of the tape. Each sample measurement data was subsequently analyzed using five key surface roughness parameters: texture aspect ratio of the surface (Str), peak density of the surface (Spc), developed interfacial area ratio (Sdr), maximum height (Sz), and arithmetic mean height (Sa). In the equations, z (x, y) denotes the signed normal distance between the reference surface and the scale-limited surface. Symbol A denotes the evaluated area's numerical measure.\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\varvec{S}\\varvec{a}=\\frac{1}{\\varvec{A}}\\underset{\\varvec{A}}{\\overset{}{\\iint\\:}}\\left|\\varvec{Z}\\left(\\varvec{x},\\varvec{y}\\right)\\right|\\varvec{d}\\varvec{x}\\varvec{d}\\varvec{y}$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:\\varvec{S}\\varvec{z}=\\varvec{S}\\varvec{p}+\\varvec{S}\\varvec{v}$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\:\\varvec{S}\\varvec{p}\\varvec{c}=-\\frac{1}{2}\\frac{1}{\\varvec{n}}\\sum\\:_{\\varvec{k}=1}^{\\varvec{n}}(\\frac{{\\partial\\:}^{2}\\varvec{z}\\left(\\varvec{x},\\varvec{y}\\right)}{\\partial\\:{\\varvec{x}}^{2}}+\\frac{{\\partial\\:}^{2}\\varvec{z}\\left(\\varvec{x},\\varvec{y}\\right)}{\\partial\\:{\\varvec{y}}^{2}})$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equd\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equd\" name=\"EquationSource\"\u003e\n$$\\:\\varvec{S}\\varvec{t}\\varvec{r}=\\frac{\\underset{{\\varvec{\\tau\\:}}_{\\varvec{x},}{\\varvec{\\tau\\:}}_{\\varvec{y}}\\in\\:\\varvec{R}}{\\mathbf{min}}\\sqrt{{{\\varvec{\\tau\\:}}_{\\varvec{x}}}^{2}+{{\\varvec{\\tau\\:}}_{\\varvec{y}}}^{2}}}{\\underset{{\\varvec{\\tau\\:}}_{\\varvec{x},}{\\varvec{\\tau\\:}}_{\\varvec{y}}\\in\\:\\varvec{Q}}{\\mathbf{max}}\\sqrt{{{\\varvec{\\tau\\:}}_{\\varvec{x}}}^{2}+{{\\varvec{\\tau\\:}}_{\\varvec{y}}}^{2}}}$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Eque\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Eque\" name=\"EquationSource\"\u003e\n$$\\:\\varvec{S}\\varvec{d}\\varvec{r}=\\frac{1}{\\varvec{A}}\\underset{\\varvec{A}}{\\overset{}{\\iint\\:}}（\\sqrt{\\left[1+（{\\frac{\\partial\\:\\varvec{z}\\left(\\varvec{x},\\varvec{y}\\right)}{\\partial\\:\\varvec{x}}）}^{2}+（{\\frac{\\partial\\:\\varvec{z}\\left(\\varvec{x},\\varvec{y}\\right)}{\\partial\\:\\varvec{y}}）}^{2}\\right]}-1）\\varvec{d}\\varvec{x}\\varvec{d}\\varvec{y}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eArithmetic Mean Height (Sa\u003cb\u003e)\u003c/b\u003e is defined as the arithmetic mean of the absolute deviations in vertical distance between the measured surface profile and the reference plane within the sampling area. As a height parameter, a higher Sa value indicates greater amplitude variation and increased surface roughness.\u003c/p\u003e \u003cp\u003eMaximum Height (Sz), represents the sum of the highest peak and deepest valley within the sampling area. This parameter quantifies the extreme vertical deviations of the surface.\u003c/p\u003e \u003cp\u003eTexture Aspect Ratio (Str) is a spatial parameter that characterizes surface isotropy or anisotropy, where \u003cb\u003eλ\u003c/b\u003e\u003csub\u003e\u003cb\u003emin\u003c/b\u003e\u003c/sub\u003e and \u003cb\u003eλ\u003c/b\u003e\u003csub\u003e\u003cb\u003emax\u003c/b\u003e\u003c/sub\u003e denote the dominant texture wavelengths perpendicular and parallel to the lay direction. Values approaching 0 indicate striated patterns, Str\u0026thinsp;\u0026asymp;\u0026thinsp;1 denotes isotropic topography.\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{S}\\varvec{t}\\varvec{r}=\\frac{\\varvec{\\lambda\\:}\\varvec{m}\\varvec{i}\\varvec{n}}{\\varvec{\\lambda\\:}\\varvec{m}\\varvec{a}\\varvec{x}}\\)\u003c/span\u003e \u003c/span\u003e \u003cb\u003e(0\u003c/b\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:\\le\\:\\varvec{S}\\varvec{t}\\varvec{r}\\le\\:1\\)\u003c/span\u003e \u003c/span\u003e \u003cb\u003e)\u003c/b\u003e \u003c/p\u003e \u003cp\u003eArithmetic Mean Peak Curvature (Spc) calculates the average principal curvature of the surface summits. As a feature shape parameter, lower Spc values indicate blunt or rounded asperities, whereas higher values correspond to sharp surface features that influence contact mechanics.\u003c/p\u003e \u003cp\u003eDeveloped Interfacial Area Ratio (Sdr) is a hybrid parameter that quantifies the percentage increase in surface area relative to an ideal flat plane:\u003cdiv id=\"Equf\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equf\" name=\"EquationSource\"\u003e\n$$\\:\\varvec{S}\\varvec{d}\\varvec{r}=\\left(\\frac{\\varvec{A}\\varvec{a}\\varvec{c}\\varvec{t}\\varvec{u}\\varvec{a}\\varvec{l}-\\varvec{A}\\varvec{p}\\varvec{r}\\varvec{o}\\varvec{j}\\varvec{e}\\varvec{c}\\varvec{t}\\varvec{e}\\varvec{d}}{\\varvec{A}\\varvec{p}\\varvec{r}\\varvec{o}\\varvec{j}\\varvec{e}\\varvec{c}\\varvec{t}\\varvec{e}\\varvec{d}}\\right)\\times\\:100\\varvec{\\%}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere A actual is the true surface area and A \u003csub\u003eprojected\u003c/sub\u003e is the projected area. An Sdr value\u0026thinsp;\u0026gt;\u0026thinsp;0 reflects surface porosity and roughness complexity, while Sdr\u0026thinsp;=\u0026thinsp;0 indicates perfect flatness.