Audio files for spectrum analysis demonstration

This data set consists of 6 real-world audio files (in .wav format, 48000Hz, mono) that are carefully crafted as examples for spectral analysis (i.e for teaching or as sample test data for algorithms). The files have clear discernible sound with an added true random background ambient noise (composed of: distant fan noise + distant street traffic + close harddisk clicking noise). The sound is clearly discernible by a human, despite the noise. - There are some musical sounds (the note G3 on several instruments; fundamental frequency 196Hz) on a tubular bell, classical piano, trumpet, violin. The sounds of the instrument was generated from MIDI banks with FluidSynth software, played on a loudspeaker and re-recorded with an analogical microphone (with the ambient noises). Audio processing was performed with Tenacity software; - Sample from human speech (wovel “o”), with the same processing as above; - The “Noise” file is purely digitally generated (white noise). Each audio set is composed of three files: - The audio file (*.wav), each sampled at 48000 Hz, Mono. - an amplitude file (*_amplitude.csv, corresponding linear amplitudes recorded by the microphone of the .wav file). Numeric format in simple text format (.csv) with labeled column names. -a spectrum file (*_spectrum.csv, frequency/amplitude(dB) ) with the results of a FFT (Fast Fourier Transform). Numeric format in simple text format (.csv) with labeled column names. Detailed description of each set is provided below. Bell_G3.wav Bell_G3_amplitude.csv: Length processed: 113851 samples 2.37190 seconds. Sample Rate: 48000 Hz. Sample values on linear scale. 1 channel (mono). Length processed: 113851 samples, 2.37190 seconds. Peak amplitude: 0.59001 (linear) -4.58276 dB. Unweighted RMS: -19.87796 dB. DC offset: 0.00069 linear, -63.18732 dB. Bell_G3_spectrum.csv: FFT transform (Hz / dB) Noise.wav Noise_amplitude.csv: Length processed: 31765 samples 0.66177 seconds. Sample Rate: 48000 Hz. Sample values on linear scale. 1 channel (mono). Length processed: 31765 samples, 0.66177 seconds. Peak amplitude: 0.52797 (linear) -5.54788 dB. Unweighted RMS: -16.61153 dB. DC offset: -0.00013 linear, -77.53051 dB Noise_spectrum.csv) FFT transform (Hz / dB) Piano_G3.wav Piano_G3_amplitude.csv: Sample Rate: 48000 Hz. Sample values on linear scale. 1 channel (mono). Length processed: 133063 samples, 2.77215 seconds. Peak amplitude: 0.41070 (linear) -7.72946 dB. Unweighted RMS: -23.95101 dB. DC offset: 0.00030 linear, -70.58058 dB. Piano_G3_spectrum.csv: FFT transform (Hz / dB) Trumpet_G3.wav Trumpet_G3_amplitude.csv: Sample Rate: 48000 Hz. Sample values on linear scale. 1 channel (mono). Length processed: 66931 samples, 1.39440 seconds. Peak amplitude: 0.29551 (linear) -10.58853 dB. Unweighted RMS: -22.00611 dB. DC offset: 0.00076 linear, -62.39740 dB. Trumpet_G3_spectrum.csv: FFT transform (Hz / dB) Violin_G3.wav Violin_G3_amplitude.csv: Sample Rate: 48000 Hz. Sample values on linear scale. 