This audio file editor is designed to study the spectra of short duration fragments of audio files. Spectra can be viewed in a time window of up to 1 ms.
Construct and study spectra of speech sounds for audio fragments of up to 1 ms
Select speech pauses in the audio file automatically
Calculate signal-to-noise ratio (SNR) automatically
Use “human ear” filter
Solveyg audio editor relies on the methodology of spectral transformations for the construction and the following study of spectra for speech sound patterns. When constructing the spectra, a discrete non-orthogonal Morlet transform is used. This allows to obtain spectra with a frequency step of 1 Hz., even for shorter audio fragments with the duration less than 20 ms.
While using the orthogonal discrete Fourier transform, the number of discrete frequencies of the spectrum is determined by the audio fragment’s length. For the duration of 20 ms the Fourier spectrum has components with a multiplicity of 50 Hz while auditory analyzers of the human brain only have a frequency resolution of 1 Hz.
This is why “human ear” filter is one of the important features of the Solveyg editor. This filter simulates the dependence of the amplitude of the spectrum of the sounds of the human auditory apparatus as a function of the conversion frequency.