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Acoustic characterization of simulation samples\u003c/h2\u003e \u003cp\u003e \u003cb\u003ePortable tape recorder\u003c/b\u003e (6503, Nanjing Panda Electronics Co., Ltd.) was used to record audio in MP3 format before and after processing.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePraat speech analysis software\u003c/b\u003e (version Intel64) developed by the Institute of Phonetic Sciences at the University of Amsterdam's Faculty of Humanities was used. Praat is a comprehensive speech analysis software package developed and maintained by the Institute of Phonetic Sciences at the University of Amsterdam \u003csup\u003e[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]\u003c/sup\u003e. It enables a wide range of acoustic-phonetic analyses, including but not limited to fundamental frequency (F0), spectral characteristics, and formant tracking, in addition to voice analysis. Wideband spectrogram is a type of spectrogram optimized for high time resolution at the expense of frequency resolution.\u003c/p\u003e \u003cp\u003eThe center of gravity (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{f}}_{\\text{c}}\\)\u003c/span\u003e\u003c/span\u003e) is a measure for how high the frequencies in a spectrum are on average, expressed as \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{f}}_{\\text{c}}\\)\u003c/span\u003e\u003c/span\u003e in Hz.\u003cdiv id=\"Equg\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equg\" name=\"EquationSource\"\u003e\n$$\\:{\\varvec{f}}_{\\varvec{c}}={\\int\\:}_{0}^{\\varvec{\\infty\\:}}{\\varvec{f}\\mid\\:\\varvec{S}\\left(\\varvec{f}\\right)\\mid\\:}^{\\varvec{P}}\\varvec{d}\\varvec{f}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe standard deviation (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{S}\\text{D}\\)\u003c/span\u003e\u003c/span\u003e)is a measure for how much the frequencies in a spectrum can deviate from the center of gravity. Standard deviation represents the standard deviation in the spectrum, denoted by SD, and is used to measure the degree of dispersion of the frequency with respect to.\u003c/p\u003e \u003cp\u003eThe skewness (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{S}}_{\\text{k}}\\)\u003c/span\u003e\u003c/span\u003e) is a measure for how much the shape of the spectrum below the center of gravity is different from the shape above the mean frequency. Skewness represents the skewness in the spectrum, and is used to measure the direction and degree of skewness of the data distribution (relative to the standard normal distribution).\u003c/p\u003e \u003cp\u003eThe kurtosis (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{K}_{u}\\)\u003c/span\u003e\u003c/span\u003e) is a measure for how much the shape of the spectrum around the center of gravity is different from a Gaussian shape. Kurtosis denotes the degree of kurtosis in the spectrum, and is used as a measure of the sharpness of the data distribution (relative to the standard normal distribution), with a high kurtosis usually indicating a sharper distribution curve, often with more extreme values.\u003c/p\u003e \u003cp\u003eThe band energy difference (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{B}\\text{E}\\text{D}\\)\u003c/span\u003e\u003c/span\u003e) represents the energy difference between the low-frequency band (20\u0026thinsp;~\u0026thinsp;3000 Hz) and the high-frequency band (3000\u0026thinsp;~\u0026thinsp;6000 Hz).\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results and Discussion","content":" \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003ea, the characteristic peak at 630.8 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in the FTIR spectrum is attributed to the lattice vibration of Fe-O, indicating that the audio tape contains γ-Fe\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e particles. The peak at 1060.4 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is due to the stretching vibration of the C-O bond, and the peak at 1170.4 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is caused by the stretching vibration of the C-O bond connected to the acetyl group. The peak at 1725.5 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is attributed to the stretching vibration of the ester group (C\u0026thinsp;=\u0026thinsp;O), and the peak at 2921.6 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is the absorption peak caused by the -CH\u003csub\u003e2\u003c/sub\u003e in the polyester \u003csup\u003e[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]\u003c/sup\u003e, suggesting that the binder is a polyester type. Combined with the manufacturing process of the audio tape, it is speculated that the audio tape selected for this experiment contains