TF32 is the time-frequency analysis software program for 32-bit Windows (95/98/NT/2000/XP) for the analysis of speech and other audio-frequency waveforms. TF32 supercedes CSpeechSP for DOS.

TFpamph.pdf gives current prices.

TF32.pdf explains how to use TF32.

TF32.exe downloads TF32 Demo Level (Revised July 26, 2005), which opens all the popular sound file formats and plays back waveforms through any Windows-compatible sound card. Recording and saving waveforms to files requires the Basic Level or Lab Automation Level -- click here for prices. The July 9, 2003 revision corrected the numeric readouts of waveform and analysis displays regarding the initial position of the analysis interval . The May 7, 2004 revision introduced an improved epoch marking algorithm used in pitch analysis and formant tracking. It is possible to revert to the earlier-version epoch marking algorithm if needed.

TF32 (Demo Level and higher) displays multiple waveform channels, pitch, RMS/dB trace, LPC inverse filter, time-frequency gray-scale digital sound spectrogram with editable LPC formant track overlay, text labels, time-slice spectrum (Fourier, LPC, moments), and voice perturbation measures of jitter, shimmer and periodicity SNR. Waveforms, spectrogram plots and numeric text readouts may be copied and pasted into Windows-compatible word processors for preparing publication manuscripts. This is copyrighted software; users of the Demo Level may publish displays and data extracted from TF32; all that is required is that you cite TF32 as the origin.

E-mail cspeech@chorus.net for more information.