1 channel (mono). Length processed: 59252 samples, 1.23442 seconds. Peak amplitude: 0.41633 (linear) -7.61119 dB. Unweighted RMS: -18.36738 dB. DC offset: 0.00021 linear, -73.46784 dB. Violin_G3_spectrum.csv: FFT transform (Hz / dB) Wovel_O.wav Wovel_O_amplitude.csv Sample Rate: 48000 Hz. Sample values on linear scale. 1 channel (mono). Length processed: 4975 samples, 0.10365 seconds. Peak amplitude: 0.46174 (linear) -6.71214 dB. Unweighted RMS: -14.88086 dB. DC offset: 0.00009 linear, -80.54917 dB. Wovel_O_spectrum.csv: FFT transform (Hz / dB) These files are created by A. Iftime and released under Creative Commons Licence, 2024. You might cite the dataset as: “Audio files for spectrum analysis demonstration” [dataset] (2024), in “Medical Biophysics for 1st year medical students”, by Călinescu O., Babeș R., Iftime A., Băran I., Ionescu D., Ganea C., in publishing本数据集包含6个真实世界音频文件（格式为.wav，采样率48000Hz，单声道），经精心制作，可作为频谱分析的示例素材，用于教学或作为算法测试样本数据。 这些文件带有清晰可辨的主体声音，并叠加了真实随机的背景环境噪声，噪声由远处风扇声、远处街道交通声以及近距离硬盘点击声混合而成。即便存在噪声，人类仍可清晰辨识其中的主体声音。 本数据集包含四类乐器声音：管钟、古典钢琴、小号、小提琴演奏的G3音符（基频196Hz）。上述乐器的声音均通过FluidSynth软件从MIDI音色库生成，经扬声器播放后，使用模拟麦克风重新录制，同时收录环境噪声。音频后期处理使用Tenacity软件完成。 此外包含人类语音样本：元音“o”的录音，采用与上述音频相同的处理流程。 另有“噪声”文件：纯数字生成的白噪声。 每个音频组均包含三个文件： 1. 音频文件（*.wav）：均为48000Hz采样率的单声道音频； 2. 振幅文件（*_amplitude.csv）：对应.wav文件的麦克风录制线性振幅数据，采用带标注列名的纯文本CSV格式存储； 3. 频谱文件（*_spectrum.csv）：包含快速傅里叶变换（Fast Fourier Transform, FFT）得到的频率/振幅（dB）结果，同样采用带标注列名的纯文本CSV格式存储。 以下为各样本的详细参数： ### Bell_G3.wav 对应 Bell_G3_amplitude.csv 处理后时长为113851个采样点，即2.37190秒。采样率48000Hz，样本值为线性刻度，单声道。峰值振幅为0.59001（线性刻度），对应-4.58276 dB。未加权均方根振幅为-19.87796 dB。直流偏移为0.00069（线性刻度），对应-63.18732 dB。Bell_G3_spectrum.csv：FFT变换结果（格式为Hz / dB）。 ### Noise.wav 对应 Noise_amplitude.csv 处理后时长为31765个采样点，即0.66177秒。采样率48000Hz，样本值为线性刻度，单声道。峰值振幅为0.52797（线性刻度），对应-5.54788 dB。未加权均方根振幅为-16.61153 dB。直流偏移为-0.00013（线性刻度），对应-77.53051 dB。Noise_spectrum.csv：FFT变换结果（格式为Hz / dB）。 ### Piano_G3.wav 对应 Piano_G3_amplitude.csv 采样率48000Hz，样本值为线性刻度，单声道。处理后时长为133063个采样点，即2.77215秒。峰值振幅为0.41070（线性刻度），对应-7.72946 dB。未加权均方根振幅为-23.95101 dB。直流偏移为0.00030（线性刻度），对应-70.58058 dB。Piano_G3_spectrum.csv：FFT变换结果（格式为Hz / dB）。 ### Trumpet_G3.wav 对应 Trumpet_G3_amplitude.csv 采样率48000Hz，样本值为线性刻度，单声道。处理后时长为66931个采样点，即1.39440秒。峰值振幅为0.29551（线性刻度），对应-10.58853 dB。未加权均方根振幅为-22.00611 dB。直流偏移为0.00076（线性刻度），对应-62.39740 dB。Trumpet_G3_spectrum.csv：FFT变换结果（格式为Hz / dB）。 ### Violin_G3.wav 对应 Violin_G3_amplitude.csv 采样率48000Hz，样本值为线性刻度，单声道。处理后时长为59252个采样点，即1.23442秒。峰值振幅为0.41633（线性刻度），对应-7.61119 dB。未加权均方根振幅为-18.36738 dB。直流偏移为0.00021（线性刻度），对应-73.46784 dB。Violin_G3_spectrum.csv：FFT变换结果（格式为Hz / dB）。 ### Wovel_O.wav 对应 Wovel_O_amplitude.csv 采样率48000Hz，样本值为线性刻度，单声道。处理后时长为4975个采样点，即0.10365秒。峰值振幅为0.46174（线性刻度），对应-6.71214 dB。未加权均方根振幅为-14.88086 dB。直流偏移为0.00009（线性刻度），对应-80.54917 dB。Wovel_O_spectrum.csv：FFT变换结果（格式为Hz / dB）。 本数据集由A. Iftime制作，2024年以知识共享（Creative Commons）许可协议发布。该数据集的引用格式为："用于频谱分析演示的音频文件"[数据集] (2024)，出自《一年级医学生医用生物物理学》，作者为Călinescu O.、Babeș R.、Iftime A.、Băran I.、Ionescu D.、Ganea C.，待出版。