a polyester polyurethane binder. It can be seen from Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, the peak at 723.6 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is attributed to the out-of-plane bending vibration of the benzene ring, the peak at 1097.5 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is due to the symmetrical stretching vibration of the C-O bond in the ethylene glycol segment, the peak at 1244.5 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is attributed to the asymmetrical stretching vibration of C-O-C, and the peak at 1714.1 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is due to the stretching vibration of the ester group (C\u0026thinsp;=\u0026thinsp;O), indicating that the audio tape selected for this experiment is made of PET (polyethylene terephthalate). Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003ec and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003ed shows the SEM micrographs of the cross-section of the audio tape. It shows that the audio tape is composed of three layers with an average thickness of approximately 3.5 \u0026micro;m, 8.0 \u0026micro;m, and 2.5 \u0026micro;m (from left to right). The energy spectrum indicates that the main element of the first layer structure is Fe, while the main elements of the second- and third-layer structures are C. From the comprehensive Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003ea-d, it can be concluded that the three layers from left to right are the magnetic layer, the base layer, and the back coating layer, respectively.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows the 2D surface morphology of localized areas of audio samples after mildew erosion during different stages of fungal erosion. The surface of the magnetic layer without mildew erosion (D0) is relatively smooth and contains fine black magnetic particles. When the mildew begins to erode the magnetic layer (D1), long and thin hyphae emerge, arranged in a grid pattern \u003csup\u003e[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]\u003c/sup\u003e. As time progresses and spore concentration increases (D2-D5), patchy aggregated colonies gradually form, leading to an increase in coverage of the magnetic layer. Notably, the mildew colonies on the D5 sample nearly encompass the entire surface of the magnetic layer \u003csup\u003e[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the grayscale histograms representing the gray levels of the scanned image and the corresponding number of pixels was used to represent the varying coverage degrees of the mildew simulation samples. The pixel distribution of the sample U0 was the most concentrated, mainly around the gray level of 25\u0026ndash;50, the corresponding number of pixels was about 14 x 10\u003csup\u003e5\u003c/sup\u003e. When the sample is covered with mildew (D1), there is a noticeable decrease in the number of pixels within the gray level range of 25\u0026ndash;50. This change is accompanied by a shift in the peak shape towards higher gray values, while an increase in the number of pixels within the gray level range of 50\u0026ndash;100 becomes evident. These observations indicate that the overall image brightness increases, resulting in diminished detail in darker areas. As the degree of coverage progresses from slight (D2) to severe (D5), there is a gradual decline in pixel count for the gray level range of 25\u0026ndash;50, contrasted by a steady rise in pixel count for the gray level range of 50\u0026ndash;100. This phenomenon can be attributed to the higher whiteness associated with mildew compared to that of the black magnetic layer found on audio tape. Consequently, as mildew coverage intensifies, there is an overall increase in sample whiteness, leading to enhanced brightness within the image \u003csup\u003e[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003e, the trend of average surface roughness change of mildew-covered samples at varying coverage levels can be clearly seen. The average Sa value of the surface of sample D0 is 2.34 \u0026micro;m (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). As the degree of mildew coverage increases from slight (D1) to severe (D5), the Sa value exhibits a monotonically increasing trend, indicating greater amplitude variation and enhanced surface roughness. Figure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eb illustrates the variation in Sz values for samples with differing degrees of mildew coverage. As the degree of mildew coverage progresses from D0 to D5, Sz demonstrates a clear upward trend. This suggests that increased mildew coverage results in an elevation in both the maximum height and deepest valley on the surface, revealing a strong positive correlation between Sz, Sa. Upon examining Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ec, it is evident that the Str values for all mildew-covered samples (D0 through D5) exceed 0.5 \u003csup\u003e[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]\u003c/sup\u003e, signifying that these sample surfaces exhibit more isotropic characteristics. Figure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ed and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ee have similar variation patterns. The changes from D1 to D3 are relatively gentle, and the overall trend is monotonically increasing. These results suggest that the growth of mildew on the magnetic layer surface is uneven, and the masking effect on the local magnetic layer is different \u003csup\u003e[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e6\u003c/span\u003e illustrates the wideband spectrograms of audio samples subjected to varying degrees of mildew coverage. These spectrograms are commonly employed to depict the energy distribution of sound signals across different frequencies over time. The depth of color typically signifies the intensity of energy, with darker hues indicating higher energy levels and lighter shades representing lower energy levels. From the wideband spectrogram corresponding to R-D0, it is evident that color depth fluctuates across various frequency bands. Specifically, in the low-frequency range of 1000\u0026ndash;2000 Hz, the coloration is darker compared to that in the high-frequency range, suggesting a predominant concentration of energy within this lower frequency spectrum. This phenomenon can be attributed to low-frequency sound waves possessing longer wavelengths, which render them more susceptible to diffraction or reflection; thus, they retain their energy during propagation \u003csup\u003e[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]\u003c/sup\u003e. When mildew begins to cover the sample (R-D1), all frequency bands exhibit a noticeable lightening in color within a span of 5 seconds, accompanied by a decrease in total or average sound energy. This observation indicates that mildew coverage diminishes both overall amplitude and volume of sound-transforming it from \"full-bodied\" (characterized by rich harmonics typical in music) into a \"thin\" quality-and results in reduced speech clarity. As surface coverage escalates from slight (R-D0) to severe (R-D5), there is an increasingly pronounced lightening effect observed across all frequency bands within 5 seconds. Notably, for sample R-D5, nearly all high-frequency energies ranging from 2000\u0026ndash;6000 Hz dissipate entirely while some degree of intensity persists within the low-frequency range. This suggests that high-frequency waves-with their shorter wavelengths-are more readily obstructed or scattered by mildew particles than their low-frequency counterparts, consequently leading to a swifter loss of energy \u003csup\u003e[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]\u003c/sup\u003e. High-frequency sounds contribute significantly to attributes such as brightness, clarity, and detail within auditory perception. The absence of energetic presence in this high-frequency domain manifests as continued existence but with timbral alterations shifting from \"bright\" towards \"dull,\" further accompanied by diminished clarity and loss of intricate details.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e7\u003c/span\u003e, the intensity curves of audio samples were exhibited with varying mildew-covered levels. In comparison to R-D0, the presence of mildew coverage diminishes the amplitude of the sound wave. The severity of this coverage correlates with a reduction in audio energy, with these changes being particularly pronounced at both peaks and troughs. This phenomenon occurs because as the magnetized audio tape moves toward the playback head, the fluctuating magnetic field generates magnetic flux, resulting in an output electrical signal that continuously varies to record information. When there is close contact between the head and the magnetic tape, magnetic coupling is enhanced. Consequently, more magnetic field lines traverse through the head coil, generating a stronger induced signal. However, increased mildew coverage elevates the distance between the playback head and the magnetic layer, thereby decreasing the number of magnetic field lines passing through the head coil. This results in a weakened induced signal and subsequently reduces sound volume. Furthermore, mildew coverage can alter both peak height and shape within intensity curves. For instance, as mildew coverage intensifies, a curve peak occurring around 1s becomes higher and sharper. It was indicated that sound transitions from low to high levels during this interval which may lead to noise formation. Moreover, the concentrated frequency point (peak) of energy at 2s shifts to the low-frequency direction, which may cause the attack time to be advanced.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eUsing Praat software to perform a Fourier transform allows for the decomposition of complex sound waves into simpler components, which are then distributed across various frequency bands to create a formant chart \u003csup\u003e[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]\u003c/sup\u003e. The horizontal axis represents time, while the vertical axis indicates the frequency of the sound wave. Formants correspond to specific regions in the sound spectrum where energy is relatively concentrated, and they appear as thick black \"horizontal bars\" in the spectrogram \u003csup\u003e[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]\u003c/sup\u003e. Figure\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e8\u003c/span\u003e illustrates the formant charts of audio samples exhibiting varying degrees of mildew coverage. The audio sample R-D0 displays relatively concentrated distributions of sound wave frequencies at 900\u0026ndash;1100 Hz, 2900\u0026ndash;3100 Hz, 3900\u0026ndash;4100 Hz, and 5000\u0026ndash;5200 Hz. As mildew coverage increases (D1 - D5), the high-frequency changes are obvious, but the shape and distribution range of the formants do not change much overall. On one hand, mildew presence obstructs direct contact between the magnetic head and magnetic layer, leading to attenuation of remanent magnetic signals read by the magnetic head. Given that high-frequency signals possess shorter wavelengths, these changes become more pronounced at higher frequencies. On the other hand, the presence of hypha reduces the smoothness of the surface of the audio tape, and the local magnetic field distribution changes during playback, causing changes in the frequency distribution. However, since the magnetic domain disorder is random, the impact on signals of different frequencies is relatively consistent, so the shape and distribution range of the formants do not change much overall.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e9\u003c/span\u003ea illustrates the variations in 5s spectrogram parameters (center of gravity, standard deviation, skewness, kurtosis and band energy difference) of mildew-covered audio samples under varying coverage levels. As the coverage changes from slight (D0) to severe (D5), center of gravity and standard deviation of the 5s audio spectrogram decrease monotonically, and the skewness and kurtosis increase monotonically. This is because the overlay process attenuates the remanence signal read by the magnetic head, some of the sound information lost. The loss of part of the frequency information during reading directly leads to a decreasing trend in the standard deviation of the spectrogram, and the curve becomes sharper and sharper, which is manifested as higher and higher kurtosis. Due to the short wavelength of the high frequency, the coverage treatment has a greater impact on the high frequency region, so the average frequency shows a decreasing trend with the increase of mildew coverage \u003csup\u003e[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]\u003c/sup\u003e. In addition, due to the higher energy in the low-frequency region, more information in the low-frequency region is lost due to the coverage processing, so the peak area in the spectrogram is positively biased and the skewness increases. As evidence from the Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e10\u003c/span\u003eb, with the increase of coverage, the energy difference between the low frequency region (20-3000 Hz) and the high frequency region (3000\u0026ndash;6000 Hz) becomes larger, indicating that the treatment of surface anisotropy will increase the energy difference to a certain extent.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e10\u003c/span\u003e, surface morphology of the mildew corrosion sample (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e10\u003c/span\u003ea and a\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) and sample after removing mildew (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e10\u003c/span\u003eb and b\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) reveal notable. In Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e10\u003c/span\u003ea and a\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, fungal hyphae proliferate across the tape surface, forming an interconnected network. The colonies form pitting on the surface of the magnetic layer. It is worth noting that after removing the mildew, we observe not only localized pitting but also extensive irregular line-like corrosion in severely affected areas on the magnetic layer's surface (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e10\u003c/span\u003eb). As can be seen from the 3D surface morphology of sample after removing mildew (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e10\u003c/span\u003eb1), these pitting or linear corrosion patterns exhibit a concave structure. Based on the SEM cross-section image of mildew sample (Fig.\u0026nbsp;10a2), it can be concluded that during its reproductive process, mildew mycelium penetrates into the magnetic layer, resulting in both pitting and linear concave structures due to corrosion. This suggests that as mildew metabolizes, it produces organic acids which corrode the recording medium, thereby compromising the durability of the magnetic carrier. Mildew degradation affects components of this medium leading to structural damage. Additionally, it causes surface pitting and movement of magnetic particles which ultimately results in distortion and attenuation of magnetic recording signals.\u003c/p\u003e"},{"header":"4. Conclusions","content":"\u003cp\u003eIn this study, a combination of chemical characterization and acoustic measurement to systematically evaluate the impact of mildew contamination on audio materials was explored. The common strain Penicillium and Aspergillus was used to prepare mildew corrosive samples with different coverage degrees, and the corresponding audio samples were analyzed by Praat. Refine the data on surface structure and changes in acoustic information, it was found that the coverage of mycelium on the magnetic layer surface was uneven and would cause the image to brighten. On the one hand, the energy was mainly concentrated in the low-frequency range (1000\u0026ndash;2000). The coverage of mildew would cause the total energy or average energy of the sound would decrease. This indicates that the covering effect of mildew would reduce the overall amplitude or volume of the sound and decrease the clarity of speech. Notably, the energy loss in the high-frequency range was faster than that in the low-frequency range, resulting in detail loss and reduced intelligibility. On the other hand, it would cause changes in the peak values of the sound intensity curve or shifts in the peak shape, leading to noise or alterations in the attack time.\u003c/p\u003e \u003cp\u003eFrom the parameter change graph of the 5-second spectrum of the sample, it can be concluded that the average frequency and standard deviation monotonically decreased, while the skewness and kurtosis monotonically increased. The above information indicates that the covering effect of mildew increases the distance between the magnetic head and the magnetic layer, resulting in weakened induction signals and the loss of some sound information. Combining the surface and cross-sectional morphology and roughness information, mildew degradation affects the components of this medium, leading to surface pitting and movement of magnetic particles, which ultimately results in distortion and attenuation of magnetic recording signals. Based on the information presented above, the dual effects of mildew coverage and corrosion cause a decrease in sound clarity and distortion. This work highlights the necessity of mildew removal and provides valuable theoretical references for the evaluation of sound information after the removal of contaminants from magnetic materials in the future.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThe authors acknowledge financial support from the National Natural Science Foundation of China (22572112), National Natural Science Foundation of China (22002080), Science and Technology Project of the National Archives Administration (2022-B-005), The Key Research and Development Program of Shaanxi Province (2021GY-172), The Key Scientific Research Project at the Museum Level of Hubei Provincial Museum (25A05)\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eJ.Z.H. and C.L. provided research design, research guidance, data analysis and writing Original Draft; X.H.P. and L.Y.H. provided research guidance; W.Y.N. and Z.Y.H. participated in data analysis; Y.S.J. was responsible for research design and data collection; C.Q.F. and L.B.D involved in data collection.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBressan F, Hess RL, Sgarbossa P. Chemistry for Audio Heritage Preservation: A Review of Analytical Techniques for Audio Magnetic Tapes[J]. Heritage. 2019;2(2):1551\u0026ndash;87.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo LH. Development and Applications of Magnetic Recording[J]. Wuli (Physics). 1984;13(3):145\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBereijo A. The conservation and Preservation of Film and Magnetic Materials (2): Magnetic Materials[J]. Libr Rev. 2004;53(7):372\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLantz MA, Furrer S, Peterman M. Magnetic Tape Storage Technology[J]. ACM Trans Storage. 2025;21(1):1\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBogart JWV. Magnetic Tape Storage and Handling: A Guide for Libraries and Archives[M]. ERIC. 1995;1:34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeitkamp MA, Freeman JP, McMillan DC. Fungal Metabolism of tert-Butylphenyl Diphenyl Phosphate[J]. Appl Environ Microbiol. 1985;50(2):265\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCappitelli F, Sorlini C. From Papyrus to Compact Disc: The Microbial Deterioration of Documentary Heritage[J]. Crit Rev Microbiol. 2005;31(1):1\u0026ndash;10.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVivar I, Borrego SGE. Fungal Biodeterioration of Color Cinematographic Films of the Cultural Heritage of Cuba[J]. Int Biodeterior Biodegrad. 2013;84(3):372\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoleymani S, Russ L. Mould on Magnetic Media: What Are the Current Preservation Practices by Audiovisual Conservation Practitioners?[J]. Stud Conserv. 2023;68(7):704\u0026ndash;19.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichardson E, Giachet M, Schilling M. Assessing the Physical Stability of Archival Cellulose Acetate Films by Monitoring Plasticizer Loss[J]. Polym Degrad Stab. 2013;107(2):231\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBressan F, Bertani R, Furlan C, An ATR-FTIR. ESEM Study on Magnetic Tapes for the Assessment of the Degradation of Historical Audio Recordings[J]. J Cult Herit. 2016;18(1):313\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBressan F, Canazza S, Bertani R. A study on Thermal Treatment for the Recovery of Magnetic Tapes Affected by Soft Binder Syndrome-Sticky Shed Syndrome[J]. IASA J. 2015;44(1):53\u0026ndash;64.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavis AR, Monroe E, France FG. Understanding Magnetic Tape Degradation by Polymeric and Material Testing[C]. \u003cem\u003eProceedings of\u003c/em\u003e 2018 \u003cem\u003eAES International Conference on Audio Archiving, Preservation \u0026amp; Restoration, Audio Engineering Society\u003c/em\u003e, 2018.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJudge JS, Schmidt RG, Weiss RD. Media Stability and Life Expectancies of Magnetic Tape for Use with IBM 3590 and Digital Linear Tape Systems[C]. \u003cem\u003eProceedings of the 20th IEEE/11th NASA Goddard Conference on Mass Storage Systems and Technologies (MSST 2003)\u003c/em\u003e, 2003, 97\u0026ndash;100.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026oacute;mez-S\u0026aacute;nchez E. ATR-FTIR Spectroscopy for the Characterization of Magnetic Tape Materials[J]. E Preservation Sci. 2011;8(1):2\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCassidy BM, Lu Z, Fuenffinger NC. Minimally Invasive Identification of Degraded Polyester-Urethane Magnetic Tape Using Attenuated Total Reflection Fourier Transform Infrared Spectroscopy and Multivariate Statistics[J]. Anal Chem. 2015;87(18):9265\u0026ndash;72.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHobaica S. Analysis of Audio Magnetic Tapes with Sticky Shed Syndrome by ATR-FTIR[J]. J Appl Polym Sci. 2013;128(3):1962\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang WG, Ha JH, Kim SG. Spectroscopic Determination of Alkyl Resorcinol Concentration in Hydroxyapatite Composite[J]. J Anal Sci Technol. 2016;7(1):9\u0026ndash;16.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKatayama K, Chinda Y, Shimizu O. Long-Term Stability of Magnetic Tape for Data Storage Under an Accelerated Condition[J]. IEEE Trans Magn. 2016;52(7):1\u0026ndash;4.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThi\u0026eacute;baut B, Vilmont LB, Lav\u0026eacute;drine B. Characterization of U-matic videotape Deterioration by Size Exclusion Chromatography and Pyrolysis Gas Chromatography/mass Spectrometry and the Role of Adipic Acid[J]. J Cult Herit. 2009;10(2):183\u0026ndash;97.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThi\u0026eacute;baut B, Lattuati-Derieux A, Hocevar M. Application of Headspace SPME-GC-MS in Characterisation of Odorous Volatile Organic Compounds Emitted from Magnetic Tape Coatings Based on Poly(urethane-ester) After Natural and Artificial Ageing[J]. Polym Test. 2007;26(2):243\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoersma P, Weenink D. Praat, a System for Doing Phonetics by Computer[J]. Glot Int. 2001;5(3):341\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBressan F, Bertani R, Furlan C. An ATR-FTIR and ESEM Study on Magnetic Tapes for the Assessment of the Degradation of Historical Audio Recordings[J]. J Cult Herit. 2016;18(3):313\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCassidy BM, Lu Z, Fuenffinger NC. Minimally Invasive Identification of Degraded Polyester-Urethane Magnetic Tape Using Attenuated Total Reflection Fourier Transform Infrared Spectroscopy and Multivariate Statistics[J]. Anal Chem. 2015;87(18):9265\u0026ndash;72.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatsky L. Mildew Treatment for Analog Audio Tapes: A Technical Guide[C]. \u003cem\u003eProceedings of the Association for Recorded Sound Collections (ARCS) Annual Conference\u003c/em\u003e, 2014.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePelaez J, Torres C, Ortiz JE. Efecto de Dos Campos Magn\u0026eacute;ticos sobre el Crecimiento Micelial y Propiedades Sensoriales del Hongo Pleurotus Ostreatus[J]. Revista Colombiana de Ciencias Hort\u0026iacute;colas. 2013;7(1):89\u0026ndash;97.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHussein I, Hasan R. The Effect of Zirconium Oxide Nanoparticle on the Tear Strength of Maxillofacial Silicone[J]. Al-Rafidain Dent J. 2021;21(3):193\u0026ndash;201.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRashidi M. Effect of Zirconium Oxide-Titanium Dioxide Nanoparticles on Mechanical and Physical Properties of Soft Denture Lining Materials[J]. J Nanostruct. 2022;12(1):1\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOelze T. Sound Reflection and Energy Loss at Boundaries[J]. Appl Acoust. 2018;137(3):211\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInspur General Software Co. Ltd. Lightweight High-Frequency Speech Restoration Method, System, Device, and Medium: CN118072750A[P]. 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKing R, Grau-Bov\u0026eacute; J, Curran K. Plasticiser Loss in Heritage Collections: Its Prevalence, Cause, Effect, and Methods for Analysis[J]. Herit Sci. 2020;8(1):123\u0026ndash;31.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu QQ, Li XG. Research Progress on Mildew Corrosion of Metals and Their Protective Layers in Atmospheric Environment[J]. J Univ Sci Technol Beijing. 2017;39(10):1463\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-heritage-science","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"hsci","sideBox":"Learn more about [Heritage Science](http://heritagesciencejournal.springeropen.com)","snPcode":"40494","submissionUrl":"https://submission.nature.com/new-submission/40494/3","title":"npj Heritage Science","twitterHandle":"@SpringerOpen","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"sound quality, magnetic audio tape, mildew contamination","lastPublishedDoi":"10.21203/rs.3.rs-8597238/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8597238/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSince 1947, analog tape recording has become the main method of sound recording, and has been widely used in music dissemination, cultural popularization, scientific research, news interviews, etc. For more than 40 years, giving birth to a large number of important audio recording archives. Due to the presence of a large amount of adhesive in the tape, mildew have been found on the surface of the magnetic layer and the edges of many tapes, posing a serious threat to the long-term preservation and the sound quality of the recordings. In this paper, ATR-FTIR and SEM were used to extensively characterize the chemical and physical of magnetic tape materials. To assess the effects of mildew contamination, the common strain \u003cem\u003ePenicillium and Aspergillus\u003c/em\u003e was used to prepare mildew corrosive samples with different coverage degrees, and the corresponding audio samples were analyzed by Praat, a professional speech analysis software. The relationship between wideband spectrogram, sound intensity contours, formants plots, center of gravity, standard deviation, skewness, kurtosis, band energy difference and coverage degree of mildew contaminated samples were analyzed. Additionally, the surface roughness and morphology of the analog tapes were observed using the laser microscopy system. Based on the above information, the mechanism by which the dual effects of mildew coverage and corrosion affect sound quality was revealed. This research provides a theoretical foundation for improved and restored strategies to mitigate mildew damage contaminated analog tape archives in the future.\u003c/p\u003e","manuscriptTitle":"Research on the mechanism of mildew contamination affecting the sound quality of analog tape archives","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-28 09:36:24","doi":"10.21203/rs.3.rs-8597238/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-09T05:40:12+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-18T07:40:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"146570732616748892001372455133015330942","date":"2026-01-27T13:38:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"255926677748846710651626298886781622403","date":"2026-01-23T10:21:46+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-01-23T09:43:37+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-20T04:18:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-20T04:17:51+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Heritage Science","date":"2026-01-14T03:58:57+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-heritage-science","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"hsci","sideBox":"Learn more about [Heritage Science](http://heritagesciencejournal.springeropen.com)","snPcode":"40494","submissionUrl":"https://submission.nature.com/new-submission/40494/3","title":"npj Heritage Science","twitterHandle":"@SpringerOpen","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"854042ec-7c60-4652-8c4d-105396e61255","owner":[],"postedDate":"January 28th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-20T14:27:31+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-28 09:36:24","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8597238","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8597238","identity":"rs-8597238","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}